A MOLECULAR ANALYSIS OF THE OESTROGEN RECEPTOR AND
SURROUNDING CHROMOSOMAL REGION IN EARLY BREAST CANCER
Thesis submitted for the degree of
Doctor of Philosophy
at the University of Leicester
by
Stephen Chappell
BSc(Hons) (Sheffield Polytechnic), MSc (Leicester)
Department of Pathology
University of Leicester
October 1997
UMI Number: U116227
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI U116227
Published by ProQuest LLC 2013. Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
For Annie Chappell, Linda Chappell
and Michelle Bickerdike
I dreamt that I dwelt in marble halls
With vassals and serfs at my side
And o f all who assembled within those walls
That I was the hope and the pride
I had riches too great to count, could boast
O f a high ancestral name
But I also dreamt, which pleased me most
That you loved me just the same
(James Joyce, Dubliners)
ABSTRACT
A Molecular Analysis Of The Oestrogen Receptor And Surrounding Chromosomal
Region In Early Breast Cancer
Stephen Chappell
Despite extensive research, the pathways of breast cancer development remains largely
unknown. Identification of key genetic alterations, particularly at early stages of the disease,
are central to elucidating the development of the disease.
Analyses for abnormal forms of the oestrogen receptor (ER) were performed to determine if
they may be functionally important in the evolution o f breast cancer. Forty four
mammographically detected, node negative, 15 mm or less, moderately or well differentiated,
invasive carcinomas were screened by RT-PCR and single stranded conformational
Polymorphism analysis, identifying variant forms of exon 3, 5, and 7 in addition to 3 missense
mutations. ER splice variants were detected in 15/44 tumours and included exon 5 (2/44), exon
3 (4/44) and exon 7 (13/44). 10/37 grade I and II tumours had variants compared with 5/7
grade III tumours. In addition, 11/39 ER+ and either progesterone (PgR)+ or PgR- tumours
had variants compared with 4/5 ER-/PgR- tumours.
Pure populations of tumour cells were microdissected from well defined groups of breast
lesions including ductal carcinoma in situ (DCIS), tubular carcinomas, and a similar tumour
group screened for variant ERs, and analysed for alterations to polymorphic microsatellite
sequences within, and flanking the ER gene (6q25.1-27) itself. This enabled evaluation of loss
o f heterozygosity (LOH) indicating the presence of tumour suppressor genes, and
microsatellite instability (MI) indicative of a mutator phenotype in colorectal cancers.
Frequent LOH, apart from at the M6P/IGF2R locus, was observed for all types and grades of
disease studied; identifying loss in this region as an early event involved in breast
development/progression. In comparison, MI at multiple loci and LOH at the M6P/IGF2R
were only found in high grade DCIS. LOH at this chromosomal interval in all types and grade
o f disease suggests inactivation of other as of yet uncharacterised tumour suppressor genes
mapped on chromosome 6q25.1-27.
A cknow ledgem ents
First and foremost I would like to say a heartfelt thank you to my supervisor Dr Rosemary
Walker, whose unwavering belief in me, and whose guidance, support, and understanding has
been a constant source of encouragement throughout my time at the Breast Cancer Research
Unit. I would also like to thank Dr Jacqui Shaw for her encouragement, frequent advice and
invaluable assistance and patience during the writing of this thesis, which without whose help
would have taken considerably longer to complete.
I would like to convey my dearest thanks to two individuals who I have been most fortunate to
work alongside. To Dr Tom Walsh, for his continuing friendship, camaraderie and unparalleled
support and encouragement, without which I would not have been able to achieve as much as I
have. Likewise to Dr Louise Jones, “Young Jonsie”, for her friendship, advice, answers to
numerous questions pathological, and for her infectious and seemingly unquenchable
enthusiasm for research.
Many thanks to the numerous individuals for friendship and frequent advice, from the BCRU,
the department of pathology, and other departments at the Glenfield General Hospital,
especially Sheila Dearing, Dr Howyda Hassan, Dr Steve Fenwick (Orthopaedics), Gilbert
Evans, Dr Gareth Bicknall (for numerous PC tutorials), Emma Lees, Dr Debbie Ireland and Dr
Alison Woods (Cardiology), and intercalated BSc students Matthew Brooks and Karen-Ann
Scott.
I'd like to say a special thank you to Dr Anne Willis in the Department of Biochemistry, for her
encouragement, understanding, support and willingness to take me on board during the writing
of this thesis.
I would like to expresses my most deepest love, thanks and appreciation to my family. To my
mother, Annie Chappell, and my sister Linda Chappell. I can never thank you enough for your
unquestioning love and support throughout my whole life. Mum, your boy finally done good!!
Also love and thanks to the other members of my family who have had such a profound
influence on my life. In particular my grandparents, John and Mary Joyce, and numerous
aunts, uncles and cousins, especially, Brian, Elizabeth, Colin and Paul Chappell and John and
Maria Joyce.
Finally, to my fiancee Michelle Bickerdike for her complete love, support (both emotional and
financial) and total belief in me, without which I would never have completed my PhD. Foxy, I
love you lots!!
L IS T O F F I G U R E S
F igure 1.1 Normal structure o f the human mammary gland
2
F igure 1.2 Genetic alterations associated with colorectal carcinogenesis
26
F igu re 1.3 Caretaker and Gatekeeper pathways o f inherited susceptibility to cancer
28
F igure 1.4 Hypothetical m odel for infiltrating ductal carcinoma
30
F igure 1.5 M odel o f breast cancer pathways
31
F igure 2.1 Schem atic representation o f the 5 ’ region o f the oestrogen receptor
gene
42
F igure 2.2 Organization o f the human oestrogen receptor gene
44
F igure 2.3 The hypothetical structure o f the human oestrogen receptor DNA binding domain
(region C am ino acids 180-262)
47
F igure 2.4 R ole o f the oestrogen receptor in gene activation
59
F igure 2.6 A garose electrophoresis o f R NA isolated using TR lzol reagent from T47-D breast
carcinoma cell line stained with ethidium bromide
107
F igu re 2.7(a-f) Agarose electrophoresis o f oestrogen receptor PCR products
109
F igu re 2.8 Agarose electrophoresis o f ER S et 2 nested PCR products using 6 71/1316 first
round products on ER cD N A and asym ptom atic carcinoma reverse transcription product
111
F igu re 2.9 Polyacrylamide gel electrophoresis o f ER S et 2 and A ctin RT-PCR products
113
Figure 2.10 Polyacrylam ide gel electrophoresis o f E R S e t2 RT-PCR products including a
reverse transcription negative control
113
F igu re 2.11 Polyacrylam ide gel electrophoresis o f ER S et 2 and Actin RT-PCR products with
varying concentrations o f T 47-D total R N A and after additional precautions to m inim ize
aerosol contamination
113
F igu re 2.12 Polyacrylam ide gel electrophoresis o f ER S et 2 RT-PCR products using D N A
isolated from a small and large sam ple o f asym ptom atic breast carcinomas
113
F igu re 2.13(a-c) A garose electrophoresis o f ER S et 2, ER S et 4, and Actin RT-PCR products
o f Dynabead oligo d(T)25 isolated R N A from frozen, asymptomatic breast carcinoma
114
F igu re 2.14 Single stranded conform ational polym orphism analysis o f ER S et 3 35S labeled
PCR products using, high m olecular w eight D N A extracted from breast carcinomas, and
peripheral blood lym phocytes from 10 patients as template
116
F igure 2.15 Single stranded conform ational polym orphism analysis o f ER S et 2 35S labeled
RT-PCR products using lOOng o f total R N A extracted from asymptomatic breast carcinomas
as template
117
F igure 2.16(a-c) Single stranded conform ational polymorphism analysis o f ER Set 2, ER S et 4,
and ER Set 5 35S labeled RT-PCR products using Dynabead oligo d(T)25 m RNA extracted
from asymptomatic breast carcinom as as tem plate
117
F igu re 2.17(a-b ) Non isotopic sin gle stranded conformational polymorphism analysis o f p53
controls and ER S et 4 PCR products using p53 and ER cD N A control plasmids and
Dynabead oligo d(T)25m RN A isolated reverse transcription products as template
119
F igure 2.18(a) A garose gel analysis o f ER S et 2 reamplification PCR products
119
v
F ig u re 2 .1 9 (a-b ) x 250 and x 400 Form alin-fixed mammary carcinom a tissue, m icrowave
oven pretreated, and im m unohistochem ically stained for ER by ABC-HRP diam inobenzidine
H 2 O 2 procedure
123
F ig u re 2.2 0 (a -b ) x 250 Form alin-fixed mammary carcinom a tissue, m icrow ave pretreated
and stained by immunohistochemistry for PgR and MIB-1 by ABC-HRP diam inobenzine
124
F ig u re 21(a -g ) Single stranded conform ational polym orphism analysis o f ER S et 2 and
nested ER S et 2 35S labeled, PCR products using D yneabead oligo dT(25) m RNA isolated
from asym ptom atic breast carcinomas as template
127
F ig u re 2.22 Single stranded conformational polym orphism analysis o f ER S et 4 35S labeled
RT-PCR products o f Dynabead oligo d(T )25 isolated m RN A from frozen, asymptomatic
breast carcinom as
128
F ig u re 2.23 Single stranded conformational polym orphism analysis o f ER Set 5 35S labeled
RT-PCR products o f Dynabead oligo d(T )25 isolated m RN A from frozen, asymptomatic
breast carcinom as
128
F ig u re 2.2 4 Single stranded polymorphism analysis o f p53 positive and negative mutation
controls
128
F ig u re 2 .2 5 (a ) Sequence analysis o f asym ptom atic 8 ER S et 2 SSCP bandshift identifying
129
an A -> G substitution
F ig u re 2.25(b ) Sequence analysis o f asym ptom atic 43 ER S et 4 SSCP bandshift identifying
129
an T -> C substitution
F igu re 2.25(c) Sequence analysis o f asym ptom atic 43 ER S et 2 SSCP bandshift
identifying
129
an A ~>T substitution
F igu re 2 .26(a) Agarose gel analysis o f nested RT-PCR analysis
by Dynabead oligo d(T)25 from asym ptom atic breast carcinom as
from
m RNA
isolated
131
F ig u re 2 .26(b ) Agarose gel analysis o f nested RT-PCR analysis
by Dynabead oligo d(T)25 from asym ptom atic breast carcinom as
from
m RNA
isolated
131
m RNA isolated
131
F ig u re 2 .27(a) Sequencing analysis o f aberrant asym ptom atic 51 ER S et 2 nested PCR
product cloned into pCR-Script Am p SK + cloning vector identifying an in frame exon 3
deletion
133
F ig u re 2 .27(b ) Sequencing analysis o f asym ptom atic 56 ER S et 4 nested PCR product cloned
into pCR-Script Amp SK+ cloning vector identifying a partial deletion o f exon 4, the insertion
o f an additional 27 nucleotides and com plete deletion o f exon 5
133
F ig u re 2 .27(c) Sequencing analysis o f asym ptom atic 31 ER S et 5 nested PCR product cloned
into pCR-Script Amp SK+ cloning vector identifying an in frame deletion o f exon 7
134
F ig u re 2.27(d ) Sequencing analysis o f asym ptom atic 26 ER S et 5 nested PCR product cloned
into pCR-Script Amp SK+ cloning vector identifying an com plete deletion o f exon 7
134
F ig u re 2.28(a) Representative NIRCA analysis o f p53 controls with digestion using R N A se
system s #1, #2, and #3 on T7/SP6 transcript duplexes
137
F igu re 2.28(b ) Representative NIRCA analysis o f asym ptom atic carcinomas with digestion
using R N A se system #2 on ER S et 6 generated w t T7 and test SP6 transcript duplexes
137
F ig u re 2 .26(c) Agarose gel analysis o f nested RT-PCR analysis
by Dynabead oligo d(T)25 from asym ptom atic breast carcinom as
vi
from
F igu re 2 .28(c) Representative NIRCA analysis o f asymptomatic carcinom as with digestion
using R N A se system #1 on heterogeneous wt and mt ER Set 7 generated T7 and SP6 transcript
duplexes
137
F igu re 2 .2 9 (a ) Oestrogen receptor CpG island m ethylation
142
F igu re 2.2 9 (b ) A garose gel analysis o f oestrogen
products using ESR 1 + 2 PCR primers
PCR method
receptor CpG island m ethylation PCR
142
F igu re 2 .2 9 (c) A garose gel analysis o f oestrogen receptor CpG island m ethylation PCR
products using ESR 1/ER Meth 2 PCR primers
143
F igu re 2.2 9 (d ) Polyacrylam ide gel analysis o f ER CpG island methylation PCR products using
ESR 1+ 2 PCR primers
148
F igu re 2 .29(e) Polyacrylam ide gel analysis o f ER CpG island methylation PCR products using
ESR 1 + 2 PCR primers
143
F igu re 2.3 0 Restriction Map o f the human ER promoter
148
F igu re 2.31 Schem atic representation o f putative ER splice variants
162
F ig u re 3.1 A general m odel for mechanism eliciting loss o f heterozygosity
150
F ig u re 3.2 Schem atic representation o f the sources for m ultiple mutations in cancer as initially
proposed by Loeb, 1991, 1994
170
F ig u re 3 .3 Early stages o f mismatch repair D isplaced single bases (e.g. -A -) are recognized by
the m ism atch binding heterodimer M SH 2:M SH 6 which bind to the mispaired region
173
F igu re 3.4 (a -b ) Representative m icrodissection o f individual ducts from a case o f ductal
carcinom a in situ
187
F igu re 3 .5 (a -b ) Representative m icrodissection o f individual tumour foci from an invasive
breast carcinom a
188
F ig u re 3 .6 (a -e) Representative carcinom as w hich were heterozygous for all m icrosatellite
markers utilized
189
F igu re 3 .7 (a -e) Representative carcinomas which were h om ozygous for all m icrosatellite
markers utilized
190
F ig u re 3 .8 (a -b ) LOH analysis at the ESR locus illustrating the advantages o f the tumour
m icrodissection approach compared with w hole tumour sections
190
F ig u re 3 .9 (a -h ) Representative m icrodissection LOH in m am m ographically detected, breast
carcinom as
191
F ig u re 3 .1 0 (a -g ) Representative m icrodissection
192
LOH
in preinvasive
lesions
o f DCIS
F igu re 3 .1 1 (a -d ) Representative m icrosatellite instability in m am m ographically detected breast
carcinom as
199
F ig u re 3 .1 2 (a -d ) Representative m icrosatellite instability in m am m ographically detected breast
carcinom as
200
F low D iagram 1 General overview o f the planned approach for ER mutational analysis
80
F low D iagram 2 O verview o f approach undertaken for mutational analysis o f the oestrogen
receptor gene
121
vii
L IS T O F T A B L E S
T ab le 1.1 Relative risk for breast cancer with and without atypia
7
T ab le 1.2 Stage and survival at 5 years
8
T ab le 1.3 The main staging system s used to assess the extent o f spread o f breast carcinomas
9
T ab le 1.4 Five year breast cancer survival rates (%) by tumour size and lymph node status
10
T ab le 1.5 Relative frequency
o f the different types o f invasive breast tumours
12
T ab le 1.6 Ten year survival
rate (%) for the different carcinoma types
14
T ab le 1.7 A ssessm ent scores
for tubule form ation, and nuclear pleomorphism
14
T a b le 1.8 The prognostic significance o f m easuring oestrogen and progesterone receptors
together
16
T ab le 1.9 C hrom osom es altered in primary breast tum ours/cell lines identified by karyotypic
analysis
22
T ab le 1.10 Exam ples o f com m on chrom osom al abnormalities in breast tumour DNA detected
by Southern analysis
23
T a b le 2.1 The sequence, position and proposed function o f m otifs within the 5 ’ flanking
region o f the human oestrogen receptor gene
43
T a b le 2.2 A gents and extracellular signals w hich activate the oestrogen receptor in the absence
o f ligand
64
T a b le 2.3 Oestrogen receptor variants in human breast cancer tissue and cell lines
66 -6 7
T ab le 2.4 Oestrogen receptor splice variant expression in normal breast tissue
71
T ab le 2.5 Estimation o f oestrogen receptor sp lice variant m RNA present in normal and
tumour cells
73
T ab le 2.6 A verage (%) com position o f total oestrogen receptor m RN A from a panel o f breast
cancer cell lines
73
T a b le 2.7 C linicopathological features o f 44 frozen “early”, sporadic, invasive carcinom as
83
T ab le 2.8 Oestrogen receptor PCR primer sequences
93
T ab le 2.9 Effect o f tumour sam ple size and quantity ( m l) o f T R izol reagent on quantity and
quality o f RNA isolates from fresh frozen breast tumours
108
T ab le 2.10 Summary O f A nnealing Temperatures And S izes O f PCR Products
110
T ab le 2.11 Comparison o f ER H-score and proliferation index in relation to type and grade
o f m am m ographically detected Carcinoma
125
T ab le 2.12 Summary sequence data for oestrogen receptor SSCP bandshifts and PCR variants
135
T ab le 2.13 Correlation o f altered oestrogen receptor t transcripts in relation to pathological
data o f m am m ographically detected carcinom as
139
T ab le 2.14 Correlation o f the presence o f altered oestrogen receptor species with grade and
steroid receptor expression
140
T a b le 3.1 Summary o f loss o f heterozygosity data in breast cancer
164
T ab le. 3.2. Summary o f m icrosatellite instability in invasive breast carcinomas from other
groups
175
T a b le 3.3 Clinicopathological features o f 59 paraffin em bedded "early" sporadic, invasive
carcinom as
181
T a b le 3.4 Microsatellite repeat sequence PCR primer sequences
184
T a b le 3.5 Heterozygosity frequencies for the invasive and non invasive tumour groups at 5
m icrosatellite loci
186
T a b le 3.6 Pattern o f loss o f heterozygosity and MSI observed using 4 m icrosatellite markers
from 59 early invasive breast carcinomas
195
T a b le 3.7 Pattern o f loss o f heterozygosity and MSI observed using 5 microsatellite markers
from 16 non-invasive lesions o f DCIS
196
T a b le 3.8 Summary o f chrom osom e 6q LOH data in early invasive tumours and non-invasive
197
cases of DCIS
ix
Abbreviations
aa
ACF
AH
ALH
Amp
APC
APP
AR
AT
ATPgS
bp
BSA
BW
°C
cAMP
cDNA
cER
CEF
Ci
cm
DAB
dATP
DBD
DCC
DCIS
dCTP
ddNTP
DEPC
dGTP
DMEM
dNTP
DM
DNA
dTTP
EAT
E. coli
EDTA
EGF
ER
ER a
ERp
FAP
FCS
GR
GRE
h
HBD
H&E
hER
Amino Acid
Aberrant Cryptic Foci
Atypical Hyperplasia
Atypical Lobular Hyperplasia
Ampicillin
Adenomatous Polyposis Coli
Alternating Purine/Pyrimidine
Androgen Receptor
Ataxia Telangiectasia
Adenosine-5 ’-0-(3-Thiotriphosphate)
Base Pair
Bovine Serum Albumin
Binding And Washing
Degrees Centigrade
Cyclic Adenosine Monophosphate
Complimentary Deoxyribonucleic Acid
Chicken Oestrogen Receptor
Chicken Embryo Fibroblast
Currie
Centimeter
Diaminobenzidine
2 ’-Deoxyadenosine 5 ’-Triphosphate
DNA Binding Domain
Deleted In Colorectal Cancer
Ductal Carcinoma In Situ
2’-Deoxycytidine 5’-Triphosphate
Di-deoxynucleotide 5 ’-Triphosphate
Diethyl pyrocarbonate
2’-Deoxyguanosine 5’-Triphosphate
Dulbecco’s Modification Of Eagles Media
Deoxynucleotide 5 ’-Triphosphate
Distant Metastasis
Deoxyribonucleic Acid
2’-Deoxythymidine 5’-Triphosphate
Estimated Annealing Temperature
Escherichia coli
Disodium Ethylene Diamine Tetraacetate
Epidermal Growth Factor
Oestrogen Receptor
Oestrogen Receptor Alpha
Oestrogen Receptor Beta
Familial Adenomatous Polyposis
Foetal Calf Serum
Glucocorticoid Receptor
Glucocorticoid Response Element
Hour
Hormone Binding Domain
Haematoxylin And Eosin
Human Oestrogen Receptor
hGR
HNPCC
HRE
HRP
HRT
IDC
IGFI
IGFII
IGFIIR
IMS
INR
Inter
ILC
IPTG
kb
kD
1
LB
LCIS
lob
LOH
LNM
M
mA
Mamm
MAP
MDE
mg
Pg
MI
min
ml
Ml
mm
mM
pM
pm
MMR
MPC
M6P/IGF2R
mRNA
mt
NA
ND
NE
ng
NI
NIBH
NIRCA
nM
nM
Human Glucocorticoid Receptor
Hereditary Non-Polyposis Colorectal Cancer
Hormone Response Element
Horseradish Peroxidase
Hormone Replacement Therapy
Infiltrating Ductal Carcinoma
Insulin-Like Growth Factor I
Insulin-Like Growth Factor II
Insulin-Like Growth Factor II Receptor
Industrial Methylated Spirit
Initiator Element
Intermediate
Infiltrating Lobular Carcinoma
Isopropyl-1-Thio-p-D-Galactopyranoside
Kilobases
Kilodalton
Litre
Luria Broth
Lobular Carcinoma In Situ
Lobular
Loss O f Heterozygosity
Lymph Node Metastasis
Molar
Milliamps
Mammographically
Mitogen Activated Kinase
Mutation Detection Enhancement
Milligram
Microgram
Microsatellite Instability
Minute
Millilitre
Microlitre
Millimeter
Millimolar
Micromolar
Micrometer
Mismatch Repair
Magnetic Particle Concentrator
Mannose 6-Phosphate/Insulin-Like Growth F
Messenger RNA
Mutant
No Amplification
Not Described/Not Determined
Normal Epithelium
Nanogram
Not Informative
N-iodacetyl-N’-biotinylhexylenediamine
Non Isotopic RNAse Cleavage Assay
Nanomolar
Nanometre
xi
NST
NT
NTA
OCP
OD
ORF
PAGE
PCR
PE
PEG
PgR
pmol
rATP
RAR
RER
RFLP
RNA
rNTP
rpm
RT-PCR
S. cerevisiae
SDS
Sec
Soln
SSC
SSCP
StrepABC
Taq
TAE
TBE
TBP
TBS
TDLU
TE
TEMED
temp
TGF-a
TGF-P
TGF-piIR
TIFS
TNM
TR
tub
UK
UTR
UV
V
Vols
VNTR
W
WHO
Non Specific Type
Not Tested
No Tissue Available
Oral Contraceptive Pill
Optical Density
Open Reading Frame
Polyacrylamide Gel Electrophoresis
Polymerase Chain Reaction
Predisposed Epithelium
Polyethylene Glycol
Progesterone Receptor
Picomoles
riboadenosine 5’-Triphsophate
Retanoic Acid Receptor
Replication Error
Restriction Fragment Length Polymorphism
Ribonucleic Acid
Ribonucleotide Triphosphate
Revolutions Per Minute
Reverse Transcriptase-Polymersase Chain Reaction
Saccharomyces cerevisiae
Sodium Dodecyl Sulfate
Seconds
Solution
Saline Sodium Citrate
Single Stranded Conformational Polymorphism
Streptavidin Biotin Complex
Thermus Aquaticus
Tris Acetic Disodium Ethylene Diamine Tetraacetate
Tris Boric Disodium Ethylene Diamine Tetraacetate
TATA Box Binding Protein
Tris Buffered Saline
Terminal Duct Lobular Unit
Tris Disodium Ethylene Diamine Tetraacetate
N, N, N ’, N ’-teramethylethylenadiamine
Temperature
Transforming Growth Factor Alpha
Transforming Growth Factor Beta
Transforming Growth Factor Beta type II Receptor
Transcription Intermediary Factors
Tumour Node Metastasis
Thyroid Hormone Receptor
Tubular
United Kingdom
Untranslated Region
Ultraviolet
Volts
Volumes
Variable Number O f Tandem Repeats
Watts
World Health Organization
xii
wt
X-gal
Wild Type
(5-bromo-4-chloro-3-indoyl-p-D-galactopyranoside)
xiii
CONTENTS
TITLE
i)
DEDICATION
ii)
ABSTRACT
iii)
ACKNOWLEDGMENTS
iv)
LIST OF FIGURES
v-vii)
LIST OF TABLES
viii-ix)
ABBREVIATIONS
x-xiii)
xiv-xxi)
CONTENTS
1. CHAPTER I - INTRODUCTION
1-36
1. 1 The Normal Structure And Function O f The Breast
1-2
1.2 Breast Cancer
2-7
1.2.1 Family History Of Breast Cancer
4
1.2.2 Age
4
1.2.3 Age At Menarche
4
1.2.4 Age At Birth Of First Child And Parity
4-5
1.2.5 Age At Menopause
5
1.2.6 Obesity
5
1.2.7 Oral Contraceptive Pill (OCP)
5
1.2.8 Hormone Replacement Therapy (HRT)
5
1.2.9 Radiation
6
1.2.10 Geographical Variation
6
1.2.11 Benign Breast Disease
6-7
1.3 Predicting The Behavior O f Breast Cancer
7-21
1.3.1 Staging
7-10
1.3.2 Pathological Features
11-14
1.3.2.1 Histological Classification
11
1.3.2.1.1 Non Invasive Carcinoma
11
1.3.2 .1.1.1 Ductal Carcinomas In Situ
11
1.3.2.1.1.2 Lobular Carcinoma In Situ
11
1.3.2.1.2 Invasive Carcinoma
11
1.3.2.1.2.1 Infiltrating Ductal Carcinoma
11
1.3.2.1.2.2 Infiltrating Lobular Carcinoma
12
1.3.2.1.2.3 Special Types
12
xiv
1.3.2.1.2.3.1 Medullary Carcinoma
12
1.3.2.1.2.3.2 Colloid or Mucinous Carcinoma
12
1.3.2.1.2.3.3 Tubular Carcinoma
13
1.3.2.1.2.3.4 Invasive Cribriform
13
1.3.2.2 Histological Grade
13-14
1.3.3 Biological Features
15-21
1.3.3.1 Cell Kinetics
15
1.3.3.2 Steroid Receptors
15-16
1.3.3.3 Oncogenes/Tumour Suppressor Genes
16-19
1.3.3.3.1 c-myc
16
1.3.3.3.2 ras
16
1.3.3.3.3 c-erb-B2 (neu)
17
1.3.3.3.4. CCDN1
17-18
1.3.3.3.5 Rbl
18
1.3.3.3.6 p53
18-19
1.3.3.4 Others
19-21
1.3.3.4.1 pS2
19
1.3.3.4.2 Cathepsin D
20
1.3.3.4.3 Cell Adhesion Molecules
20
1.3.3.4.3.1 E-Cadherin
20
1.3.3.4.3.2 Inteerins
20-21
1.4. Genetic Alterations In Breast Cancer
21 -24
1.5 Studies O f “Early ” Breast Cancer
24-25
1.6 The Multistep Nature O f Carcinogenesis
25-28
1.7 The Natural History O f Breast Cancer
28-35
1.7.1 DCIS
29-33
1.7.2 LCIS
33-34
1.7.3 ADH
34
1.7.4 Morphologically Normal Breast Tissue
34
1.7.5 Susceptibility Genes
34-35
1.8 Hypothesis
35
1.9 Aims
36
XV
CHAPTER II - THE OESTROGEN RECEPTOR
37-159
2.1 An historical Perspective
37-40
2.2 Structural Organization O f The Oestrogen Receptor Gene
40-46
2.2.1 5’ Flanking Region O f The Oestrogen Receptor Gene
40-44
2.2.1.1 Promoter elements A nd Upstream Binding Sites
41 -44
2.2.1.2 Purine/Pyrimidine Tracts
44
2.2.2 Exon/Intron Structure O f The Oestrogen Receptor Gene
44
2.2.3 3’ Flanking Region O f The Oestrogen Receptor Gene
54-46
2.3 Conserved Sequence A nd Function O f The Oestrogen Receptor
46-51
2.3.1 Region C - The DNA Binding Domain
46-48
2.3.1.1 Nuclear Localization
48
2.3.3 Region E - The Ligand Binding Domain
48
2.3.3.1 Ligand Binding
48
2.3.3.2 Dimerization
48-49
2.3.3.3 Hsp 90 Binding
49-50
2.3.3.4 Transactivation Domain - TAFII
50
2.3.4 Region A/B - Transactivation Domain
50
2.3.5 Regions D and F
50-51
2.4 The Subunit Structure O f The Unliganded ER
51 -52
2.5 ER Mediated Transcription Activation
52-60
2.5.1 By Ligand Binding
52-60
2.5.1.1 Ligand Induced Activation O f The ER
52-53
2.5.1.2 Specific Binding and Stable Complex Formation At the HRE
53-56
2.5.1.3 Recruitment O f Transcription Factors/RNA Polymerases
56-58
2.5.1.4 Modulation O f The Oestrogenic Response By Membrane Receptor
5 8-60
Ligands
2.6 Phosphorylation O f The ER
60-64
2.6.1 Serine Phosphorylation O f The ER
60-61
2.6.2 Tyrosine -537 Phosphorylation O f The ER
62-63
2.6.3 Phosphorylation In The Absence O f Ligand
63-64
2.7 Alterations To The ER In Breast Cancer
65-69
2.7.1 Dominant Positive ERs
65-68
2.7.2 Dominant Negative ERs
68
2.7.3 Negative ERs
68-69
xvi
2.7.4 Miscellaneous ERs
69
2.8 Variant ER Expression In Normal Tissue
70-71
2.9 Level O f ER Variant Expression
72-75
2.10 Differential Promoter Usage
75-76
2.11 Methylation O f ER CpG Islands
76-79
AIM OF CH A PTER
79
2.12 Materials And Methods
81 -106
2.12.1 M aterials
81
2.12.1.1 Chemicals
81
2.12.1.2 DNAs
81
2.12.1.3 Radioisotopes
81
2.12.1.4 Enzymes And Miscellaneous
81 -82
2.12.1.5 Antibodies And Sera
82
2.12.1.6 Tissue Culture
82
2.12.1.7 Electrophoresis
82-83
2.12.1.8 Photography And Autoradiography
82-83
2.12.1.9 Tissues
83
2.12.1.10 Commonly Used Solutions
84-86
2.12.2 M ethods
87-106
2.12.2.1 Silane Coating O f Microscope Slides
87
2.12.2.2 Haematoxylin And Eosin Staining O f Sections
87
2.12.2.3 Immunocytochemistry
87-88
2.12.2.3.1 Antigen Unmasking O f Formalin-Fixed Paraffin Embedded
87
Breast Tissue Sections
2.12.2.3.2 Immunocytochemical Staining For ER, PgR andMIB-1
87-88
2.12.2.3.3 Assessment of ER and PgR Immunocytochemistry
88
2.12.2.3.4 Assessment of MIB-1 Immunocytochemistry
88
2.12.2.4 Cell Culture
89
2.12.2.4.1 Maintenance O f Breast Cancer Cell Lines
89
2.12.2.4.2 Passage
89
2.12.2.4.3 Freezing Down
89
2.12.2.5 RNA Isolation
89-91
2.12.2.5.1 Cell Lines
89-90
2.12.2.5.2 Frozen Tissue
90
xvii
2.12.2.5.3 Oligo d(T)25 mRNA Isolation From Total RNA
2.12.2.6 Reverse Transcription
90-91
91 -92
2.12.2.6.1 AMV Reverse Transcription
91-92
2.12.2.6.1.1 Cell Line RNA
91
2.12.2.6.1.2 Frozen Tissue RNA
91-92
2.12.2.6.2 Tth DNA Polymerase Reverse Transcription
2.12.2.7 Polymerase Chain Reaction
92
92-97
2.12.2.7.1 Design O f ER PCR Primers
92-93
2.12.2.7.2 Optimization O f ER PCR
94
2.12.2.7.3 Optimization O f Nested PCR
94
2.12.2.7.4 Agarose Gels Analysis Of PCR Products
95
2.12.2.7.5 PCR Amplification O f Reverse Transcription Products
95
2.12.2.7.6 Labeled PCR
95
2.12.2.7.7 PCR Methylation Analysis Of ER CpG Islands
95
2.12.2.7.8 PAGE Analysis O f Labeled PCR Products
96
2.12.2.7.9 Southern Analysis O f RT-PCR Products
96-97
2.12.2.8 Single Stranded Conformational Analysis
97-99
2.12.2.8.1 Isotopic SSCP
97-98
2.12.2.8.2 Non Isotopic SSCP
98-99
2.12.2.9 Mismatch Detect II Analysis
99-100
2.12.2.9.1 PCR And Nested PCR
99-100
2.12.2.9.2 Production O f Sense And Antisense RNA Probes
100
2.12.2.9.3 Hybridization O f Experimental And Control Transcripts
100
2.12.2.9.4 RNAse Treatment O f Experimental - Wild Type Hybrids
100
2.12.2.9.5 Agarose Gel Analysis Of RNAse Cleavage Products
100
2.12.2.10 Recovery O f Variant ER Species From Gels
100-101
2.12.2.11 Subcloning O f Recovered Variant ER Species
101 -103
2.12.2.11.1 Pfu PCR Product Generation
101
2.12.2.11.2 Cloning Reaction
101-102
2.12.2.11.3 Transformation
102
2.12.2.11.4 Culturing O f Variant ER Subcloning Reactions
102
2.12.2.11.5 Plasmid Isolation
102-103
2.12.2.12 Sequencing
103-105
2.12.2.12.1 O f PCR Products In Agarose Gels
xviii
103
2.12.2.12.2 Of Single Stranded DNA Isolated By Streptavadin Linked
103-104
Dynabeads
2.12.2.12.3 Plasmid sequencing
104
2.12.2.12.4 Sequencing Protocol
104-105
2.12.2.13 Biotinylation O f Oligonucleotide Primers
105
2.12.2.14 DNA Isolation From Frozen Tissue Using Trizol Reagent'
105-106
2.13 Results - Optimization O f Methodologies
2.13.1 RNA Isolation
107-120
107-108
2.13.1.1 Cell Line RNA
107
2.12.1.2 Frozen Breast Tumour RNA
107-108
2.13.2 PCR Optimization
108-110
2.13.3 Nested PCR Optimization
110-111
2.13.4 AMV Reverse Transcription PCR Optimization
111-115
2.13.4.1 Cell Line RNA
111-112
2.13.4.2 Frozen Breast Tissue RNA
112-115
2.13.5 Tth Reverse Transcription PCR Optimization
115
2.13.6 SSCP Optimization
115-118
2.13.6.1 Isotopic SSCP
115-116
2.13.6.1.1 Genomic DNA (Acugel Gels)
115
2.13.6.1.2 Frozen Tumour Total RNA (MDE Gels)
116
2.13.6.1.3 Frozen Tumour Oligo d(T)25 Dynabead mRNA (MDE)
116
2.13.6.2 Non Isotopic SSCP
116-118
2.13.6.2.1 Genomic DNA (Acugel And Metaphor Agarose Gels)
116-118
2.13.6.2.2 p53 wt/mt controls / Frozen Tissue Oligo d(T)25 Dynabead
118
mRNA (MDE)
2.13.7 Recovery Of Aberrant Bands From Gels
118
2.13.8 Sequencing O f ER Variants
120
2.13.8.1 Agarose Gel Slices
120
2.13.8.2 Biotin PCR Method
120
2.13.8.3 Plasmid Sequencing
120
2.14 Results
121-144
2.14.1 ER, PgR and MIB-1 Immunocytochemical Analyses
122-125
2.14.2 Oestrogen Receptor Variant / Mutant Analyses
126-137
2.14.2.1 Single Stranded Conformational Polymorphism Analysis
xix
126-165
2.14.2.2 Mismatch Detect II Analysis
136-137
2.14.3 Correlation O f Variant ER Species With Clinicopathologic Features 138-141
2.14.4 Oestrogen Receptor CpG Island Methylation Analysis
2.15 Discussion
141 -143
144-159
2.15.1 Technical Problems
144-147
2.15.2 Interpretation Of Data
147-159
3. CHAPTER III - LOSS OF HETEROZYGOSITY AT CHROMOSOME
161-211
6q25.1-27
3 . 1 Loss O f Heterozygosity
161-163
3 . 2 Loss O f Heterozygosity In Breast Cancer
163-165
3.3 Loss O f Heterozygosity A t Chromosome 6q In Breast Cancer
165-167
3.4 Candidate Tumor Suppressor Genes On Chromosome 6q
167-169
3.4.1 The Mannose 6-Phosphate / Insulin-Like Growth Factor II
167-168
Receptor
3.4.2 The TATA Box Binding Protein
168-169
3 . 5 Microsatellite Instability
170-178
3.5.1 Mismatch Repair genes
171-172
3.5.2 Mechanism Of Mismatch Repair
172
3.5.3 Microsatellite Instability In Sporadic Cancers
172-173
3.5.4 Microsatellite Instability In Breast Cancer
173-174
3.5.4.1 Invasive Carcinomas
3.5.4.2 Non Invasive Carcinomas
3.5.4.3 Pre Malignant Disease
3.5.5 Mismatch Repair Genes In Breast Cancer
174
3.5.6 Mismatch repair Mechanisms And Cancer
176-178
3.5.6.1 Transforming Growth Factor p Type IIReceptor
176
3.5.6.2 Mannose 6-Phosphate/Insulin-Like Growth factor II Receptor
176-177
3.5.63 Bax
111
3.5.6AE2F-4
177
3.5.6.5 Other Targets
178
AIM OF CHAPTER
178-179
3.6 Materials A nd Methods
180-185
3.6.1 Materials
180-181
XX
3.6.1.1 Patients
180
3.6.1.2 Tissues
180
3.6.1.3 Histology
180
3.6.2 Methods
182-185
3.6.2.1 Immunocytochemistry
182
3.6.2.2 DNA Extraction
182-183
3.6.2.2.1 Whole Tumour Sections
182
3.6.2.2.2 Tumour Microdissection And DNA Extraction
182-183
3.6.2.3 PCR Analysis At 6q25.1-27
183
3.6.2.4 Polyacrylamide Gel Electrophoresis Analysis O f Labeled PCR
183-185
Products
3.7 Results
186-200
3.7.1 Loss O f Heterozygosity Analysis At 6q25.1-27
186-197
3.7.2 Microsatellite Instability Analysis At 6q25.1-27
198-200
3 . 8 Discussion
201-215
3.8.1 Loss Of Heterozygosity At 6q25.1-27
201-207
3.8.1 Microsatellite Instability At 6q25.1-27
207-211
4. CHAPTER IV - CONCLUSIONS AND FUTURE WORK
212-215
4.1 Conclusion
212
4.2 Future Work
212-215
APPENDICES
Appendix I Asymptomatic Breast Tumour Clinicopathological Data (ER
Variant Analysis)
216
Appendix II Human Oestrogen Receptor 5’ Flanking Region And Gene
Coding Sequence
217-222
Appendix III Asymptomatic Breast Tumour Clinicopathological Data (LOH
and MI Analysis)
223-224
PUBLICATIONS AND ABSTRACTS
225
REFERENCES
226-277
xxi
1.
INTRODUCTION
1.1
The Normal Structure A nd Function O f The Breast
The female breast is a hormonally-regulated gland which is largely undeveloped until
puberty. Shortly before menarche, extensive branching and lengthening of the ductal system
occurs with the appearance o f terminal ducts and an increase in fat and connective tissue.
Growth continues until the mid-20s and is accelerated if pregnancy intervenes.
The main function o f the breast is the production and expression of milk. Secretory units of
the breast called lobules are composed of numerous acini. These consist o f epithelial cells
which secrete milk and myoepithelial cells which contract to expel it. The lobule, together
with the intralobular and terminal extralobular ducts, forms the terminal duct lobular unit
(TDLU). There is a network o f ducts leading to 15-20 lactiferous ducts which carry milk to
the nipple where it is expressed ( Fig 1.1).
During the menstrual cycle, the breast undergoes minor changes. Proliferation is tightly
linked to the menstrual cycle suggesting oestrogen and progesterone are likely factors in its
regulation (Ferguson and Anderson, 1981). In pregnancy, regulated proliferation and
enlargement are seen in, preparation for lactation which requires prolactin for initiation. On
cessation of breast-feeding, rapid involution occurs and the breast returns to its pre­
pregnancy structure.
Involution changes occur with increasing age, commencing pre-menopausally and
continuing past the menopause. Changes include connective tissue becoming denser,
thickening of the basement membranes around the acini and loss of cells lining the acini as
well as an increase in the proportion of adipose tissue.
Steroid receptors can be detected in normal breast, which are localized to nuclei o f epithelial
cells. Differences have been found throughout the menstrual cycle with a decline in the
oestrogen receptor (ER) in the luteal phase (Howell et al., 1994). The number o f cells with
detectable ER are low in premenopausal breast tissue with an increased detection o f ER after
the menopause (Walker et al., 1992).
l
Figure 1.1 Normal structure of the human mammary gland
L actiferous
sinus
Lactiferous
duct
Lobules
c o n ta in in g acini
(site of milk
fo r m a tio n )
Duct
(d ra in a g e system )
N ip ple
Skin
(site of miik
expression;
1.2
Breast Cancer
Breast cancer is one of the most com m on cancers affecting women in the industrialized
countries o f the West (Boring et a l, 1993). The incidence of breast cancer in the UK is 25,
000 newly diagnosed cases each year, with a mortality rate of 15, 000, and has with the
poorest 5 year survival rate in Europe and compared to the USA (Coleman et a l, 1993). The
age specific incidence begins to rise after 35 years of age and it is a leading cause of
morbidity and mortality in women over the age o f 50.
The incidence of breast cancer has been rising steadily since the early eighties (Miller,
1992). Improvement in cancer registration is unlikely to account for more than 25 % of the
increase that has occurred, and the introduction of screening mammography is thought to be
responsible for the surge in increase.
Although mortality from breast cancer has generally been increasing worldwide, there have
been recent reports of declining mortality in the UK, USA, Norway, Sweden and elsewhere
(Blot et al., 1987; Beral et al., 1995; Hermon and Beral, 1996).
Identification of environmental, biochemical and genetic factors that might contribute to the
aetiology and progression of breast cancer are therefore essential in terms of improving
prevention, diagnosis and therapy.
Epidemiological studies have evaluated the potential of several aetiological factors and other
variables associated with the development of breast cancer in women. However, none of
these factors alone or in combination can predict the occurrence or explain the variability of
the disease. These studies generally state the relative risk to indicate the strength of a risk
factor. However since these studies are purely observational, only an association, and not a
causation can be inferred. The following is a list of risk factors where a significant
association has been suggested:
Family history o f breast cancer
Age
Age at menarche
Age at birth of first child and parity
Age at menopause
Oral contraceptive pill
Hormone replacement therapy
Radiation
Geographical variation
Benign breast disease
3
1.2.1
Family History O f Breast Cancer
Genetic factors contribute to an ill defined proportion of breast cancer incidence, and
although not absolute, a genetic predisposition undoubtedly exists. Many statistically
significant data suggest this: the closer the familial risk the greater the risk (Sattin et al.,
1985); the greater the number o f first degree relatives affected, the greater the risk (Claus et
al.. 1990). The overall proportion of breast cancer patients with a genetic disposition has
been estimated to be 5% o f all cases, but approximately 25% of all cases diagnosed before
the age of 25 (Claus et al., 1991).
Inherited mutations in the p53 (Malikin et al, 1990), BRCA1 (Miki et al, 1994), and
BRCA2 (Wooster et al, 1995) genes are known to confer a predisposition to breast cancer.
It has also been suggested that heterozygous carriers of defective forms of the gene
predisposing to ataxia telangiectasia are at a higher risk for breast carcinoma (Swift et al,
1991).
1.2.2
Age
The incidence of breast cancer increases with age, doubling about every 10 years until after
the menopause, when the rate o f increase slows dramatically. Although breast cancer is still
a rare disease among young women, it has been suggested that the development of breast
cancer before the age o f 35 is an independent prognostic factor (Bonnier et al, 1995).
1.2.3
Age At Menarche
Women who begin to menstruate early in life have an increased risk o f developing breast
cancer. In addition, for a fixed age at menarche, women who establish regular menstrual
cycles within 1 year of the first menstrual period have more than double the risk o f breast
cancer than women with a 5 year longer delay in onset of regular cycles (Henderson et a l,
1988). A delay in establishment of regular ovulatory cycles is thought to be protective
(Apter etal, 1983).
1.2.4
Age At Birth O f First Child And Parity
Nulliparity and late age at first birth both increase the lifetime incidence of breast cancer.
Single and nulliparous women were found to have an increased risk of breast cancer
approximately 1.4 times the risk of parous married women (MacMahon et a l, 1970). The
4
risk of breast cancer in women who have their first child after the age of 30 is about twice
that of women who have their first child before the age of 20 (Anderson 1974).
1.2.5
Age At Menopause
Women who experience a natural menopause before the age of 45 have half the breast
cancer risk of those whose menopause occurs after the age of 55 (Trichopoulos et al., 1972).
Artificial menopause, by lateral oopherectomy or pelvic irradiation, also markedly reduces
breast cancer risk.
1.2.6
Obesity
The relationship between weight and breast cancer risk is critically dependent on age.
Subsequent to the menopause, the data suggests a modest increase in risk in women with
higher relative weight only amongst older post-menopausal women (Chi et al., 1974; Lubin
et al., 1985). Obese women have a higher breast cancer mortality, thought to be due to
delayed detection (Lew 1979).
1.2.7
Oral Contraceptive Pill (OCP)
The relationship between oral contraceptive use and the risk of breast cancer continues to be
a source of controversy (Pike et al., 1983). The relative risk of breast cancer in women who
have taken the OCP has been estimated to be 1.15 (WHO study of neoplasia and steroid
contraceptives 1990). The risk did not increase with duration of use and dropped when the
pill was stopped. However, studies that have targeted younger women have shown that the
use of the pill for more than a few years is associated with increased risk of breast cancer
irrespective of when they were used (UK national case control study group 1989; Miller et
al., 1989) although this is not definitive as other studies have shown (Paul et al., 1986).
1.2.8
Hormone Replacement Therapy (HRT)
The association between HRT and the development of breast cancer is controversial. HRT is
popular with post-menopausal women as it has been shown to reduce the risk of
cardiovascular disease, osteoporosis, stroke, Alzheimer’s disease and increase the users
overall quality of life (Lobo 1995). Women who have take HRT for 5 years or more have an
increased risk of developing breast cancer (La Vecchia et al., 1995; Colditz et al., 1995).
The risk among women of developing breast cancer using progestins in combination with
oestrogens are similar to the risk o f women using oestrogen alone (La Vecchia et al., 1995).
1.2.9
Radiation
The carcinogenic effect of high doses of radiation on breast tissue has been well studied in
atomic bomb survivors. From this data, it is known that risk is highest amongst women
exposed before the age of 20, with definite evidence of a dose-response effect. The
carcinogenic effect of radiation on the breast appears to diminish with increasing age and
exposure, and is very small following exposure after the age of 40 (Tokunaga et al., 1984).
1.2.10 Geographical Variation
Age adjusted incidence and mortality for breast cancer varies greatly between countries,
with Western countries generally having a higher incidence of breast cancer than Eastern.
Studies have documented the gradual acquisition of the breast cancer rates of their adoptive
countries by migrants from areas of lower incidence (Haenszel, 1961). For example, the
incidence o f breast cancer in Asia is low, yet studies o f Japanese who have emigrated to
America show that the second generation have an incidence of breast cancer approaching
that o f the native white population (Dunn, 1977). These studies suggest the importance of
environmental factors in the development of breast cancer.
1.2.11 Benign Breast Disease
Benign breast diseases encompasses a heterogeneous group of lesions that clinically and
radiographically span the entire spectrum of breast abnormalities. Dupont and Page (1985)
classified benign breast diseases into three main groups. Non proliferative lesions, e.g.,
cysts, papillary apocrine change, epithelial-related calcifications and fibroadenomas;
Proliferative lesions without atypia, e.g., intraductal hyperplasia’s, sclerosing adenosis and
moderate or florid hyperplasia o f usual type; Proliferative lesion with atypia, such as
atypical ductal or lobular hyperplasia. All these groups have different relative risks for
subsequent development of breast cancer. Studies by Dupont and Page (1985) and London
et al, (1992) have suggested a link between breast cancer and proliferative disease with
atypia (Table 1.1).
It is likely that more than one o f the aforementioned factors may play a role in the
pathogenesis o f breast cancer in any given patient. However, 75% of all breast cancer
patients do not have a recognizable risk factor (Strax, 1989).
6
Table 1.1 Relative risk for breast cancer with and without atypia
Dupont and Page (1985)
Relative Risk
95% confidence
Interval
Relative Risk
95% confidence
Interval
Without atypia
1.9
1 .2-2.9
1.6
1.0-2.5
With atypia
5.3
3.7
2.1
1.3
1
0°
CO
Proliferative
disease
CJ
London et al., (1992)
- 6 .8
Predicting The Behaviour O f Breast Cancer
The overall five year survival from breast cancer is approximately 62% (OPCS, 1988).
Prognosis in relation to breast cancer can be measured in terms of interval to recurrence after
treatment, and length o f survival. In general, these are closely related factors associated with
a high frequency of recurrence which correlate with reduced survival. The time to recurrence
may be delayed by aduvant therapy, as may overall survival but it may not reduce overall
mortality.
Various factors are used to try and predict the behaviour of a breast cancer. These relate to
clinical, pathological and biological features of the tumour.
1.3.1
Staging
Staging of a cancer is the grouping o f patients according to the extent of spread of the
disease. It is an established indicator for overall survival (Table 1.2). The two main systems
used are the International Classification o f Staging and the TNM system (Table 1.3).
The nodal status is best evaluated histologically as clinical evaluation has a high false
positive and negative rate (Schottenfeld, 1976). The histologic involvement o f axillary
nodes, and the number of nodes involved correlates with prognosis (Valagussal et al., 1978;
Ferguson et al., 1982).
Tumour size has also been cited as a prognostic indicator, although Fisher et al., (1969)
concluded that size alone was not as consequential to the patient’s survival as lymph node
7
involvement. A large study by the National Cancer Institute concluded that tumour size and
node status were independent but additive prognostic indicators (Table 1.4). As tumour size
increased, the prognosis was less favorable regardless of lymph node status, and as lymph
node involvement increased, survival status decreased regardless of tumour size (Carter et
al., 1989).
Table 1.2 Stage and survival at 5 years (Adapted from Dixon 1995)
Stage
Survival at 5 years
I
84%
II
71%
III
48%
IV
18%
8
Table 1.3 The main staging systems used to assess the extent of spread of breast carcinomas
International Classification
I
Lump with slight tethering to skin but node negative
II
Lump with lymph node metastasis or skin tethering
III
Tumour which is extensively adherent to skin and / or
underlying muscles, or ulcerating or lymph nodes are
fixed
IV
Distant metastasis
TNM
T,
Tumour 20mm or less; no fixation or nipple retraction,
including Paget’s disease
Tumour 20-50mm, or less than 20mm but with tethering
t3
Tumour greater than 50mm but less than 100mm; or less
than 50mm but with infiltration, ulceration or fixation
t4
Any tumour with ulceration or infiltration wide of it, or
chest wall fixation, or greater than 100 mm in diameter
N,
Node negative
N,
Axillary nodes mobile
n2
Axillary nodes fixed
n3
Supraventricular nodes or oedema of arm
M„
No distant metastasis
M,
Distant metastases
9
Table 1.4 Five year breast cancer survival rates (%) by tumour size and lymph node status
(Carter et al., 1989)
LN Status
Size (mm)
<5
5-10
10-19
20-50
>50
Negative
99.2
94.9
90.6
89.4
82.2
1-3 Positive
Nodes
95.3
94
86.6
79.9
73
4 + Positive
Nodes
59
54.2
67.2
58.7
45.5
96.2
94.9
90.6
79.8
62.7
Total
10
1.3.2
Pathological Features
1.3.2.1 Histological Classification
The following outlines the differing histological types adapted from the WHO classification
of breast cancer (1981).
1.3.2.1.1 Non Invasive Carcinoma
Lesion in which tumour cells are confined to the ductal or lobular units of the breast without
evidence of invasion through the basement membrane into the surrounding stroma. They are
characterized as either ductal or lobular, depending on the cytological features and growth
patterns, although there may be overlap between the two lesions.
1.3.2.1.1.1 Ductal Carcinoma In Situ (DCIS)
DCIS is divided into high, intermediate and low grades (Royal College of Pathologist
Working Group NHS screening Program, 1996). In the past most women with a diagnosis of
DCIS underwent mastectomy, so there is only limited data as to whether DCIS progresses to
invasive carcinoma.
1.3.2.1.1.2 Lobular Carcinoma In Situ fLCIS)
These lesions originate from the terminal ductules or acini. It is multicentric, does not form
a palpable tumour and has no characteristic radiographic signs. The true incidence of LCIS
is consequently unknown. LCIS is a marker of increased risk of developing cancer.
1.3.2.1.2 Invasive Carcinoma
The incidence of different histological types of invasive carcinoma are given in Table 1.5.
1.3.2.1.2.1 Infiltrating Ductal Carcinoma fIDC. No Special Type orNST)
Invasive ductal carcinoma is the single largest group of malignant mammary tumours
(Rosen et al., 1979), often characterized by hardness on palpation. To be defined as
infiltrating ductal, over 90% o f the tumour should not contain features of special tumour
types and it is thought o f as a classification of exclusion (WHO, 1981). There is
considerable histological variation within this group, with cells in trabeculae or solid groups,
with the cells themselves ranging from regular to pleomorphic. In addition, there is variable
or absent gland formation.
11
1.3.2.1.2.2 Infiltrating Lobular Carcinoma (ILC)
Typical clinical presentation is a palpable area of ill defined thickness in contrast to the
prominent lump of ductal carcinoma and characterized by single cell infiltration, often
around pre-existing breast structures.
Infiltrating lobular carcinoma is far more frequently associated with bilateral disease than
NST tumours and tends to be multicentric within the same breast.
Table 1.5 Relative frequency of the different types of invasive breast tumours (Page and
Anderson, 1987)
Type
Frequency (%)
Infiltrating ductal
73
Infiltrating lobular
10
Special Types
Medullary
5
Mucinous
2
Tubular
3
Cribriform and Papillary
5
Mixed Tumours
2
1.3.2.1.2.1 Special Types
1.3.2.1.2.1.1 Medullary Carcinoma
These tumours are circumscribed, often large (up to 5-10cm) and are composed of syncytial
cells with vesicular nuclei. Necrosis is common and the mitotic rate is often high. They have
extensive lymphocytic infiltrate and scanty fibrous tissue.
1.3.2.1.2.1.2 Colloid or Mucinous Carcinoma
The commonest tumour amongst older women and are large due to late presentation rather
than rapid growth rate. Histologically, the tumour is composed of nests, cords and even
isolated cells lying in lakes o f mucin which often accounts for more than half the volume of
the tumour. It may be pure colloid or associated with other histological types. The presence
of other histological types excludes it from the colloid category to that of mixed ductalspecial type.
12
1.3.2.1.2.1.3 Tubular Carcinoma (Tub)
These are highly differentiated infiltrating carcinomas composed of uniform cells arranged
in well developed tubules. Diagnosis is made if tubule formation is greater than 90%, the
exception being a combination of tubular and cribriform patterns where it is classified as
tubular if the latter component is greater than 50%. Generally, they are small lesions (1-2
cm). They are well differentiated and have an excellent prognosis. Ellis et al., (1992)
showed this type to be associated with a five year survival of approximately 90%.
1.3.2.1.2.1.4 Invasive Cribriform Carcinoma
Although similar to tubular carcinoma this type of tumour is classified as a special type due
to its excellent prognosis. It is characterized by the presence of islands of small cells that are
similar to cribriform DCIS. To be classified as cribriform carcinoma this pattern should be
present in >90% of the section.
Histological typing provides useful prognostic information in patients with primary operable
breast cancer (Ellis et al., 1992). Mucinous, tubular, and cribriform carcinomas have a better
prognosis than ductal carcinoma (Ellis et al., 1992). These data are summarized in Table
1. 6 .
1.3.2.2 Histological Grade
The term grade refers to the level o f differentiation within a tumour. The method of grading
generally employed is a modification o f the Bloom and Richardson method described in
1957 (Elston and Ellis 1991) and is based on evaluation of tubule formation, nuclear
pleomorphism and mitotic count. Each of these variables is given a score of 1-3 for each
feature that is assessed (Table 1.7).
Scores are added together to obtain the overall tumour grade:
3-5 points - Grade 1 i.e.: Well differentiated
6-7 points - Grade 2 i.e.: Moderately differentiated
8-9 points - Grade 3 i.e.: Poorly differentiated
The correlation between high histological grade and poor prognosis has been demonstrated
in many studies (Fisher et al., 1975; Contessa et al, 1987; Elston and Ellis 1991).
Table 1.6 Ten year survival rate (%) for the different carcinoma types (adapted from Ellis et
al., (1992))
Type of Carcinoma
Ten Year Survival
Ductal Carcinoma In Situ
92
Lobular Carcinoma In Situ
NA
Infiltrating ductal
47
Infiltrating lobular
54
Special Types
Medullary
51
Mucinous
80
Tubular-Cribriform
90
Table 1.7 Assessment scores for tubule formation, and nuclear pleomorphism
Assessed Feature
Score
Tubule formation
>75%
1
10-75%
2
<10%
3
Nuclear pleomorphism
Small, regular, uniform cells
1
Moderate increase in size and variability
2
Marked variation
n
J
Mitotic counts
Based on the number of mitoses per high power fields,
which relates to the microscope
14
1.3.3
Biological Features
1.3.3.1 Cell Kinetics
The relevance of cell kinetics as a tool to investigate growth pattern, biological
heterogeneity and clinical progression of human tumours has been supported by a number of
reports (Silvestrini et al., 1989; Merkel and McGuire 1990). Consequently the measurement
of the proliferative fraction of tumour cell populations (the ratio of cycling to non-cycling
cells) has become increasingly important as a complement to the clinicopathologic finding
in making clinical decisions.
The growth rate of a tumour can be defined as low, medium or high, with a tumour with a
lower rate o f cell growth generally associated with an improved prognosis. It is important to
keep in mind that a high rate of cell differentiation may also be accompanied by a high rate
of cell death, and that aberrant cell division occurs such that one division fails to result in
the doubling o f the number o f cells. Cell proliferation can be measured by the counting of
mitotic figures, [ H]-thymidine uptake, flow cytometry and by use o f monoclonal antibodies
mainly directed against the Ki-67 antigen (Walker and Camplejohn, 1988; Bouzubar et al.,
1989).
1.3.3.2 Steroid Receptors
Many risk factors point towards the role of steroid hormones and their receptors in the
progression o f breast cancer. The steroid implicated are oestrogen and progesterone, with
their hormonal action most likely via their cognate receptors, the oestrogen receptor (ER)
and the progesterone receptor (PgR).
It is thought that two thirds of breast tumours are ER positive (Lippman and Allegra, 1980)
with approximately half of these responding to anti-oestrogen therapy , while approximately
10% o f ER negative tumours do respond (Anderson and Poulsen, 1989). Patients with ER
negative tumours have shorter disease free intervals, earlier recurrence rates, and shorter
survival times when compared to patients with ER positive tumours (Maynard et al., 1978;
Hartveit et al., 1980). As PgR synthesis is regulated by ER (Lippman et a l, 1986), PgR is
regarded as an additional marker of response endocrine therapy and overall prognosis (Table
1.8). PgR negative tumours are more likely to fail endocrine therapy than PgR positive
tumours (Pertschuk et al., 1988), and in addition patients with PgR positive tumours have a
longer disease free interval and longer overall survival (Fisher et al., 1988; Chavellier et al.,
15
1988). In conclusion, evidence suggests that both the ER and PgR are both prognostic
indicators but neither is a strong predictor of behaviour.
Table 1.8 The prognostic significance of measuring oestrogen and progesterone receptors
together (reviewed by Leong and Lee, 1995)
Combination of
receptors
ER+ PgR+
Incidence
50%
5 year disease free
interval
73%
Overall 5 year
survival
91%
ER+ PgR-
20%
75%
93%
ER- PgR+
5%
68%
88%
ER- PgR-
25%
64%
11%
L3.3.3 Oncogenes/Tumour Suppressor Genes
Alterations to a number o f proto-oncogenes appears to be of importance in a proportion of
breast cancers and include c-myc, ras, c-erbB2, and CCND1.
Oncogenes
1.3.3.3.1 c-myc
The c-myc gene encodes a nuclear phosphoprotein which acts as a transcriptional regulator
controlling cell proliferation, differentiation and apoptosis (Evan et al., 1992). Alterations to
the c-myc gene, predominantly by amplification, have been found in approximately 25% of
carcinomas which has been considered to be associated with the development o f breast
cancer (Bonilla et al., 1988), with other studies correlating c-myc alterations with aggressive
features and/or poor prognosis (Escot et al., 1986; Varley et al., 1987; Bems et al., 1992).
1.3.3.3.2 ras
The ras gene encodes a G protein involved in signal transduction, but the role of activated
ras genes in human cancer is not very clear. Mutations are rarely identified (Rochlitz et al.,
1989), although loss of one W-ras-\ allele has been correlated with aggressive features
(Theillet et al., 1986). Linkage o f H-ras-1 to a minisatellite locus consisting o f four
common alleles and several rare alleles has been found to exhibit a significant association
with cancer and as many as 1 in 11 breast cancers might be attributed to this (Krontiris et
al., 1993).
16
1.3.3.3.3 c-erb-B2 (neu)
The c-erb-B2 gene encodes a 185 kD transmembrane glycoprotein, that has extensive
homology to the epidermal growth factor receptor and is a putative growth factor receptor
(Coussens et a l, 1985). Amplification o f the c-erb-B2 gene has been found in 20-30% of
invasive carcinomas (Varley et al, 1987; Slamon et al, 1987; Zhou et al., 1987; Borresen et
al, 1990) with a correlation found between amplification and aggressive features and poor
short term prognosis, although not by all researchers (Clark and McGuire, 1991; Zhou et al,
1989). Such c-erb-B2 gene amplifications have been shown to correlate with a
corresponding over-expression of mRNA and protein levels (Walker et al, 1989b; Borresen
et al, 1990; Venter et al., 1987).
Many researchers have found c-erb-B2 overexpression to be an independent predictor of
poorer disease free survival (Walker et al., 1989c; Gullick et al., 1991; Lovekin et al, 1991;
Winstanley et al., 1991; Press et al, 1993) and associated with more aggressive forms of the
disease. Surprisingly, c-erb-B2 protein has been detected in ductal carcinoma in situ (van
der Vijver et al., 1988), with more extensive studies demonstrating expression in 40-60% of
cases but always associated with high grade (comedo) type (Bartkova, et a l, 1990;
Ramachandra et al, 1990; Lodata et a l, 1990).
Tumours over expressing c-erb-B2 oncoprotein are more likely to-be ER and PgR negative
(Heintz et al, 1990; Marx et al, 1990) and can therefore be of value in determining therapy
since z-erb-B2 positive tumours show a poor response to endocrine therapy (Wright et a l,
1992; Nicholson et al, 1993; Klijn et a l, 1993).
1.3.3.3.4 CCND1
CCND1 encodes cyclin D1 (Xiong et a l, 1992) which when complexed with its associated
cyclin-dependent kinase, controls cell cycle progression in G1 by phosphorylating
retinoblastoma protein (Scherr, 1994). Cyclin D1 over-expression has been demonstrated in
breast cancer cell lines in both the presence and absence of amplification and may be a
potential factor in the pathogenesis of breast cancer (Buckley et al, 1993). Cyclin D1 over­
expression has also been demonstrated in breast cancers, with and without amplification
(Bartkova et a l, 1994; Gillet et al, 1994). Co-expression of cyclin D1 with epidermal
growth factor receptor in carcinomas have been shown to have a poorer prognosis
(McIntosh et al, 1995). Cyclin D1 expression has also been found to correlate with
17
oestrogen receptor levels (Gillet et a l, 1996) and consequently a good marker of likely
response to endocrine therapy.
Tumour Suppressor Genes
1.3.3.3.5 Rbl
The retinoblastoma gene has been mapped to 13ql4 (Friend et al., 1986) and encodes for a
105 kD protein which in its unphosphorylated form restricts cell cycle progression in G l, by
interacting with E2F transcription factor (Chellapan et al., 1991). Alterations to 13q
(Lundberg et al., 1987) and to the RB gene itself (T’Ang et al., 1988) have been found in
breast carcinomas. Comparison of allele loss and protein levels have shown both loss of
protein with loss of allele (Varley et a l, 1989) and high protein expression with allele loss
(Borg et al., 1992). Alterations were found either in advanced cases (Varley et al., 1989) or
in aneuploid high S phase cancer (Borg et al., 1992) suggesting that RB alterations were not
an initiating event in breast cancer but as a consequence of an unstable genome.
1.3.3.3.6 p53
Germline p53 mutations have been found in families with the Li-Fraumeni Syndrome
(Malakin et al, 1990), a rare syndrome in which there is young onset sarcoma associated
with a number of cancers including the breast in a first degree relative under the age of 45
years. However only half the Li-Fraumeni families have p53 mutations (Santibanez-Koref et
al., 1991) and germline mutations are rarely found in cases of early-onset breast cancer and
those with a strong family history (Sidransky et al., 1992; Warren et al., 1992).
Most p53 mutational studies have focused on sequence changes in exons 5, 6, 7 and 8 with
highly conserved domains (Varley et al., 1991; Osbourne et al., 1991; Mazars et al., 1992;
Coles et al., 1992; Merlo et al., 1993; Andersen et al., 1993; Caleffi et al., 1994; Eyijord et
al., 1995) although complete sequencing has been undertaken (Bergh et al., 1995). These
studies have demonstrated an association between the presence of mutations and aggressive
features within breast carcinomas, e.g. lack of ER (Mazars et al., 1992; Andersen et al.,
1993; Caleffi et al., 1994), high S phase index (Merlo et al., 1993). A significant association
was found between p53 mutations and disease-free and overall survival (Andersen et al.,
1993), and differences between sites of mutation between node positive and node negative
cases (Caleffi et al., 1994).
18
Prominent protein p53 expression detected by immunohistochemistry has been shown to be
associated with lack o f ER expression, poor differentiation, high proliferation rates, and the
presence of epidermal growth factor receptor (Jacquemier et al, 1994; Thor et al., 1992;
Walker et al., 1991; Poller et a l, 1992). It has also been shown to be an independent marker
of prognosis (Thor et al, 1992; Barnes et al, 1993; Lipponen et al, 1993).
1.3.3.4 Others
1.3.3.4.1 pS2
pS2 is a trefoil peptide (Poulson and Wright, 1993) and its function remains unclear. The
gene was initially identified by differential screening of cDNA from hormone treated and
untreated MCF-7 breast cancer cell line (Masiakowski et al, 1982). Expression of pS2
mRNA and protein is related to oestrogen receptor within breast carcinomas (Henry et al,
1989; Koemer et al, 1992; Walker et a l, 1995) and is a useful marker of potential hormone
responsiveness (Henry et al, 1991).
pS2 is also responsive to epidermal growth factor, tissue plasminigen activator, c-H-ras and
c-jun (Nunez et al, 1989), consequently there may be factors other than oestrogen
regulating its expression in breast cancers, which may account for the complex staining
pattern seen with immunohistochemistry and the finding of pS2 in ER negative tumours
(Dookeran et al, 1993).
1.3.3.4.2 Cathepsin D
Cathepsin D, a major lysosomal protease, was initially identified in the medium of MCF-7
breast cancer cells cultured in the presence of oestrogen (Westley and Rochefort, 1980).
However, it is constitutively overexpressed in ER negative breast cancer cell lines and
several studies have failed to find a relationship between cathepsin D and ER in primary
breast carcinomas (Thorpe et a l, 1989; Tandon et al, 1990; Duffy et a l, 1991; Walker et
al, 1995). Transfection of cathepsin D resulted in transformed cells increasing their
metastatic capacity (Garcia et al, 1990). A number of clinical studies have shown that
cathepsin D is a marker o f poor prognosis (Thorpe et al, 1989; Tandon et al, 1990; Duffy
et al, 1991) but immunohistochemical studies have shown that the cathepsin D may be in
the stromal component, rather than the cells and that this may be a reflection of macrophage
infiltrate (Walker et al, 1994)
19
1.3.3.4.3 Cell adhesion Molecules
1.3.3.4.3.1 E-cadherin
E-cadherin is a calcium-dependent adhesion molecule which plays a critical role in initiating
and maintaining cell-cell contacts (Kembler, 1992). E-cadherin molecules are located within
adheran junctions and are transmembrane structures: the cytoplasmic region interacts with
catenins, which are in turn connected to the actin microfilament network (Kembler, 1993).
Transfection of E-cadherin cDNA has demonstrated that it is important in suppressing
invasion (Vleminckx et al., 1991).
Clinical studies showed that reduced E-cadherin has (Moll et al., 1993; Oka et al., 1993) and
has not (Lipponen et al., 1994; Jones et al., 1996) been related to poorer differentiation of
infiltrating ductal carcinomas. Some researchers have found a relationship between reduced
membrane staining and lymph node metastasis (Oka et al., 1993; Jones et al, 1996).
Infiltrating
lobular and lobular carcinoma in situ have been shown not to express E-
cadherin (Moll et al., 1993; Lipponen et al., 1994; Jones et al., 1996; Lindblom et al., 1993)
which may be relevant when explaining the infiltrative nature of this type of invasive
carcinoma.
The E-cadherin gene has been mapped to 16q, deletions of which are frequent and
associated with distant metastasis (Rasbridge et al., 1993). Mutations have been found in a
small number of infiltrating lobular carcinomas (Kanai et al., 1994; Berx et al., 1995a, b),
but not in infiltrating ductal carcinomas (Kashiwaba et al., 1995)
I.3.3.4.3.2. Integrins
Integrins are cell adhesion molecules which are involved in cell-stromal, and possibly cell­
cell interactions. They are heterodimers composed of non-covalently linked a and p
subunits providing a transmembrane link between the cytoskeleton and specific extracellular
matrix proteins (Hynes, 1992). The integrins are classified according to their p subunit.
Several studies have examined integrin expression in primary breast carcinomas both at the
protein and mRNA levels (Pignatelli et al., 1994; Natali et al., 1992; Jones et al., 1992;
Zutter et al., 1993). All have shown a reduced expression at the protein and mRNA levels
for a 2Pi, a 6pj and a 6p4 with no relationship to grade or node status (Jones et al., 1992),
20
although others found that loss of a 2p, to be greater in poorly differentiated carcinomas
(Zutter et al., 1990; Pignatelli et al., 1994).
Jones et al., (1992) observed an alteration in a 2Pi expression in non-involved tissue from
cancer containing breasts, in that two thirds of the cases showed loss of reactivity, identical
to that of the corresponding tumour suggesting that altered a 2pj expression may be an early
event in the neoplastic process.
1.4
Genetic Alterations In Breast Cancer
The main approach to identifying genes that are involved in breast cancer has been to assess
chromosomal alterations that occur within these breast tumours. Up until relatively recently
this has been accomplished in two different ways. The first was by karyotypic analysis
which identified chromosomal regions that were consistently altered and therefore likely to
harbour aberrant genes. Although multiple cytogenetic abnormalities were detected in breast
cancers, few consistent abnormalities were noted (Mars and Saunders, 1990). This was in
part due to the source of tumour cells, e.g., established cell lines, primary tumours;
differences in culturing systems which may confer a selective growth advantage or
disadvantage to a particular cytogenetic abnormality; study of tumours prior or after
administration of adjuvant therapies while others may have not; and due to tumour
heterogeneity, where there is no reason to believe that karyotypic changes common to one
type o f carcinoma are common to another.
Karyotypic analysis of primary tumours, primary cultures and cell lines demonstrated
multiple abnormalities, although it was possible to discern that some chromosomes were
more commonly affected. Chromosomes most commonly affected are summarized in Table
1.9. The types of alterations identified included deletion of parts or whole arms of
chromosomes, e.g., deletion of lp21-lpter, and translocation, e.g., translocation of 14ql 1.
Due to the difficulties encountered in obtaining consistent cytogenetic data from breast
tumours, e.g. obtaining mitotic cells, many investigators began to directly study changes in
tumour DNA itself by Southern analysis which was also more sensitive than karyotypic
analysis. Using Southern analysis, amplification and loss of chromosomal regions were
more readily detected at greater resolution than karyotypic methods (Table 1.10). In this
21
Table 1.9 Chromosomes altered in primary breast tumours and cell lines identified by karyotypic analysis (Adapted from Mars and Saunders, 1990)
Chromosome
Reference
1
Rodgers et al, 1984, 1985; Ferti-Passantonopoulou and Panani, 1987; Gebhart et al,
1986; Zhang et al, 1989; Cruciger et al, 1976; Bello and Rey, 1989
2q
Weinberg, 1989; Rodgers et al, 1985
3
Ferti-Passantonopoulou and Panani, 1987; Bello and Rey, 1989; Rodgers et al, 1984,
1985; Zhang et al, 1989
5
Ferti-Passantonopoulou and Panani, 1987
6
Rodgers et al, 1984; Ferti-Passantonopoulou and Panani, 1987; Bello and Rey, 1989;
Trent, 1985; Rodgers et al, 1985
8p
Rodgers et al, 1984, 1985; Hill et al, 1987
1lp
Ferti-Passantonopoulou and Panani, 1987; Bello and Rey,1989; Rodgers et al, 1985
12
Rodgers et al,
1984, 1985
13
Rodgers et al,
1984, 1985; Hill et al, 1987
14q
Gebhart et al, 1986
16
Rodgers et al, 1984
17p
Rodgers et al, 1984
Table 1.10 Examples o f common chromosomal abnormalities in breast tumour DNA
detected by Southern analysis (Adapted from Mars and Saunders, 1990)
Chromosomal
Region
DNA Alteration
Lidereau et al., 1988; Guerin et al., 1988;
Adnane et al., 1989; Garcia et al., 1989
8q24
11 ql 3
Amplification
Genuardi et al., 1989
lp36
Chen et al., 1989; Merlo et al., 1989
Iq21-q32
llp l-1 5
13ql2-34
17pl 3.3
Adnane et al., 1989; Zhou et al., 1989; Ali
et al., 1989a
Guerin et al., 1988; Adnane et al., 1989;
Garcia et al, 1989; Slamon et al, 1987; Ali
et al., 1988; Zeillinger et al., 1989; Slamon
etal., 1989
17ql2-21.3
3p21-25
Reference
LOH
Devilee et al, 1989; Ali et al., 1989b
Lidereau et al., 1988; Garcia et al., 1989;
Devilee et al,
1989; Ali et al., 1987;
M ackayetu/., 1988a
Devilee et al., 1989; Lundberg et al., 1987;
T’Ang etal., 1988; Varley etal., 1989
Devilee et al., 1989; Mackay et al, 1988a
23
way amplification of c-myc, c-erb-B2, bci-\ and hst genes was demonstrated. In addition, a
number of regions where common loss of DNA sequences, suggesting the potential
localization of tumour suppresser genes, were identified.
Despite the obvious advantages of these techniques, analysis was almost exclusively
confined to larger, possibly later stage tumours which could be obtained as fresh tissue
breast tumours. This was due predominantly to the requirement of large amounts of high
molecular weight DNA to undertake Southern analysis, whereas karyotypic analysis
required fresh tissue and consequently was not suitable for the analysis of paraffinembedded tumours. Molecular alterations seen in tumours at this later stage of the disease
may not reflect changes which play an important role in the development and earlier stages
of the disease. As a result the involvement of many genes known to be important in the
biology o f breast cancer, e.g., the oestrogen receptor, has been limited.
1.5
Studies O f “Early ” Breast Cancer
Screening for breast cancer by mammography is increasing in frequency throughout the
world due to finding from a number of studies, which have shown that a reduction in
mortality can be achieved (Verbeek et al, 1984; Shapiro et al., 1985; Tabar et al, 1985).
The key factors are considered to be the smaller size of screen-detected cancers and the
lower frequency of nodal metastasis.
The introduction of screening has also resulted in the identification o f small, invasive
cancers and pre-invasive cancers which represent earlier stages of the disease. Consequently
small, mammographically detected breast cancers form a useful group for molecular
analysis into the involvement of many genes and genetic alterations in development and
early stages of the disease.
Additionally, advances in laboratory technology, especially the advent o f the Polymerase
Chain Reaction (PCR) and techniques based on this process, also allows the analysis of
considerably smaller amounts of tissue. Development of techniques, such as Reverse
Transcription-PCR (RT-PCR) means that genes of importance in breast cancer, e.g., the
oestrogen receptor, which have been studied predominantly in well established cancers, can
now be studied in greater detail at an earlier stage of the disease.
24
Studies within this laboratory (Rajakariar and Walker, 1995) have been undertaken to
examine early invasive breast cancers using immunocytochemistry for various markers
associated with different behaviours, e.g., oestrogen receptor, the presence of which is
associated with better differentiation (Bruun Rasmussen et a l, 1981), c-erb-B2 protein and
p53, which are associated with more aggressive features (Walker et al., 1991) and compared
to that o f a group clinically presenting, invasive carcinomas, in aged-matched patients.
Rajakariar and Walker (1995) found that the screen-detected carcinomas differed from
clinically presenting carcinomas in many ways. They were more likely to be lymph node
negative, with a higher prevalence o f oestrogen receptor, pS2, and Cathepsin D, although
there was no significant differences in progesterone levels. p53 and c-erb-B2 were
detectable in fewer screen-detected carcinomas which were also found to have a lower mean
proliferation index than the clinically presenting carcinomas.
Rajakariar and Walker (1995) also identified a small subset of tumours with an unusual
pattern in that they had high oestrogen receptor levels with high proliferation indices, as
well as tumours with the expected high oestrogen receptor levels and low proliferation
indices. Consequently, this could be an area which would provide information about
divergence of pathways of progression which involve the oestrogen receptor and increase
our knowledge concerning the natural history of the disease.
7 .6
The Multistep Nature O f Carcinogenesis
Nowell (1976) was the first to propose that cancer can develop in a multistep manner and a
substantial amount of evidence has been collected over the past decade which indicates that
the genesis of malignancy requires the sequential accumulation of a number of genetic
alterations (Fearon and Vogelstein, 1990). Colorectal carcinoma is an excellent model of the
multistep nature of cancer in which distinct clinical stages of disease development are
identifiable.
Epidemiological studies have suggested that approximately 15% of colorectal cancers occur
in a dominantly inherited pattern (Houlston et al., 1992). Two familial forms are now well
defined, familial adenomatous polyposis (FAP) and hereditary non-polyposis colorectal
cancer (HNPCC). FAP patients develop between hundreds and thousands of benign
colorectal tumours (adenomas), some of which eventually progress to carcinomas. The
disease is characterized at the molecular level by gross chromosomal alterations leading to
25
losses of chromosome regions. This is identified at the molecular level as Loss of
Heterozygosity (LOH), which is recognized as the hallmark of tumour suppressor gene
inactivation (described in detail in Chapter 3). LOH is particularly prevalent at
chromosomes: 5q, 17p and 18q, suggesting the involvement of tumour suppressor genes at
these regions (Vogelstein et al., 1988).
In the late 1980s, the adenomatous polyposis coli (APC) gene was identified on
chromosome 5q and mutant APC alleles were shown to co-segregate with the affected FAP
kindreds (Bodmer et al., 1987; Leppert et al., 1987). However, patients with germline
mutations o f APC do not necessarily develop colon cancer, but they are at an elevated risk
of doing so over the general population. In order for such tumours to form, additional
genetic alterations are required (Figure 1.2).
O ther
alterations
MMR
deficiency
Normal
Epithelium
Dysplastic
ACF
Early
Adenoma
Intermediate
Adenoma
Late
Adenoma
►
C arcinom a
M etastasis
F igu re 1.2 G enetic alterations associated with colorectal carcinogenesis. APC alterations occur at the
dysplastic aberrant cryptic foci (ACF), which are believed to be precursors o f adenom as. These in turn
progress to carcinom a as they acquire subsequent mutations (K -RAS proto-oncogene activation, allelic loss o f
Chromosome 18, which is likely to represent aberration to the DPC4 and JV18-1 tumour suppressor genes,
and finally p53 alterations). Inactivation o f the mismatch repair (M M R) proteins are b elieved to accelerate the
tum ourigenic process, as indicated by the “rolling circle” (Adapted from Kinzler and V ogelstein, 1996).
The other genetic alterations identified in the sequence of events include: the activation of
the proto-oncogene K-ras, and the inactivation of the tumour suppressor gene, p53 (Fearon
and Vogelstein, 1990). The identity of the putative tumour suppressor gene at the 18q locus
remains controversial. The initial candidate was DCC {deleted in colon cancer) (Fearon et
al., 1990). However, recent evidence suggests that this gene may not have tumour
suppressive functions (Fazeli et al, 1997). Members of the Mad family, encoding proteins
that transduce signals from the transforming growth factor p (TGFp) family of cytokines,
are possible candidates (Thiagalingam et al., 1996) which are currently under investigation.
26
As the multistep progression for colorectal carcinogenesis has been refined over the years, it
has emerged that the order in which the molecular aberrations occur is not important, except
for in the initiation stage. Mutations in genes other than APC (e.g. p53 or K-ras) do not
efficiently initiate the neoplastic process. Therefore, it has been proposed that APC acts as
cellular ‘gatekeeper’, that directly regulates colonic epithelial growth by inhibiting growth
or promoting cellular death (apoptosis) (Morin et al., 1996). The mechanisms by which
APC normally control this switch are now emerging (Peifer et al., 1997). In normal colonic
epithelial cells, APC and a serine-threonine protein kinase (GSK) function together to
maintain low levels of a key signaling molecule (p-catenin). However, in cells with inactive
APC this control is abolished and the level of p-catenin increases. In turn, high levels of pcatenin lead to the formation o f complexes with a group of transcription factors (Korinek et
al., 1997; Morin et al., 1997), which may activate expression of genes involved in the
stimulation of proliferation and inhibition of apoptosis.
In contrast to FAP, tumours o f the HNPCC syndrome do not typically display loss of
chromosomal regions. Instead these tumours are characterized by an elevated mutation rate
which is manifest at the molecular level by replication errors or microsatellite instability
(MI). HNPCC has been shown to be caused by germline mutations in four mismatch repair
genes (Marra and Boland, 1995). Although the increase in genetic instability can be simply
monitored by genome-wide changes at arbitrary loci, it is emerging that this defect may also
target critical cancer-associated loci (Markowitz et al., 1995) (described in detail in Chapter
3). The mismatch repair genes responsible for HNPCC play a role in maintaining genome
integrity, and have been described as ‘caretaker’ genes (Kinzler and Vogelstein, 1997). In
HNPCC, it is believed that adenomas form at approximately the same rate as in the general
population. However, an adenoma cell from a HNPCC patient will acquire mutations at an
elevated rate, and the resultant accumulation of mutations in oncogenes and tumour
suppressor genes may lead to a rapid progression to malignancy.
It has been assumed that the genetic instability, associated with MI, occurred in the later
stages of the model, accelerating the rate of tumour progression (Kinzler and Vogelstein,
1996). However, a recent study has demonstrated that tumours with MI also have mutations
of APC (Huang et al., 1996). The specific type of APC mutations in the MI patients are
distinctly different from those o f the APC mutations in patients without MI, and are
characteristic of tumours with the MI phenotype. This suggests that the genetic instability
27
may precede APC mutations. In these patients with MI tumours, genetic instability may lead
to tumourigenesis via APC inactivation and therefore is indirectly responsible for tumour
initiation and progression.
A model of how ‘gatekeeper’ and ‘caretaker’ genes may co-operate in pathways of inherited
susceptibility to cancer, and in particular the development of colon cancer, has recently been
proposed (Kinzler and Vogelstein, 1997), and is illustrated in Figure 1.3. Inherited
mutations of either a caretaker or gatekeeper gene can predispose an individual to neoplasia,
however, additional genetic alterations are needed to convert a predisposed cell to a
neoplastic cell. In the caretaker pathway, a mutation of one caretaker allele must be followed
by three other mutations (the second caretaker allele and two gatekeeper alleles). In the
gatekeeper pathway, mutation of one gatekeeper allele is followed by mutation of the other
gatekeeper allele, in order to initiate neoplasia.
Gatekeeper pathway
N
O
R
M
A
L
Caretaker pathway
Mutation of
caretaker allele
Mutation o f second b
caretaker allele B
leads to genetic B
instability
B
Mutation of
gatekeeper allele
k
B
B
I
Mutation o f second
gatekeeper allele
leads to tumour
initiation
F igu re 1.3 Caretaker and Gatekeeper pathways o f inherited susceptibility to cancer. In the caretaker pathway,
an inherited mutation is follow ed by three additional mutations, although the genetic instability that follow s
inactivation o f the second caretaker allele accelerates the accumulation o f the later mutations. In the
gatekeeper pathway, one mutation coupled w ith the inherited mutation is needed to initiate neoplasia (Kinzler
and V ogelstein, 1997).
1.7
The Natural History O f Breast Cancer
A model of breast tumourigenesis has been postulated in which a normal epithelium cell of
the ducts or lobules gives rise to a focus of proliferating cells (hyperplastic without atypia
and subsequently, with atypia) and then, through an accumulation of molecular
abnormalities, evolves into a neoplasm, initially carcinoma in situ, followed by invasive
disease and finally to metastatic disease (Russo and Russo, 1991).
28
This progression has been depicted by Devilee et al., (1994), and is illustrated in Figure 1.4.
As can be seen the proposed model is very complex, with the potential for breast cancer to
develop through a variety of different pathways. For example, invasive carcinoma may
develop from normal epithelium without progressing through a non-invasive step. Similarly
it is possible that not all in situ disease progresses to invasive carcinoma. This model has
been proposed mainly from histological observations and epidemiological data which
suggest that certain histological subtypes confer an increased risk of developing breast
cancer (Section 1.3).
Over the last few years, a major area of research has been in attempting to understand how
the various precursor lesions behave at the molecular level. The studies described below
provide some molecular evidence to underpin each of the pathways proposed by Devilee et
al., (1994).
The model proposed by Devilee et al., (1994), only accounts for the development of
infiltrating ductal carcinoma. Another model has recently been proposed which
encompasses some of the other histological subtypes of breast cancer (Walker, 1997),
illustrated in Figure 1.5. The early steps in this model are also supported by the molecular
analysis of precursor lesions described below. However, this model also positions
previously defined genetic alterations (Section 1.3.3.3 and 1.4) in the proposed pathways of
breast cancer development, as well as the genes responsible for the familial component of
the disease.
1.7.1
Ductal Carcinoma In Situ
Histological observations indicate that DCIS is a precursor to infiltrating ductal carcinoma
o f the breast because of finding it in association with invasive carcinoma (Page and
Anderson, 1987). However, not all cases of DCIS appear to have the same potential of
progressing to invasive cancer. This depends on the nature of DCIS; low grade DCIS,
misdiagnosed as benign, and incompletely excised can recur up to 24 years later (Page et al.,
1995). High grade DCIS recurs, with invasion, earlier and at a higher frequency (Lagios et
al., 1989). A greater understanding o f the biology of DCIS is a major concern, because
DCIS has been diagnosed with increasing frequency in recent years, largely as a result of the
29
Figure 1.4. Hypothetical model for infiltrating ductal carcinoma (Adapted from D evilee et al., 1994). Normal (NE) or predisposed epithelium (PE) may develop into infiltrating ductal
carcinoma (IDC), without proceeding through atypical ductal hyperplasia (AD H ) or ductal carcinoma in situ (DCIS), as proposed by D eng et al., (1996). DCIS may progress to larger
DCIS, with the accumulation o f further genetic alterations, or develop into IDC, or remain as DCIS, as proposed by Fujii et al., 1996b. DM indicates distant metastasis, and LNM is
lymph node metastasis.
Lymph Node
Metastasis
Atypical
Hyperplasia
Infiltrating
Ductal
Carcinoma
/\
( 1)
Normal
Epithelium
Distant
Metastasis
Predisposed
Epithelium
Ductal C a rcin o m a \(2)
In Situ
| (2)
Infiltrating
Ductal
Carcinoma
(Sub-clinical Stages)
Ductal Carcinoma
In Situ
Ductal Carcinoma
In Situ
(Clinically
M anifest)
MONOCLONAL GROWTH
Figure 1.5. M odel o f breast cancer pathways. The small arrows indicate possible pathways where the evidence is limited The larger arrows indicate there is better evidence (Walker,
1997).
Normal Breast
C aretak er genes
(BR C A 1 & 2, A TM )
‘At risk lesions’
AH, LCIS
DCIS (Non-H igh)
DCIS (High grade)
Rb alterations
m utation
I
p53 m utations
c-erb-B2 am plification
]
Low incidence p53
♦
Other IDC
A ggressive IDC
Tubular carcinomas
ILC
Low incidence p53
m utation
introduction of screening mammography. In the absence of known specific genes, the
identification of allelic losses is often used to analyze DCIS. LOH at chromosomes 1 (Munn
et al, 1995), 1lq (Zhuang et al., 1995), 17q (Munn et al, 1996a), and 16q and 17p (Radford
et al., 1995; Stratton et al., 1995; Munn et al., 1996b; Chen et al., 1996) have been
demonstrated. To date, all of the chromosomal
regions identified in DCIS, have been
previously documented in invasive carcinomas (Section 3.3).
A number of these studies have analyzed pure DCIS, without associated invasion, and DCIS
with synchronous invasion. Microdissection of individual in situ and invasive foci have
identified loss of the same allele for markers on chromosome 1 l q l 3 in all of the cases
studied by Zhuang et al., (1995), suggesting that in these cases, the in situ tumour may have
represented a precursor lesion of invasive carcinoma. This has also been observed at the
majority of cases for other chromosomal regions: lq (Munn et al., 1995), 16q and 17p
(Stratton et al., 1995; Munn et al., 1996a), and 17q (Munn et al., 1996b).
In general, the investigations described above analyzed one or two in situ and invasive foci
from individual tumours. Fujii et al, (1996a), reasoned that this approach may not be
adequate to define the sequence of genetic events in the progression of an individual tumour,
since distinct foci at various stages of progression may be present within one tumour
sample. In their study, multiple individual foci were microdissected and analyzed at 7
chromosomal regions, in two groups of lesions: cases with DCIS and synchronous
infiltrating breast cancer, and cases of pure DCIS. In 8 of 20 (40%) cases with synchronous
DCIS and invasive cancer, loss of different alleles was observed in some of the in situ foci.
This phenomenon has been described as ‘allelic heterogeneity’ (Chen et al, 1992). In
contrast the frequency of allelic heterogeneity between ducts from the pure DCIS was only
8% (3 o f 23). Further evidence for this difference between the pure DCIS, and DCIS with
invasion, came from an extended study of pure DCIS, which also described low frequencies
of allelic heterogeneity (Fujii et al., 1996b).
Fujii et al., (1996a, 1996b) proposed a model whereby the involved ducts in pure DCIS have
evolved along a common pathway. As the tumour becomes invasive (DCIS with
synchronous invasive cancer) several in situ tumour foci diverge along different pathways of
genetic progression. In contrast allele heterogeneity was not observed between multiple foci
32
of invasive carcinomas. Therefore it is possible that growth of invasive cancer is the result
of a rapid expansion of one population from the divergent in situ cancer.
Clinically, DCIS is a heterogeneous group of pathological and biological subtypes, which
may differ in behaviour. The lesions can be categorized by their nuclear morphology into
high, intermediate and low nuclear grade (Bobrow et al, 1995). The nuclear grade of a
particular DCIS lesion can be predictive of the biological behaviour of the tumour. For
example, high grade lesions may be at a greater risk of local recurrence if incompletely
excised and be associated with the development of invasion (Lagios et al, 1982). Therefore
it is important to characterize the genetic changes of the different categories of DCIS, and to
determine the sequence of genetic changes in tumour evolution. The allelic loss studies
described by Fujii et al, (1996a) have suggested that the cumulative allelic loss for all loci
and LOH at each of the chromosomal loci studied (with the exception of chromosome 16q)
are significantly greater in intermediate or high-grade DCIS than in the low grade DCIS.
This is consistent with the model proposed by Walker (1997), whereby high grade DCIS
accumulates more genetic alterations than non-high grade DCIS, and in turn may progress to
a more aggressive form of invasive carcinoma than that of non-high grade DCIS.
Two of the groups of investigators who had investigated allele loss on chromosome 11 q 13
and chromosome 16q and 17p in DCIS, have extended their analyses to lesions of atypical
hyperplasia (ADH) and lobular carcinoma in situ (LCIS), as described below.
1.7.2
Lobular Carcinoma In Situ
LCIS and atypical lobular hyperplasia (ALH) are generally regarded as being markers of an
increased risk for breast cancer rather than direct precursors of invasive breast carcinoma
(Page and Anderson, 1987). A study by Nayar et al, (1996), noted 40% LOH at 11 q l 3 in
ILC and in LCIS associated with ILC. In all cases in which synchronous LCIS and ILC
from the same case showed LOH. the same allele was lost in both foci, similar to the
previous study of DCIS (Zhuang et a l, 1995). This provides molecular evidence for the
hypothesis that invasive lobular carcinoma develops from foci of LCIS. However, LCIS
without associated ILC showed a very low frequency of LOH, comparable to a study of
ALH (Chuqui et al, 1997). The authors proposed that the observed differences in LOH
frequency may indicate two sub-populations of ALH/LCIS, one that shows little genetic
changes and indicates lesions that do not progress or possibly regress, and the second
population, characterized by genetic alterations, which have the potential to progress to
invasive carcinoma.
Cases of pure LCIS, and LCIS with associated DCIS or invasive cancer were examined for
LOH at chromosomes 16q and 17p by Lakhani et al, (1995a). In this study, multiple loci
from seven cases were analyzed. In one case, two distinct foci of pure LCIS showed allelic
heterogeneity, suggesting the presence o f independent genetic clones, as described for
DCIS.
1.7.3
Atypical Ductal Hyperplasia
The study by Lakhani et al, (1995b), demonstrated LOH for loci on chromosomes 16q and
17p in approximately 55% (5 o f 9), and 20% (2 of 9) of ADH lesions, respectively. These
findings suggest that ADH is a monoclonal, and therefore neoplastic proliferation which
exhibits at least some o f the genetic changes which characterize established in situ and
invasive breast carcinoma. The incidence of LOH at chromosome 16q is similar to that
reported in DCIS (Stratton et a l, 1995) and LCIS (Lakhani et al, 1995a), suggesting that
ADH, LCIS and DCIS may share a common pathology of evolution, as illustrated in Figure
1.5.
1.7.4
Morphologically Normal Breast Tissue
A recent investigation has demonstrated that some of the genetic aberrations found in
invasive cancers are also present in morphologically normal breast epithelium from the
same patient (Deng et al, 1996). This study identified 8 of 30 cases with LOH, at three
chromosome regions in morphologically normal ducts. In each of these cases the same allele
was lost from the adjacent carcinoma cells. The highest frequency of LOH was observed at a
marker positioned at 3p22-25. LOH was not found in normal ducts from surrounding tissue
distant to the area of carcinoma, suggesting that the alteration was confined to normal ducts
surrounding the carcinoma. This has provided the first evidence that some invasive
carcinomas may develop from morphologically normal epithelium (Figure 1.4).
1.7.5
Susceptibility Genes
A major incentive for studying the genetics of hereditary breast cancer was the belief that
the genes responsible would also be somatically mutated in a proportion o f the non-familial
or sporadic cases of the disease, similar to findings for colorectal cancer. To date, somatic
34
mutations of BRCA1 and BRCA2 have not been identified outside of the familial cases.
However, recent evidence suggests that these genes may fulfill the criteria o f ‘caretakers’, as
proposed by Kinzler and Vogelstein (1997). These have demonstrated that BRCA1 and
BRCA2 bind to Rad51, a protein involved in maintaining genome integrity (Sharan et al.,
1997; Scully et a l , 1997). A model has been proposed whereby Rad51, BRCA1 and
BRCA2 act as a complex together to repair damaged DNA (Brugarolso and Jacks, 1997).
The ATM gene may also fulfill the criteria of a caretaker gene, since it has a central role in
surveillance of DNA damage. It is part of a signal transduction pathway which activates
cellular function in response to DNA damage (Meyn et al., 1995).
Further insight in to the development and progression of breast cancer is likely to come with
the identity of ‘gatekeeper’ genes. In addition to these factors, additional factors are also
required
to
promote
the
growth
of
breast
cancer.
Oestrogens
are
implicated
epidemiologically in this process and consequently the ER could be a key factor in breast
cancer development and progression. Increasing interest about alterations/mutations to the
ER gene and ER transcripts which result in altered structure/function of the receptor (to be
discussed in chapter 2) has made it apparent that mutants and variants do occur (section 2.8).
Their presence even in premalignant conditions such as typical hyperplasia raises the
question as to their role in the early stages of the disease and how they relate to other genetic
alterations
1.8
Hypothesis
Many studies have identified numerous genetic alterations in breast carcinomas, with
different alterations (profiles) possibly resulting in different pathways of evolution. Certain
critical genetic alterations must be necessary for the development of the disease. Although
multiple alterations may already be present in small, early stage or preinvasive breast
tumours, these will form a more relevant group for study than the larger, metastatic breast
carcinomas previously studied by others. Given the considerable body of evidence
demonstrating the importance o f oestrogens and the ER in breast cancer my hypothesis is
that alterations to the ER gene, including mutations, expression of splice variants and/or loss
of chromosomal regions to which the ER gene has been mapped could be critical for the
development of breast cancers. Analysis of aspects of the ER in early stage breast cancers is
the most appropriate way of confirming or refuting this.
35
1.4.5
Aims
To characterize 44 mammographically detected, node negative, 15mm or less, moderately or
well differentiated invasive carcinomas for steroid receptor expression and proliferative
activity
To screen this group for alterations/mutations to the oestrogen receptor using RT-PCR and
single stranded conformational polymorphism analysis and a non isotopic RNase cleavage
assay. Cases demonstrating altered ER species were to be characterized by sequence
analysis and their expression correlated with pathological features of these carcinomas.
The methylation status o f the oestrogen receptor gene promoter from these carcinomas, and
a selection of ER positive and negative cell lines, were to be investigated using a novel PCR
based approach.
To determine if loss of heterozygosity occurred at the region of chromosome 6, to which the
oestrogen receptor gene has been mapped (6q25.1), was present in cases of preinvasive
ductal carcinoma in situ and mammographically detected, node negative, 15mm or less,
moderately or well differentiated invasive carcinomas. Microdissected tumour foci were to
be obtained and analyzed using microsatellites and PCR to determine allele loss.
To determine if loss o f heterozygosity occurred at the region flanking 6q25.1 (6q25.1-27)
was present including loci to which putative tumour suppressor genes have been mapped, in
particular the M6P/IGF2R and TBP genes. The presence of LOH was then to be correlated
with pathological features of the tumour group
To determine simultaneous the evaluation of microsatellite instability, which if present, was
to be correlated with pathological features of the tumour group. These data were to
contribute to a parallel study carried out by a fellow Ph.D. student, Tom Walsh.
36
2.
THE OESTROGEN RECEPTOR
2.1
A n Historical Perspective
Oestrogens, e.g., oestradiol and oestrone, are steroid hormones produced by the adrenal
glands, ovary and in peripheral tissue, e.g., fat cells, which are involved in the control of
female sexual development, promoting growth and function of female sexual organs, e.g.,
menstrual cycle and secondary sexual characteristics, e.g., breast development.
Oestrogens have been implicated in breast cancer for almost a century following the
discovery of a relationship between the ovaries and malignant disease of the breast made in
1896 by George Beatson who described response of a locally recurrent cancer of the breast
following ovariectomy (Beatson, 1896).
The modem era of work on the subcellular mechanism of oestrogen action began in the late
fifties with the synthesis o f radioactive steroids (Glascock et al., 1959). They demonstrated
the selective accumulation o f tritium-labeled hexoestrol by the reproductive organs of
immature goats and sheep. Later this observation was extended to show the accumulation of
labeled oestrogens by mammary gland tissue, pituitary, hypothalamus, vagina (Jensen and
Jacobson, 1962) and breast cancer tissue from both humans and animals (Folcan et al.,
1961; King et al., 1963; Terenius et al., 1968; Pearlman et al., 1969).
Later studies eventually revealed that there was a protein present in normal oestrogen target
tissue and certain breast tumours that was responsible for the intracellular accumulation of
oestrogens (Gorski et al., 1968), and this protein became known as the oestrogen receptor.
In 1972, Jensen and his associates were able to propose a model of oestrogen action (Jensen
et al., 1972) based on the knowledge at that time which was subsequently supported by
other workers. The model was broken down into steps as follows: i) Oestrogen freely
diffused into the cell and ii) bound specifically to a cytoplasmic receptor (Gorski et al.,
1968; Jensen et al., 1968; Rochefort et al., 1980). iii) The receptor-hormone complex
underwent a transformation ("activation") (Jensen et al., 1968) whereupon iv) there was a
hormone- and temperature- dependent translocation into the nucleus (Jensen et al., 1972;
Williams and Gorski, 1972), determined by an equilibrium whereby the amount of nuclear
receptor was proportional to the total amount of hormone bound receptor in the cell (Jensen
and Jacobson, 1962). v) The complex then bound to DNA and stimulated oestrogen37
dependent effects including oestrogen receptor synthesis (Gorski and Gannon, 1976). vi)
The oestrogen receptor could then be released from the nucleus and recycled (Mester and
Beaulieu, 1975). Therefore it could be seen that the gateway of oestrogen action was the
receptor within target tissue.
The discovery of the oestrogen receptor led to speculation that presence might signal
hormone dependence of a tumour. This led to correlations being made with a number of
clinical parameters including age and menopausal status (Martin et al., 1979; Allegra et a l,
1980) biopsy site, i.e., primary or metastatic sites, (McGuire et al., 1975), histopathology
(type of tumour), e.g., lobular carcinoma being more frequently oestrogen receptor positive
(Rosen et al., 1975).
Several investigators applied cell kinetic values in relation to the study of oestrogen receptor
in breast cancer. The reason for the interest in kinetic studies was the hypothesis that "If
oestrogen receptor was a biochemical marker for the degree of tumour differentiation, it was
reasonable to speculate that it might correlate with an index of proliferative activity" (Harris
et al., 1991). For example, Meyer et al., (1977) used the technique of measuring tritiated
thymidine-labelling index to study proliferative activity of human breast cancer.
This
method estimated the proportion of tumour cells engaged in replication of DNA during a
brief incubation with [^H] thymidine. Tumours with a low index were more frequently
oestrogen receptor positive, whereas those with a high index infrequently contained
oestrogen receptor. Other workers used different methods of analysis such as flow
cytometry (Kute et al., 1981). Findings from these studies suggested that oestrogen receptor
content was not only associated with histologic features of differentiation, but also with
evidence of low proliferative potential. These results therefore provided a rationale for the
possible use of oestrogen receptor assays as a prognostic factor for recurrence and survival
of patients with operable breast cancer.
Investigators began to measure oestrogen receptors and correlate their presence with tumour
responses to endocrine treatment. The hypothesis that oestrogen receptors may be a marker
for endocrine responsiveness was soon confirmed by cumulative data from an international
meeting held in 1974 (McGuire et al., 1975).
38
Numerous laboratories using various methodologies to measure oestrogen receptor content
came to the same conclusion that the presence of oestrogen receptor was associated with
clinical responses to all endocrine therapies, either ablative, e.g., oophorectomy, or additive,
e.g., antioestrogens. More recent studies have confirmed these observations and extended
them to include clinical response to newer endocrine therapies, e.g., aromatase inhibitors
(Allegra et al., 1980; Rubins et al., 1980; Young et al., 1980).
Thus, patients with oestrogen receptor negative tumours infrequently benefited from
endocrine therapy and were therefore spared major ablative surgery or the long delays
waiting for a response to additive hormone treatment. These patients were then generally
considered as candidates for cytotoxic chemotherapy. On the other hand, one half to two
thirds of patients with oestrogen receptor positive tumours were found to respond to
endocrine therapy. In one large study, oestrogen receptor content was found to be the most
significant predictor for tumour response to endocrine treatment (Byar et al., 1979).
In 1985, the hypothesis originally postulated in 1972 by Jensen that lipid soluble oestrogen
molecules diffused through cell membranes into the cytoplasm where it bound to a
cytoplasmic receptor, underwent an activation before translocation of the entire complex to
the nucleus was updated. Immunocytochemical data from King and his colleagues indicated
that unbound oestrogen receptor actually resided de novo in the nucleus and that
’’cytoplasmic oestrogen receptor’’ activity was an artifact of tissue processing (King et al.,
1985).
In later studies the oestrogen receptor complex was found to bind tightly to the hormone
response element of target genes, and to promote the binding of a second complex to form a
dimer (Kumar and Chambon, 1988; Klein-Hitpass et al., 1989). When the oestrogen
receptor was activated by oestrogen, the complex was found to assume a conformational
change that promoted specific gene transcription. Examples of oestrogen receptor mediated
transcription include the progesterone receptor, and a number of growth factors such as
TGFa, IGF-II and even the putative growth inhibitor TGFp which were found to be
expressed by breast cancer tissue and secreted into the surrounding medium of cultured
breast cancer cells (Osborne and Arteaga, 1990). It is thought that progesterone receptor
may be important for control of critical metabolic processes in the cell and that growth
39
factors may be involved in regulatory events leading to cell proliferation (Harris et al.,
1991).
Knowledge of the oestrogen receptor (ER) status of a carcinoma is of value in aiding
prediction of hormone responsiveness, and can provide some prognostic information. In the
search for better predictors a number o f oestrogen-regulated genes have been identified,
such as pS2 (Maisiakowski, et a l, 1982), hsp27 (Edwards et al., 1981), PL 1VI (Manning et
al., 1994), which have been isolated from ER positive breast cancer cell lines after oestrogen
stimulation and by differential screening. These provide differing information with regard to
hormone responsiveness (Henry et al., 1991) and metastasis (Manning et al., 1994).
2.2
Structural Organization O f The Oestrogen Receptor Gene
2.2.1
5 ’ F la n k in g R e g io n
The ER gene seems to be regulated in a developmental, sex and tissue specific manner and
at various levels in different oestrogen responsive tissues. This suggests that gene expression
is tightly regulated, but until recently the underlying mechanisms were poorly understood.
In order to understand the transcriptional regulation of the ER, identification of the
transcriptional start sites were examined by a number of investigators.
Walter et al., (1985) and Green et al., (1986) were the first to describe the ER gene (Figure
2.1) and identified a 5’ untranslated region of 232 nucleotides with a single mRNA cap site
(PI), a TATA box-like sequence located 27 bp upstream from the putative start site with a
CAAT box like sequence situated at -103. A 5’ ATG codon, at position +119 was shown to
be followed 20 codons downstream by a TGA stop codon. It was not known whether this 20
codon open reading frame (ORFs) was expressed or had any physiological role. Short
upstream ORFs are a feature o f the steroid receptor multigene family and their conservation
implies some functional role, e.g., they may serve to reduce the translational efficiency of
the protein product. A second ATG codon, at position +233, and in the same translational
reading frame as the first was flanked by nucleotides which had a close homology to the
Kozak’s consensus sequence (Kozak, 1984) and therefore represented the start of the open
reading frame.
Based on homology with the mouse ER gene, Keaveney et al, (1991) postulated the
presence of an exon upstream o f the originally mapped cap site. Using RT-PCR on uterine
40
RNA, and primers within exon 1 and the potential upstream exon they established the
presence of a previously unidentified exon (exon 0) and an alternative mRNA cap site (P0)
and hypothesised the existence of two ER promoters. Exon 0 was located further 5’ at -1994
to -1884. When the alternative mRNA (mRNA 2) was transcribed, exon 1’ was found to
splice into the previously described exon 1 at position +164 indicating that two transcripts
existed for the hER gene (Fig 2.1). (Green et al, 1986; Keaveney et al, 1991; Grandien et
al, 1993). Identification of a major transcriptional start site was subsequently reported at 1994 (Piva et al, 1992; 1993) which resulted in a 110-base exon 0, giving rise to an ER
mRNA 54 bases shorter than the ER mRNA starting at PI.
Cloning of the 5' flanking sequence of the hER to nucleotide -2770 (Keaveney et al, 1992)
enabled the characterisation of the entire transcriptional unit of the gene and identified
potential binding sites for a number of trans-acting factors (Summarised in Table 2.1 and
Appendix II).
2.2.2
P r o m o te r E le m e n ts A n d U p s tr e a m B in d in g S ite s
A number of promoter elements have been identified in the 5’ region of the ER gene. They
include perfect and imperfect TATA box and CAAT-like motifs, initiator elements (INR),
Spl binding sequences, an SV40 enhancer core element, and a number of half palindromic
response elements. Many of these motifs have been found in the promoter regions of other
steroid receptors but as o f yet their significance is unknown.
2.2.3
P u r in e /P y r im id in e T r a c ts
The 5’ flanking region of the hER contains a number of alternating purine/pyrimidine (APP)
tracts. The longest is 70 bp in length at -1154 to -1084, over 1 kb upstream of the CAP site
for mRNA 1, which is largely AT rich containing a central d(AT) ] 4 and with nine d(GT)
repeats. The widespread distribution of APP sequences throughout both prokaryotic and
eukaryotic genomes suggests a biological function for these elements (Taute and Rentz,
1984; Greaves and Patient, 1985).
41
PO
<N
Intron 1
PI
v©
<N
<N
oo
Exon 1’
'
— Exon 1’
Exon 3
Exon 2
Exon 2
\ i
\ i
\i
/
v
N
Exon 3
Exon 2
Exon 3
mRNA 1
mRNA 2
(54 bases sh o rter than m R N A 1)
Figure 2.1 Schematic representation o f the 5 ’ region o f the oestrogen receptor gene. The transcriptional start site o f the ER gene for mRNA 1 begins at exon 1 and results in a transcript
(mRNA 1) with a 233 base untranslated leader. Upstream transcription for mRNA 2 from exon 1’ results in transcription o f upstream sequences that are represented by exon 1’ and are
spliced to the splice acceptor site at +164 within exon 1. This results in a transcript (mRNA 2) which contains a leader sequence derived from exons 1’ and 1 and is 54 bases shorter
than mRNA 1 (Adapted from Weigel et al., 1995).
T a b le 2.1 The sequence, position and proposed function of motifs within the 5’ flanking
region of the human oestrogen receptor gene (Position refers to the sequence described by
Keaveney et al, 1992)
M o tif S e q u e n c e
( 5 ’—>3’)
P o sitio n
P r o p o se d M o t if
F u n c tio n
TACTTAAA
-27
TATA Box
TTAATAAT
-2638
GCCAATGT
-103
CCAAT
-27270
CTCACTCT
-1926
CTCACTCC*
-2121
CTCAG*C*CTCT
-126
CCGCCC
-215
CCGCCC
-530
ATTTGCAT
-1691
ATTTGCAT
-1328
TGTGGT
-718
SV40 Enhancer Core
A GGTCA
-862
Half Palindromic ERE
TGACCT
-889
TGTTCT
-1757
CAAT BOX
INR
Spl
OTF-1
Half Palindromic GRE
Key * Denotes a mismatch; Italic sequences denote nucleotides with palindromic response
elements
43
2.2.2
Exon/Intron Structure Of The Oestrogen Receptor
The human genomic library prepared by Walter et al., (1985), spanning the entire ORF of
the hER, was sequenced, and demonstrated that the gene was split into eight exons with
sizes of 684, 191, 117, 336, 139, 134, 184. and 4537 bp respectively (6322 bp in total)
(Figure 2.2). Comparison of the intron/exon boundaries showed that each corresponded to a
characteristic splice consensus sequence (Breathnach and Chambon, 1981). A majority of
the clones did not overlap indicating that the hER gene was a very large gene with a
minimum length of 140 kb.
cDNA
Genomic DNA
>19 >16
>32
Figure 2.2 Organisation o f the human oestrogen receptor gene. The 8 exons are shown as vertical lines
(genomic DNA) with the minimum size o f each exon indicated (kb). The corresponding position o f the 8
exons (1-8) with respect to the ER cD N A is shown above (cD N A ) with numbers at the borders referring to
nucleotides (Green et al., 1986). The position o f the translation initiation (ATG) and termination (TGA)
codons are indicated. Shown above the cD N A is the division o f the hER protein into 6 regions (A-F, Krust et
al., 1986) together with the location o f the DNA (region C) and hormone (region E) binding domains. (Taken
from Pongliktmongkol et a l, 1988)
44
2.2.3
y Flanking Region O f The Oestrogen Receptor
An unusual feature of the nuclear receptor family is the sequence conservation of a very
large 3’ untranslated region (UTR), accounting for at least two thirds of the total length of
the mRNA, which exhibit a very high interspecies homology (Keaveney et al, 1993). This
sequence conservation thus strongly suggests a functional role for the 3 ’ UTR.
Sequence analysis has demonstrated that the pentamer ATTTA is over-represented in the 3’
regions of a number o f steroid receptor gene 3’ UTRs. Green et al., (1986) identified
thirteen copies of the pentamer in the 4305 bp of the 3’ ER UTR. Keaveney et al., (1989)
cloned and characterised a further 3276 bp, extending from the polyadenylation site at 6322
to position 9598, of the 3’ region of the gene and found that the 3’ UTR and 3’ flanking
sequences were 58.9 and 57.6% AT-rich respectively with extensive regions where this
percentage increased significantly. An additional eight copies of this motif were
subsequently identified in the 3’ flanking region of the hER gene (Keaveney et al., 1993).
The presence of multiple copies of the sequence AUUUA ((U)nA) in the 3’ UTR of various
mRNA molecules has been shown to have a highly destabilising effect (Caput et al., 1986;
Shaw and Kamen, 1986) and it has been postulated that this sequence could function by
conferring target sites for RNases which subsequently destabilise the message. Wreschner
and Rechavi (1988) compared 3’ UTRs from different mRNAs and showed that a number of
transiently expressed mRNAs were exceptionally (U)nA rich, and that this motif at least
partially mediated the rapid turnover. The presence of the ATTTA motif suggests that the
hER mRNA is potentially highly unstable. Moreover, these elements may feature in
regulation of tissue specific expression (Keaveney et al., 1993), as the relative activities of
RNases in different tissues have been implicated as another factor involved in the control of
tissue-specific gene expression. Thus it is possible to speculate that the hER gene and other
members of the steroid receptor family may be regulated in different cells at the post­
transcription level through differing stabilities of the message.
Two additional polyadenylation sites were present in the 3’ flanking sequence at positions
8205-8211 and 8440-8446 with unknown significance. Sequence analysis identified that a
GT element was present in the first 30 bp immediately after the polyadenylation signal at
position 6306-6312, followed by a T tract. These two motifs, in conjunction with the
45
AAUAAA sequence have been shown to be necessary for efficient polyadenylation
(McLauchlan et al.. 1985; Manley, 1988; Proudfoot, 1991).
Two inversely orientated Alu repeat sequences arranged in a head to tail arrangement from
8480 to 8774 and 8835 to 9146 have also been identified. The first repeat was flanked by
two 7 bp direct repeat elements whilst the second was flanked by two 14 bp direct repeats.
The two Alu repeats were shown to share 88.2 and 88.8% homology with the consensus Alu
family sequence (Kariya et al., 1987). The importance of the Alu repeats in the 3’ flanking
region of the hER gene remains to be determined.
Keaveney et al., (1993) identified a number of half palindromic EREs and GREs. Imperfect
EREs have the potential to operate in pairs to produce a synergistic oestrogen response,
provided they are in close proximity to each other (Martinez et al, 1987; Klein-Hitpass et
al., 1988b; Martinez and Wahli, 1989; Lannigan and Notides, 1990). Whilst a number of
half palindromic HREs were found in the 3’ flanking region, they represented two distal
halves of the HRE.
2.3
Conserved Sequence A nd Function O f The Oestrogen Receptor
The cDNAs for all the major steroid hormone receptors have been cloned and sequenced
(Evans, 1988; Beato, 1989) and comparison of these have identified interspecies homology
indicating functional importance. The human ER is a 66 kDa protein and comparison of the
human and chicken ER (cER) has identified three regions of strong homology termed A, C,
and E, separated by regions o f lower homology, B, D, F (Krust et al., 1986).
2.3.1
R e g io n C - T h e D N A B in d in g D o m a in (D B D )
Region C contains a core of 65 amino acids shown to be 100% conserved between the hER
and cER and in addition was characterised by a high cysteine and basic amino acid content
(Kumar et al., 1986). Eight cysteine residues were shown to be invariantly repeated, and it
was postulated that they folded into two zinc finger structures (Klug and Rhodes, 1987;
Evans, 1988), each resulting from the co-ordination of one Zn ++ to four cysteine residues,
and that this region corresponded to the DNA-binding domain (ER DBD, Green et al., 1986;
Krust et al., 1986) (Figure 2.3). Mutant forms lacking these structures did not bind DNA in
vivo or in vitro (Kumar and Chambon, 1988; Kumar et al., 1987). In addition it was also
shown that the zinc fingers alone were not sufficient for high-affinity DNA binding, as
46
mutants lacking a basic region located C-terminus to the zinc fingers were found not to bind
DNA efficiently in vitro (Kumar and Chamboa 1988). These observations indicated that
this region stabilises complexes of the receptor to DNA.
The hER DBD is encoded by two exons (2 and 3) with an intron located between the two
fingers (Ponglikitmongkol et al, 1988). The N-terminal finger (Cl) contains several
hydrophobic amino acids and four invariant cysteines, whereas the C-terminal finger (CII)
contains five invariant cysteines and is richer in basic amino acids.
HI [3
i
m
PAFR
[K]ET(RjYC
C U A F F g g S I Q G H N D YMC
m bi
CYEVG0MKGG1 ggD ggG G
I
I
I
216
221
245
■Exon 2 ---------------------------- 1
Exon 3
i
l
262
I
------------------------------------------------ Region C ----------------------------------------------------------------- 1
Figure 2.3 The hypothetical structure o f the human oestrogen receptor D N A binding domain (region C am ino
acids 180-262). The highly conserved 66 am ino acid core region (amino acids 185-250) is shown
schematically as two ‘zinc fingers'. H ighly conserved residues are shown in bold and basic residues are boxed.
(Taken from Ponglikitongkol et a l, 1988).
Studies using chimeric hER with altered DNA binding specificity were created by
substitution of the 65 amino acid core o f region C. An hER chimera containing a core region
of the hGR activated a glucocorticoid responsive gene in the presence of oestrogen (Green
and Chambon, 1987). Further analysis revealed that the amino acids critical for this
specificity were located within the Cl N-terminal half of the core region. ER mutants in
which only a few amino acids of Cl were mutated, indicated that the three amino acids (E203, G-204, A-207) located in the C-terminal end of Cl were critical in discriminating
between the oestrogen and glucocorticoid response elements (Madar et al., 1989).
The three amino acids in the Cl finger are probably central to this discrimination when the
finger interacts with the major groove o f the DNA (Schwable et al, 1990). The conserved
47
amino acids also within the region Cl may be important for this specific interaction with the
conserved nucleotides of the palindromic response element. It has been hypothesised that
zinc finger CII and its adjacent basic sequences are involved in nonspecific interactions with
the DNA sugar-phosphate backbone to give rise to nonspecific stabilisation of DNA binding
(Greene et al, 1988; Chalepakis et al, 1988).
2.3.1.1 Nuclear Localisation
A putative signal sequence, homologous to that of the SV40 large T antigen, was localised
in the proximal part and C-terminal domain of the ER (Picard et al, 1990). When these
amino acids were deleted, the receptor became cytoplasmic but could be shifted into the
nucleus by addition of hormone (Giochonmantel, et al, 1989) indicating that the DBD is
involved in nuclear localisation. In addition to DNA binding, the amino acids between 256
to 260 and 263 and 271, which form two lysine/argenine-rich motifs respectively, serve an
additional functions since they are also involved in the binding in vitro of the 90 kD heat
shock protein (hsp90, Chambraud et al, 1990) and correspond to two proto-nuclear
localisation signals o f the ER (Gronemeyer, 1991). In addition, hsp90 binding to this region
of steroid receptors is believed to mask the DNA binding domain, thus preventing binding
to response elements (Catelli, et a l, 1985) in the absence of ligand.
2.3.3
Region E - The Ligand Binding Domain (HBD)
2.3.3.1 Ligand Binding
A second region, region E, was 94% conserved between the hER and cER, and was
comprised of
250 amino acids encoded in five exons (4-8). This region was a
predominantly hydrophobic domain speculated to fold to form a hydrophobic pocket which
specifically binds hormone, known as the hormone binding domain (HBD, Green et al,
1986; Krust et al, 1986). Kumar et al, (1986), designed a number of mutants to test this
hypothesis, and found that the integrity of the region located between amino acids 305 and
552 to be indispensable for high affinity binding of oestradiol. Mutants lacking portions of
the domain resulted in impaired oestradiol binding irrespective of whether the hER mutants
were expressed in vivo or in vitro.
2.3.3.2 Dimerization
Gel shift assays with palindromic EREs coexpressed wild-type (wt) and mutant ERs in vivo
indicated that two receptor molecules bound to a single ERE and that no ER binding could
48
be detected using half site EREs (Kumar and Chambon, 1988). This indicated that proteinprotein interactions between two ER molecules was required for high-affmity binding (/. e.,
dimerization), suggesting the presence of a dimerization domain within the molecule. These
studies also demonstrated that a HBD-truncated ER molecule could not cobind to an ERE
with a wt molecule, whereas two such HBD-ER mutants could. In addition, HBD-truncated
ERs were shown to bind considerably more weakly to palindromic EREs than wt receptor.
This suggests that the HBD o f the wt molecules stabilised ER homodimers which
dissociated less rapidly than those o f mutants lacking the HBD, and that dimerization
stabilised ER/ERE complexes.
Subsequently, the presence of a dimerization domain was demonstrated by identifying
amino acids within the domain which were involved in both high-affmity binding in vitro
and dimerization in solution (Fawell et al., 1990). Lees et al., (1990) showed the exact
location of the dimerization domain by identifying a 22 amino acid peptide of the mouse ER
HBD that conferred high-affmity DNA binding onto a truncated mouse ER. Studies on the
three-dimensional structure of the glucocorticoid receptor DBD (GR DBD) complexes with
DNA, which have been shown to be similar to that of the hER (Schwable et al., 1993), have
demonstrated that the a-helix-tum-helix surface required for DNA binding is formed when
the receptor monomers dimerize (Luisi et al., 1991).
2.3.3.3 Hsp 90 Binding
It is well established that hsp90 is present in untransformed steroid receptor complexes
including that of the oestrogen receptor complex (Pratt, 1987; 1990). Initial observations
concerning receptor and heat shock protein interactions evolved from studies carried out
using sucrose density gradient centrifugation by Toft and Gorski (1966), which
demonstrated that steroid receptors recovered from hormone-free cells were largely 9S
complexes, whereas receptors recovered from hormone-treated cells generally sedimented at
4S. Because the 4S form of the receptor bound DNA with high affinity, whereas the larger
9S form did not, it was proposed that dissociation of the 9S complex was an initial event
resulting from steroid binding to the receptor. Purification of the 9S form allowed
characterisation of the complex and demonstrated a major protein recovered of
approximately 90 kD which was subsequently identified as the 90 kD-hsp (Pratt, 1993).
49
Studies on the GR demonstrated that hsp90 interacted with the HBD of the molecule and
that this region was sufficient to confer the property of hsp90 binding to chimeric proteins
(Scherrer et al., 1993). Chambraud et al., (1990) examined the binding of the hsp90 by
mutant forms of the hER expressed in COS cells with the recovery of receptor as 9S
complexes as a criterion of hsp90 binding. Like the GR they found that the ER HBD was
required for forming the 9S complex but no specific region of the HBD proved to be
uniquely involved. Picard et al., (1990) demonstrated that amino acids 256-303 of the hER
contained a number of basic amino acids that determined the nuclear localisation of the
receptor.
2.3.3.4 Transactivation Domain - TAFI1
Webster et al., (1988) demonstrated that deletion mutants lacking fragments of the ER HBD
activated transcription to a lesser extent than wt receptor, without complete loss of
transcriptional activity.
These results
suggested the presence of an autonomous
transcriptional activation function (TAF-II) within the HBD of the gene. Chimeric proteins
containing exons individually or in combination of the ER HBD linked to the yeast
transcription factor Gal4, were used to assess the ability to activate transcription from Gal4
responsive genes. No exons individually, or in combination, were found to activate
transcription and as a consequence the precise region responsible could not be identified
(Webster et al., 1989).
2.3.4
Region A/B- Transactivation Domain
The N-terminal A/B region is poorly conserved in both length and amino acid sequence and
is largely encoded by a single exon (exon 1). Deletion analysis of this region by Kumar et
al., (1987) and by Tora et al., (1989), demonstrated that transcriptional activation was
dramatically decreased when using certain oestrogen-responsive promoters indicating that
this region contained a transcriptional activating region (TAF-I). Analysis using A/B ER
mutants was difficult as TAF-I was shown to be only weakly active in some cell types
indicating a cell specificity.
2.3.5
Regions D and F
Using site directed mutagenesis, Kumar et al., (1987) demonstrated that the poorly
conserved domains designated D and F could be altered without affecting receptor function.
The length of region D, the hinge region, which joins the DBD and HBD, could be altered
50
without affecting receptor function, although in vivo this region seems to be functionally
important as Picard et al., (1990) demonstrated that the region encoding amino acids 256303 (which includes a small region of the ER DBD) contained a number o f basic amino
acids that determine the nuclear localisation of the receptor. In addition it was demonstrated
that complete deletion of region F did not affect hormone binding or transactivation.
In contrast, Montano et al., (1995) examined the transcriptional activity of a number of
deletion mutants completely lacking the carboxy-terminal F domain, or containing point
mutations. They demonstrated that this region affected the magnitude of liganded ER
activity in a cell-specific manner and was important in the transcription activation and
repression activities o f antioestrogens. The fact that this region makes liganded ER either
more o f less transcriptionally effective in different cells suggests that it plays an important
role in maintaining ER conformation, optimal for protein interactions needed for
transcriptional effectiveness. Its influence on the agonist/antagonist balance and potency of
antioestrogens further supports its specific modulatory role in the ligand-dependant
interaction of the ER with components of the transcriptional complex.
2.4
The Subunit Structure O f The Unliganded Oestrogen Receptor
Segnitz and Gering, (1995), deduced the subunit structure of the ER from the human
mammary carcinoma cell line MCF-7 with respect to the stoichemistry of protein subunits.
Using chemical cross-linking they isolated to homogeniety a receptor complex of
approximately 300 kD. After cleavage of cross links, only three polypeptide species were
detected either by protein staining or immunoblotting. In addition to a single receptor
polypeptide o f approximately 65 kD, they detected the hsp90 and p59, with the hsp90 band
approximately twice as intense as the p59 band, indicating a heteromeric structure consisting
of one receptor polypeptide, two hsp90 molecules, and one p59 subunit with a molecular
mass o f approximately 300 kD. Since the nonactivated receptor complex contained the
receptor monomer, rather than a dimer, this suggests that hsp90 and p59 prevent the receptor
from dimerizing and interacting with DNA.
As already discussed, in hormone treated cells, the receptor can be isolated as a smaller 4-5S
form, free of any detectable hsp90 indicating that the primary role of hormone binding is to
promote dissociation o f hsp90 from the hormone-binding receptor, allowing the receptor to
bind to ERE (Baulieu, 1987). Therefore it is generally accepted that hsp90 associates
51
selectively with unliganded receptors and inhibits DNA binding either by steric hindrance
and/or by passive interference with a receptor dimerisation step. Sabbah et al, (1996)
recently demonstrated that purified hsp90 caused ER-ERE complexes to dissociate and that
ER-ERE binding was inversely proportional to relative concentrations of hsp90, indicating
that hsp90 does not suppress ER function merely by steric hindrance.
2.5
Oestrogen Receptor Mediated Transcription Activation
A considerable amount o f evidence suggests that both ligands and phosphorylation play
important roles in the activation o f the steroid receptors, which are generally accepted to be
phosphoproteins, and that phosphorylation regulates the function of receptors. Evidence has
revealed that the progesterone receptor is allosterically activated by ligand and then becomes
a substrate for a series of separate protein kinases (Bagchi et al., 1992; Weigel et al., 1992;
Takimoto et al, 1992). Until recently it was an established belief that ligand binding was an
absolute requirement for the activation o f steroid receptors. But, considerable evidence has
accumulated that steroid receptors could also mediate other extracellular signals in the
absence of their cognate ligand.
2.5.1
B y L ig a n d B in d in g
Like many other steroid receptors, the ER is activated as a transcriptional factor by hormone
binding. As already discussed, in the absence of steroid, the ER exists as a oligomeric
complex containing p56 and hsp90 and which is thought to mask the DNA binding domain
(Catelli et al, 1985), prevent receptor dimerisation, and maintain the receptor within the
nucleus (Picard et al, 1990). The precise mechanism by which the ER regulates gene
expression upon ligand binding is unknown, but the three main steps appear to be i) ligandinduced activation of the receptor; ii) specific binding and stable complex formation at the
ERE; iii) recruitment of transcription factors and RNA polymerases to initiate transcription
of target genes.
2.5.1.1 Ligand Induced Activation O f The Oestrogen Receptor
Mendel et al, (1986) were the first to provide evidence that exposure of a steroid receptor
(the GR) to its cognate steroid hormone caused the dissociation of receptor from hsp90 and
subsequently the acquisition o f DNA binding activity (Meshinchi et al, 1990).
52
A number of researchers have examined the role of ligand in the activation of steroid
receptors. Binding of ligand to the progesterone and glucocorticoid receptors has been
shown to induce small changes in binding kinetics but not to the extent which could account
for hormone dependant gene activation (Schauer, et al, 1989; Rodriguez et al, 1989). Allen
et al, (1992) demonstrated that alternative conformations at the C-terminal end of the ER
HBD were induced by ligand and antiligands, suggesting that this region possesses a
regulatory function for transcription, since when it was available to the transcriptional
machinery, antagonist bound receptor was not active. In addition, deletion of this region has
been shown to give rise to a receptor which can be transactivated by antagonist.
2.5.1.2 Specific Binding And Stable Complex Formation At The Hormone Response Element
Analysis o f the promoter regions of hormone responsive genes have led to the identification
of their cognate enhancer elements, termed the hormone response elements, (HREs). These
short cis- elements are located within the 5’ flanking regions of these genes and confer
hormonal regulation upon receptor binding. The consensus sequence of the oestrogen
responsive element (ERE) 5’-GGTCAnnnTGACC-3’ is palindromic and contains a three
nucleotide nonconserved spacer between two halves of the palindrome (Klein-Hitpass et al,
1986, 1988a; Beato, 1989) and in some EREs, a T is also present at position 6 in the distal
half of the palindrome (Gronemeyer, 1991).
To date only a single perfect palindromic ERE, GGTCAcagTGACC, has been identified
from the vitellogenin A2 gene (Inoue et al, 1993), but recent studies have indicated that
many naturally occurring oestrogen responsive genes contain multiple sequences distinct
from this ERE, which function synergistically in a cell specific and promoter context to
regulate transcription, and may be o f more importance in facilitating ER function (Dana et
al, 1994; Anolik, et al, 1995; Webb et al, 1995). Due to the existence of a diverse group of
nonclassical EREs it is likely that a large number of genes which are directly regulated by
oestrogen remain to be identified.
Since EREs appear to be palindromic sequences, it was proposed that the ER binds to them
as a dimer (Lindstedt et al, 1986; Klein-Hitpass et al, 1988b). Receptor dimerization and
DNA binding was first confirmed by in vitro analysis (Kumar and Chambon, 1988) but
much analysis regarding dimerization by other workers required ER to bind DNA but gave
conflicting results as to whether oestrogen was required (Kumar and Chambon, 1988), or
53
not required (Klein-Hitpass et al., 1989; Curtis and Korach, 1990; Murdoch et al., 1990) for
high affinity binding of ER to the ERE. Using a yeast two-hybrid system, Wang et al.,
(1996) were able to demonstrate that ER dimerization was an oestrogen-inducible event in
vivo and that antioestrogens interfered with this process.
Martinez and Wahli (1989) demonstrated, using transient expression assays, that the
inductive capacity of the two EREs in the Xenepus vitellogenin B1 gene was much greater
than that o f the sum of either alone, i.e., they acted synergistically. In addition, they also
demonstrated that the spatial arrangement of the EREs was also critical, with ligand induced
co-operation occurring only when the EREs were separated by two helical turns.
Klinge et al., (1992a) demonstrated high-affinity co-operative binding of ligand activated
ER to three or four tandem EREs containing perfect inverted repeats and a 3’ AT rich
region, positioned such that alternate EREs were approximately 7 helical turns apart. In
contrast, no co-operative binding was observed when the EREs were 3.6 helical turns apart
suggesting that ER binds co-operatively only when EREs are present on the same face of the
DNA helix. 3’ AT-rich regions, also contribute to co-operative binding, with EREs lacking
these critical flanking sequences resulting in lower affinity and non co-operative binding.
These data suggest that the number, spacing, flanking sequences of EREs can produce fine
tuning o f the response of individual genes to oestrogen levels in vivo, through variations in
the degree of co-operative binding.
A certain degree of promiscuity has been observed for a number of HREs including the
ERE. The ERE half site motif 5’-TGACC-3’ is recognised by thyroid hormone (TR) and
retinoic acid receptors (RAR, Glass, et al., 1988; Gronemeyer, 1991). However the ER
activates transcription only from palindromic EREs with a fixed spacing of three nucleotides
between the half sites whereas the TR and RAR can recognise variably spaced half-sites
(Unesono and Evans, 1988, Unesono et al., 1989).
Only a small number of genes have been identified which are modulated directly by the ER,
including, pS2, PgR, Oxytocin, c-fos, and a 2 p-globulin (Klinge et al., 1992b), and there is
little information available regarding the interactions of the ER with EREs that deviate from
the classical palindromic sequences. This is despite the fact that nonclassical ERE sequences
predominate in endogenous genes. It is believed that sequences distinct from the classical
54
vitellogenin A2 ERE, which function in a cell and promoter specific manner, may be of
more importance in facilitating ER function.
Analysis to define the sequences in the BRCA1 gene which are responsible for its oestrogen
responsiveness in cultured breast cancer cell (Love et al, 1992) has led to the identification
of another class of ER-dependant transcriptional enhancers, a new subclass of Alu DNA
repeats (Norris et al, 1995). Within this sequence an imperfect ERE containing two base
pair changes, and a 5' half site 9 bases pairs downstream were identified. Additionally, two
half sites separated by 22 base pairs were found in non-Alu sequence associated with the
sequence. These sequences were shown to independently function as ER-dependent
enhancers when assayed in mammalian cells with the activity of the composite region
greater than that of the independent regions. This seems to suggest a synergism between
independent oestrogen responsive sequences in a similar manner to that shown for the
vitellogenin B2 ERE, where two imperfect palindromes co-operate to form a functionally
important element, but do not function independently (Klein-Hitpass et al., 1988b).
An oestradiol-induced increase in AP-1 activity (Fos and Jun dimer complexes) was
demonstrated by Gaub et al. (1990) in a study on oestradiol induction of ovalbumin in chick
embryo fibroblasts in which ER was introduced transiently by transfection. The authors
revealed that the putative ERE sequence responsible (5’-TGGGTCA-3’) differed by only 1
nucleotide from the consensus AP-1 site (TGAGTCA) and canonical AP-1 sequence from
collagenase. Transcription experiments by the same authors revealed that c-fos, c-jun and
oestradiol/ER complexes coactivated this part of the ovalbumin promoter. Direct interaction
with the target sequence was not found to be required, since an ER deleted of its DNA
binding domain was functional in the co-activation with c-fos and c-jun.
Other studies demonstrated increased AP-1 activity by oestradiol in MCF-7, T47-D and
ZR75 ER positive cell lines which was modified by epidermal growth factor (Phillips et al,
1993). Using similar experimental conditions, oestradiol decreased AP-1 mediated
transcription in ER-negative cell lines MDA-MB-231 (breast cancer) and HIH3T3 (mouse
fibroblast) cells in which ER had been transiently transfected (Chalbos et al, 1994).
Similarly, oestradiol also decreased AP-1 activity in HeLa (human cervix cancer ) and CV1
(monkey kidney cells) cells, which were cotransfected with c-fos, c-jun and an ER
expression vector (Shemshedini, et al, 1991; Ambrosino, et al, 1993). The mechanism by
55
which oestrogen ER complex has opposite effects on AP-1 activity in ER negative and ER
positive cell lines is unknown, but the recent cloning of a second ER (ERp) by Kuiper et al,
(1996) may give insight. The human homologue was subsequently cloned (Mosselman et
al, 1996) and the transactivation properties of the 2 receptors (ERa and ERp) examined
with different ligands in the context o f an ERE and an AP-1 element (Paech et al, 1997).
ERa and ERp were shown to signal in opposite ways when complexed with oestradiol from
an AP-1 site. With ERa, oestradiol activated transcription whereas ERp it inhibited
transcription. In addition a number o f antioestrogens were potent activators with ERp at AP1 sites. These data therefore indicate that the two ERs signal in different ways depending on
ligand and response element and that the 2 receptors play different roles in gene regulation.
2.5.1.3 Recruitment O f Transcription Factors And RNA Polymerases To Initiate
Transcription O f Target Genes
As already discussed, a number o f deletion experiments have demonstrated the existence of
two domains within the ER region A/B and HBD of an autonomous TAFI and TAFII,
respectively, responsible for the transcriptional activation of target genes. Deletion of these
regions resulted in transcriptional activation to a much lesser extent than the wt receptor
(Webster et al, 1988: Tora et al, 1989). Chimeric receptors containing the transcriptionally
inactive Gal4 DBD and either regions o f the ER HBD or A/B region were used to assess the
transcriptional activation of Gal4-responsive genes in chicken embryo fibroblasts (CEF,
Tora et al, 1989), HeLa cells (Berry et al, 1990), and in yeast (White et al, 1988; Metzger
et al, 1988). ER TAF-I activated transcription in CEF but not HeLa cells, whereas ER
TAFII activated transcription in HeLa and CEF cells, but not in yeast, indicating that each
TAF may have a distinct pattern o f cell specificity which may be related to its function in
viv o .
Tora et al, (1989) and Berry et a l, (1990) demonstrated that the nature of the promoter of a
given target gene also influenced ER TAF activities. Transcription from reporter genes, (e.g.
Vit-tk-CAT), containing complex promoters with several upstream elements, were strongly
enhanced by TAF-II in the presence o f oestradiol whereas TAF-II weakly stimulated a
minimal promoter composed o f the adenovirus major late TATA box region placed down
stream of a palindromic ERE. In contrast, TAF-I was shown to be less affected by such
variations.
56
In most promoter contexts, synergism between TAF-I and TAF-II is required for full ER
activity. Kraus et al, (1995) demonstrated that when the two TAF containing regions were
expressed as separate polypeptides in mammalian cells, they functionally interacted in
response to
17p-oestradiol and antioestrogen binding. The interaction was only
transcriptionally productive in response to 17p-oestradiol, and was eliminated by point or
deletion mutations that destroyed TAF-I, TAF-II or 17P-oestradiol binding. These results
suggest a definitive mechanistic role for 17p-oestradiol in the activity of the ER, perhaps
altering its conformation to promote association of the amino- and carboxyl-terminal
regions, leading to transcriptional synergism between TAF-I and TAF-II and providing a
mechanism for hormonal regulation of transcription.
Cell-specific activation of transcription by TAF-I and TAF-II has been shown to be both cell
and promoter specific and it is thought that this is due to interactions, either directly or
indirectly, with other cell specific transcription factors that are required to mediate TAF
activity to the basic transcriptional machinery.
It has been observed that transcriptional interference (squelching) has been observed
between the activation functions of the various steroid receptors (Bocquel et a l, 1989;
Meyer et al., 1989; Tasset et al, 1990). The activity of TAF-II of a given steroid receptor
can be squelched by over-expression of the TAF-II containing region E of the same receptor
(auto-interference) or of a different steroid receptor (hetero-interference) in the presence of
the cognate ligand. These transcriptional interferences together with the promoter and cell
context dependency of TAF-I and TAF-II activities are mediated by transcriptional
mediators or intermediary factors (TIFs) interacting with the A/B and E regions of the
receptor respectively (Tasset et al, 1990).
These protein-protein interactions are thought to enhance the formation of a pre-initiation
complex by RNA polymerase and signal the initiation of transcription (Hawley and Roeder,
1985). Using a series of pre-incubation and template competition analyses combined with
kinetic analyses, Klein-Hitpass et al, (1990) were able to dissect the mechanism of action of
the chicken progesterone receptor in an in vitro system. It appears that the ligand activated
receptor enhances the assembly of a template-committed complex of transcription factors at
the core promoter (Fig 2.4) by facilitating recognition of the promoter or simply by
57
stabilizing the DNA-protein complex once it is formed. This stable complex is then poised
for rapid initiation of transcription by RNA polymerase.
Multiple factors must interact at the proximal promoter, i.e., TATA box, of a regulated gene
to allow initiation of transcription (Van Dyke, et al., 1989). The sequence of events is
thought to be dependent on the initial reversible interaction of the TFIID protein with the
TATA sequence itself. The interaction is subsequently stabilised by sequential binding of
TFIIA, TFIIB but the precise role of these factors in the initiation complex are only partially
known (Burtowski, et a l, 1989). However, together they form a three-dimensional surface
conformation which attracts RNA polymerase and other transcription factors such as TFIIE
or TFIIF to the gene and transcription begins.
2.5.1.4 Modulation O f The Oestrogenic Response By Membrane Receptor Ligands
As already discussed, the response to oestrogen depends upon several important factors
including, the ligand, the promoter, and the cell context. There is however, evidence that
phosphorylation may be another mechanism by which the ER is regulated. The finding that
growth factors, activators of protein kinase A and C, and phosphatase inhibitors can induce
hyperphosphorylation and as a consequence increase the transcriptional activity of the
human ER complicates the picture further. This indicates that second messengers signaling
pathways then indirectly activate protein kinases, which alter the cellular phosphorylation
state that may in turn determine the biological effectiveness of the oestrogen-occupied ER.
Despite numerous investigations, the mechanisms by which these signaling pathways *
interact have not been determined. Numerous studies have demonstrated up-regulation of
intracellular progesterone receptor by growth factors that stimulate tyrosine kinase activity
of transmembrane receptors including the insulin-like growth factor-I (IGF-I), epidermal
growth factor (EGF), and phorbol ester in MCF-7 human breast cancer cells and uterine
cells (Aronica et al., 1991, 1993; Katzenellenbogen
and Norman, 1990; Sumida and
Pasqualini, 1989; 1990). The fact that the stimulation was blocked by antioestrogens
suggests that these agents act through the ER pathway (Aronica and Katzenellenbogen,
1991; Katzenellenbogen and Norman, 1990; Cho et al, 1994; Sumida et al., 1988; 1989).
In addition, the fact that protein kinase inhibitors blocked the effects of these
58
TATA
HRE
IID, IIA, ER
TATA
ER E
IIB
P ol I-F, E, H and J
P ol II
TATA
TATA
Figure 2.4 R ole o f the oestrogen receptor in gene activation. The general factor TFIIB is a target for
transactivation domains o f steroid receptors. The promoter o f a gene regulated by oestrogen receptors is
shown, with c/s-elem ents indicated: oestrogen responsive element (ERE) and TATA box. A ctivated receptor
(ER) binds its cognate c/s-elem ent as a dimer. The general transcription factor TFIID binds to the TATA
elem ent along with TFIIA. The binding o f these factors is reversible until TFIIB join s the protein-DNA
com plex to stabilize them. D N A looping brings the com plexes together. The stable framework allow s
subsequent binding o f RNA polym erase II (Pol II) and other general transcription factors TFIIF, E, H and J
efficiently to com plete formation o f the preinitiation com plex and rapidly transcribe an activated gene. (Taken
from O ’M alley and Tsai, (1993), In Steroid Hormone Action, Edited by M.G. Parker)
59
suggests the involvement of phosphorylation in this response.
Transactivation of ERE-mediated responses by increasing cellular cyclic adenosine
monophosphate (cAMP) have also been demonstrated including dopamine in CV1 and
HeLa cells (Power et al, 1991; Smith et al., 1993). No specific phosphorylation site has
been shown to be implicated for cAMP activation of the ER although synergism was
decreased using HBD mutants that showed diminished ligand-dependent transcriptional
activation (Cho and Katzenellenbogen, 1993).
2.6
Phosphorylation O f The Oestrogen Receptor
Phosphorylation is a common covalent modification of proteins which can greatly modulate
their function. The steroid/thyroid receptor superfamily have been shown to be
phosphorylated in vivo (Aurrichio, 1989; Beato, 1989; Gilneur et al, 1990; Fahmer et al,
1990; van Laar et al, 1991; Rochette-Egly et al, 1991; 1993; Gaub et al, 1992) generally
occurring on serine residues, although tyrosine phosphorylation has also been observed for
the ER (Aurricchio et al, 1987). These phosphorylations are often ligand induced (e.g., on
serine resides) although constitutive phosphorylation sites have also been shown to be
present (Washburn et al, 1991; Denton et al, 1992; Ali et al, 1993). Consequently a
number of studies have been undertaken to clarify the role of phosphorylation in ER
function.
2.6.1 S e r in e P h o sp h o r y la tio n O f T h e O e str o g e n R e c e p to r
Using MCF-7 cells or calf uterus, Denton et al, (1992) demonstrated an oestrogen
dependent 4-fold increase in phosphorylation of the ER at serine residues. Potato acid
phosphatase treatment o f the receptor resulted in dephosphorylation and a subsequent
decrease in the receptors affinity for DNA sequences, suggesting that transcriptional
activation by the ER involved an oestrogen dependent phosphorylation of the receptor. This
in turn resulted in its increased affinity for specific DNA sequences. This oestradiol induced
increase in affinity of the hER for its ERE was confirmed by Arnold et al, (1994) who
identified ser-167 as the major oestrogen induced phosphorylation site on the hER. In
addition, they demonstrated that oestradiol binding enhanced by 2-fold the phosphorylation
of the hER by casein kinase II in vitro, demonstrating that the hER is phosphorylated on
serine-167 by casein kinase II in a hormone dependent manner.
60
Ali et al., (1993) investigated the phosphorylation state of the hER in the absence or
presence of different ligands and showed that the hER was phosphorylated at low levels in
the absence of ligand. Oestradiol induced phosphorylation 8-fold, and antioestrogens
induced phosphorylation to a lower extent. Using deletion mutants, two main regions of
phosphorylation were identified, the N-terminal A/B region and the C-terminal D/E/F region
where the majority of sites were found. Detailed analysis mapped one of these sites within
the N-terminal region at serine-118. Mutation of this residue resulted in a significant
reduction in transcriptional activation by a hER from reporter genes containing an ERE,
present in a number of cell types, indicating that phosphorylation of this residue was
important in the action of the transcription activation function 1 (TAF-1) located within the
A/B region.
Lahooti et al., (1994) identified a number of other serine residues between residues 121 and
599 within the mouse ER and demonstrated that phosphorylation was progressively reduced
when critical amino acids were replaced to disrupt the HBD and DBD, suggesting that
phosphorylation occurs in stages, initially as a consequence of ligand binding and
subsequently after DNA binding.
Detailed analysis of this region indicated that ser-122, ser-156, ser-158 and ser-298 were
phosphorylated (Lahooti et al., 1995). Transcriptional activity of a number of substitution
mutant receptors at these residues were then assessed in transfected cells. Substitution of
ser-122 and ser-298 had negligible effects on transcriptional activity, however, a reduction
was observed when ser-122 was mutated. These results confirmed those observed by Ali et
al., (1993) since the ser-122 in mouse ER is equivalent to ser-118 in the hER.
Danielian et al., (1992) had previously identified and demonstrated the importance o f a
putative amphipathic a-helix for hormone dependent transcription and showed that despite
abolition of transcription by a mutation in this region it could be rescued by the presence of
TAF-1, indicating cooperativety between TAF-1 and TAF-2. Lahooti et al., (1995)
demonstrated that the transcriptional activity of a full length receptor missing charged
residues within the putative amphipathic a-helix of a full length receptor was similar to that
of the wt receptor, but was reduced when ser 122 was mutated confirming that ser 122 is
required for TAF-1 to cooperate optimally with TAF-2.
61
2.6.2 Tyrosine-537 Phosphorylation O f The Oestrogen Receptor
Magliacco et al, (1986) proposed that hormone binding of the calf uterine ER was
controlled by an oestradiol-dependent tyrosine kinase phosphorylation.
Subsequently, a
partially purified oestradiol-dependent tyrosine-kinase and a nuclear tyrosine phosphatase
were subsequently purified that could activate and inactivate, respectively, the hormonebinding capacity of the ER (Aurricchio et al., 1987). These researchers then identified, using
site directed mutagenesis, that tyrosine-537 of the hER was the site of phosphorylation in
vitro by the oestradiol-dependent kinases (Castoria, et al., 1993).
Arnold et al., (1995a, b) carried out a series of experiments to define the role of tyrosine-537
phosphorylation in the transcriptional activation of the hER. They
identified a
phosphorylation site at tyrosine-537 in hER from MCF-7 cells and recombinant ER
expressed in Sf9 insect cells using amino acid sequencing of 32P labeled tryptic peptides of
the
hER.
Using
phosphotyrosine
antibody
they
demonstrated
that
tyrosine-537
phosphorylation was independent of oestradiol treatment of MCF-7 cells indicating that
tyrosine-537 was a basal phosphorylation site (Arnold et al, 1995a). They subsequently
demonstrated that the DNA binding form of the hER was phosphorylated at tyrosine-537,
was localized in the nucleus o f oestradiol treated MCF-7 cells, with an apparent molecular
mass of 67-kDa and hyperphosphorylated on serine residues. The non-DNA binding form
was devoid of phosphorylation at tyrosine-537, cytosolic with an apparent molecular mass
of 66-kDa, and hypophosphorylated on serine residues (Arnold et al, 1995b).
Dephosphorylation of purified hER at phosphotyrosine-537 with a tyrosine phosphatase was
shown to eliminate binding to an ERE in a gel shift mobility assay which was restored by
rephosphorylation of tyrosine-537 with Src family tyrosine kinases, p60c'vn: or p56/c*.
Subsequent mutation of tyrosine-537 confirmed that its phosphorylation was required for
hER to bind to ERE in vivo. Interestingly, the tyrosine kinase activity of p60csrc in human
breast cancers has been shown to be elevated as compared with other cancers (Jacobs and
Rubsaamen, 1983). The results of an ‘antibody rescue experiment’ suggested that the
tyrosine-537 phosphorylation regulated the transition from the monomer to dimer transition
of the receptor, the premise being that the anti-hER is bivalent and recognizes two
molecules of hER. Therefore the antibody can bring together two inactive hER monomers
and facilitate dimerization and DNA binding (Arnold et al, 1995b).
62
From their findings, Arnold et al., (1995b) hypothesised that the 67 kDa form of the hER
was the transcriptionaly active form o f the receptor. In the absence of oestradiol treatment of
MCF-7 cells, the 66 kDa form o f the native hER exists as two populations: one that is
phosphorylated at tyrosine 537 and one that is not. After oestradiol treatment only the
tyrosine-537 population undergoes serine hyperphosphorylation giving rise to the 67 kDa
form of the hER indicating its requirement for oestradiol binding. Therefore the DNA
binding ability of the hER seems to be post-translationally regulated a two levels: first
phosphorylation of tyrosine-537 and the second the phosphorylation of ser-167 (Arnold et
al., 1994) which increase the hER affinity for the ERE.
2.6.2
Phosphorylation Of The Oestrogen Receptor In The Absence Of Ligand
Until relatively recently, it was an established belief that ligand binding was an absolute
requirement for the activation of steroid receptors. However, a considerable amount of data
has accumulated which suggests that steroid receptors also mediate other cellular signals in
the absence of their cognate ligand.
A number of agents and extracellular signals which affect protein kinases and cell
phosphorylation, and which have been shown to activate the ER, i.e., the induced expression
of the PgR, pS2 and cathepsin D (Cavailles et al., 1988, 1989; Nunez et al., 1989; Sumida
and Pasqualini, 1989; Katzenellenbogen and Norman, 1990) in the absence of ligand are
summarized in Table 2.2. The mechanism of oestrogen receptor activation by growth factors
is not yet understood but it has been proposed that ligand-independent activation of steroid
receptors may represent activation of non-receptor transcription factors that synergise with
members of the steroid receptor superfamily to increase target gene expression in the
absence of steroid ligand (Kushner et al., 1994).
In spite of many studies, both the signaling pathways and the target domains of the ER are
essentially unknown . Bunone et al., (1996) carried out analysis to elucidate the mechanism
'J
of steroid independent activation of the ER by the epidermal growth factor and
demonstrated that TAF-I was necessary for the EGF response. In addition, both the
hyperphosphorylation and transcriptional activation of unliganded ER depended on a
phosphorylatable serine residue at position 118, although its phosphorylation alone was not
sufficient, implying the presence o f other target domains or proteins to fulfill the EGF
signaling through the ER. Using dominant negative Ras and MAP kinase kinase and
63
Table 2.2 Agents and extracellular signals which activate the oestrogen receptor in the
absence of ligand
Agent / Extracellular Signal
Reference
Dopamine
Power et al., 1991; Smith et al., 1993
Epidermal Growth Factor
Ignar-Trowbridge et al, 1993
Heregulin
Pietras et al, 1995
Transforming Growth Factor a
Ignar-Trowbridge et al, 1993
Insulin and Insulin-like Growth
Factor I
Aronica and Katzenellenbogen, 1993; Ma et al,
1994; Newton et al, 1994
Okadaic Acid
Power et al, 1991
TPA (Phorbal Ester)
Bunone et al, 1996
cAMP, Protein Kinase A (Cholera
Toxin)
Aronica and Katzenellenbogen, 1993
constitutive MAP kinase kinase, they also demonstrated that EGF activated the ER by
signaling through the MAP kinase pathway, implying that MAP kinase directly
phosphorylates the critical serine 118.
Other steroid-independent activators of the ER, such as insulin-like growth factor I and
agents which increase the concentration of cAMP have also been shown to induce
hyperphosphorylation o f the unliganded ER (Table 2.2) and at least in the case of TP A the
target appears to be serine 118 (Joel et al, 1995).
As already discussed previously, activation by TAF-I alone is not sufficient, especially in
the case of HeLa cells (Tora et al., 1989; Ali et al., 1993). Yet wt ER can be activated in an
TAF-I dependent manner even in HeLa cells (Bunone et al., 1996). This lead the authors to
propose that EGF is able to render TAF-I activity so that it is less dependent on cell type and
promoter context by a mechanism which does not solely rely on the hyperphosphorylation
of serine 118.
64
2.7
Alterations To The Oestrogen Receptor In Breast Cancer
The response of breast cancer patients to endocrine therapy is limited due to the
development of hormone-independence tumours in virtually all cases. The loss of hormone
dependency has been attributed to mutated or truncated oestrogen receptors that act
constitutively, i.e., enhance transcriptional activity in the absence of hormone (Sluyser and
Mester, 1985). Analyses at the mRNA/cDNA level have demonstrated a variety of somatic
mutations/variants in human breast tumours and cell lines, which are summarised in Table
2.3
On the basis of knowledge regarding the functional domains in the ER, it is not surprising to
find that mutations to the receptor can be generalised into the following variant categories:
Dominant positives, Dominant negatives, ‘Negatives’, and Miscellaneous variants.
2.7.1
D o m in a n t P o sitiv e O e str o g e n R e c e p to r s
The first category comprises ERs with constitutive activity due to loss o f TAF-II activity,
but retention of TAF-I activity and are known as dominant positives. The dominant positive
ERs with deletion of exon 5 seems to be such a case (Graham et al., 1990; Fuqua et al.,
1991; Castles et al., 1993; Castles et al., 1995). Loss of this exon results in a shift in the
reading frame and ultimately protein truncation with novel amino acid sequence distal to the
exon deletion (out of frame deletion). In expression systems, this variant exhibits
constitutive transcription-activating activity in the absence of hormone, Transfection of this
variant into MCF-7 cells (Fuqua and Wolf, 1995) demonstrated that cells expressing the
truncated receptor constitutively expressed PgR, and exhibited anchorage independent
growth which was not inhibited by tamoxifen or 4-hydroxytamoxifen, unlike MCF-7 cells
transfected with control plasmid, but was inhibited by the pure antioestrogen ICI 164,384.
These findings are in contrast to those of Rae and Parker (1996) who demonstrated only
weak transcriptional activity, with no effect on the expression of ER target genes in MCF-7
cells. In addition, they also showed that this variant was able to stimulate transcription of a
reporter gene in chick embryo fibroblasts in the absence of hormone, concluding that its
expression alone was not sufficient to give rise to hormone independence or antioestrogen
resistance.
Initially, the exon 5 variant was first cloned from ER-/PgR+ breast tumour tissues and T47D cells because a dominant-positive variant was suspected (Graham et al., 1990; Fuqua et
T a b le 2.3 Oestrogen receptor variants in human breast cancer tissue and cell lines3(adapted front Sluyser, 1995)
Key 3 Cell lines studied: T47Dco, MCF-7. ZR75-1, BT-20, BT-474, MDA-MB-134, MDA-MB-330, MDA-MB-361; b From cap site
Dom ain
A/B
RNA C hange1*
C -»T at 489
Ala->Val at 86
C
/nmv-Splice at exon 2/intron
boundary
Effect on ER Function
Decreased E2 binding
R eferences
Garcia et al, (1989)
6 unique aa inserted after 214;
truncated 23.9 kD ER
Unknown
Dotzlaw et al (1991,1992)
Out o f frame exon 2 deletion
16.7 kD ER truncated at 152
Negative
Wang and Mikscicek (1991)
fra/w-Splice at exon 3/intron
boundary
84 unique aa inserted after
253; truncated 37 kD ER
Unknown
Dotzlaw et al, (1992)
TT insert at 982
ER truncated at 251
No E2 binding; no nuclear
localisation
Graham et al, (1990); Leslie et al,
(1992)
In-frame exon 3 deletion
61.2 kD ER (556 aa)
Dominant negative in HeLa
cells; negative in yeast cells
Wang and Mikscicek (1991);
Fuqua et al, (1993)
Duplication o f exons 3 and 4
83.3 kD ER
Unknown
Murphy et al, (1996)
In frame exon 4 deletion
254-366 deleted (53.7 kD ER)
Negative
Pfeffer et al, (1993)
Point mutation in exon 4
Lys-»Arg
Hyperactive
Fuqua et al, (1996)
In-frame 1004-1462 deletion
259-412 deleted
No E2 binding; no nuclear
localisation
Graham et al, (1990)
D
T—»C at 1120
Leu—>Pro at 296
Unknown
McGuire et al, (1991)
E/F
G ->T at 1281
Asp—>Tyr at 351
Stimulated by tamoxifen
W olf and Jordan (1994a, b);
Catherino and Jordan (1995)
P ro tein C hange
Tabic 2.3 Continued Oestrogen receptor varians in human breast cancer tissue and cell lines3 (adapted from Sluyser, 1995)
3 Cell lines studied: T47Dco, MCF-7, Z.R75-1, BT-20, BT-474, MDA-MB-134, MDA-MB-330, MDA-MB-361, bFrom cap
Dom ain
E/F
RNA C hange6
A—>1288
P ro tein C hange
Glu->Val at 352
Effect on E R Function
Stimulated by tamoxifen
R eferences
Karnik et al, (1994)
Out o f frame exon 5 deletion
ER truncated at 371 (42 kD)
Dominant positive; tamoxifen
resistance
A->G at 1418
M et—>Val at 396
Unknown
Fuqua et al., (1991; 1992); Castles
et al., (1993); Zhang et al., (1993);
Castles et a l, (1995)
Roodi et al., (1994)
69 base in frame insertion
23 novel aa inserted between
412 and 413 (68.8 kD)
Unknown
Murphy et al., (1996)
G deleted at 1463
ER truncated at 419
No E2 binding
Graham etal., (1990)
1503-1550 replaced by 13801422
ER truncated at 455
No E2 binding
Karnik et al., (1994)
T deletion at 1526
ER truncated at 438
Tamoxifen resistance
Karnik et al., (1994)
Out o f frame exon 6 deletion
ER truncated at 473
Unknown
Miksicek et al., (1993)
Duplication o f exon 6
ER truncated at 462 (51.4 kD)
Unknown
Murphy et al., (1996)
Out o f frame exon 7 deletion
ER truncated at 466 (51.3 kD)
Dominant negative
Wang and Miksciek (1991); Fuqua
et al, (1992); Castles etal., (1995)
Insert of bases at 1932
80 kD ER
Unknown
Pink e ta l, (1994)
In frame duplication o f exons
6 and 7
T—>A
77-80 kDa ER
No E2 binding; down
regulates wt receptor
Stimulated by tamoxifen
Pink et al, (1996, 1997)
Tyr—>Asn at 537
Z hang e t a l , 1996
al.. 1991; Castles et al., 1993). It eventually transpired that it was widely distributed in
primary breast tumours, not only ER-/PgR+ but also ER+/PgR+ or ER+/PgR- tumours
(Zhang et al., 1993). In addition, the variant was always co-expressed with and often in
excess of, wild-type ER. The exon 5 variant was also detected in three established breast
cancer cell lines (MCF-7, T47D, and ZR75), although to a lesser extent than wt ER.
Daffada et al., (1995) investigated the expression of this variant at the mRNA level in
relation to known tamoxifen resistance and the expression of PgR and pS2. Although they
observed no significant difference between exon 5 deleted to wt ER between tamoxifenresistant and control tumours, tumours in both groups which expressed PgR/pS2 in the
absence o f measurable ER protein had significantly higher exon 5 deleted ER mRNA levels
compared with other phenotypes. In addition, in ER positive tumours which expressed pS2,
a significantly greater exon 5 deleted ER mRNA expression was observed in tamoxifenresistant compared with control tumours. These data indicate that whilst exon 5 deleted ER
was unlikely to be responsible for tamoxifen resistance in most cancers, elevated expression
of this variant could be important in some tumours, especially those which continued to
express high levels of PgR and pS2.
2.7.2
Dominant Negative Oestrogen Receptors
The second category comprises ERs that are inactive by themselves but which more
importantly interfere with the activity of normal wt receptor in a trans-dominant fashion. A
good example is the truncated 51.3 kD exon 7 deletion mutant described by Wang and
Miksciek (1991) and by Fuqua et al., (1992). Like the exon 5 deletion variant, precise
deletion of exon 7 results in a shift in the reading frame and protein truncation with novel
amino acid sequence distal to the exon deletion (out of frame deletion). Mikscieck et al.,
(1993) demonstrated that an exon 3 deleted ER also belonged to this category, since it
showed dominant negative activity in HeLa cells. Unlike precise deletion of exon 5 and 7,
precise deletion of exon 3 results in the loss of amino acid sequence that this exon codes for,
but does not introduce a shift in the reading frame. This results in a protein which is
identical to the wt molecule but which lacks 39m amino acids (in frame exon deletion).
However this seems to be in conflict with the findings of Fuqua et al., (1993) who reported
that the variant did not inhibit the wt receptor in yeast cells. Both groups agree that this
variant does not bind DNA in vitro, which given that this exon codes for a zinc finger, is no
surprise. Mikscieck et al, (1993) suggested that protein interactions were involved in the
68
dominant negative effect, and were due either to a direct interaction of the variant with the
normal receptor or due to translational interference (squelching).
2.7.3
‘Negative’ Oestrogen Receptors
ER variants resulting from a disrupting mutation that causes complete or partial loss of
transcriptional activity, and do not exhibit a dominant-negative effect are regarded as
'negative ERs’. This category includes variants with mutations which gives rise to a
decreased affinity for hormone, e.g., those demonstrated by Garcia et al, (1989), Graham et
al., (1990). and Leslie et al (1992) or loss of nuclear localisation, e.g., that described by
Graham et al., (1990) and Leslie et al, (1992). The physiological consequences of such
inactivating mutations in ER would be to reduce the oestrogen responsiveness of cells
harbouring such mutations.
2.7.4
Miscellaneous Variant Oestrogen Receptors
There are many variant ERs which contain point mutations, duplications of exons, exon
deletions whose function are either unknown or have activities which do not fall into the
above categories. These included variants containing mutations which give rise to tamoxifen
resistance, e.g., the variant described by Fuqua et al, (1991, 1992), Castles et al, (1993),
Zhang et al., (1993) and Karnik et al., (1994); or variants which are stimulated by
tamoxifen, e.g. the variant described by Wolf and Jordan (1994a, b) and Catherino and
Jordan (1995). It is interesting to note that in 1 of the 4 tamoxifen resistant MCF-7 cell lines,
the tamoxifen stimulated variant was the major form of ER expressed, and not found in
three others, suggesting that multiple mechanisms may be possible for tamoxifen stimulated
growth.
Most studies of primary breast cancers have related to the development of hormone
resistance, i.e., groups of tumours with data on hormone responsiveness, e.g. Daffada et al.,
(1995), or just small numbers of unselected tumours, but none which have examined small,
node positive carcinomas. This group may yield information regarding the role of the ER in
the earlier stages of the disease. Analysis of such a group could be of great interest in view
of recent work by Fuqua et al., (1996) who described a variant ER cloned from typical
hyperplasia microdissected from frozen breast tissue which contained a point mutation in
exon 4 resulting in an amino acid change in the hormone binding domain. Of particular note
was that adjacent normal epithelial cells also contained the same alteration whereas tissue
69
obtained from healthy adult females without hyperplastic breast disease did not show the
alteration. Functional analysis demonstrated strong agonist properties in MDA-MB-361 and
HeLa cells, with higher levels of activity relative to normal ER, including a 200-fold
increase in oestrogen sensitivity. These results consequently suggest that hyperplastic breast
lesions containing such an ER variant may be responsible for their development, and may
facilitate tumour progression in general.
2.8
Oestrogen Receptor Variant Expression In Normal Tissue
To date only a few studies have been carried out which have examined the presence of
variant ERs in normal mammary gland tissue, thereby limiting an assessment into the
possible involvement of variant messengers in the pathological process of breast cancer.
These studies have demonstrated a similar variety of splice variants in normal breast tissue
to that shown in breast cancer cell lines and primary tumours, the results of which are
summarised in Table 2.4.
It is clear from these studies that some variants are not specific to breast tumours, but appear
to be a frequent occurrence in non-cancerous tissue not only in human breast (Gotteland, et
al., 1995; Pfeffer et al, 1995; Leygue et al, 1996) but also larynx (Marsigliante et al,
1996). In addition, ER variants have also been found in human uterine tissue (Koehorst et
al, 1993; Castles et al, 1995), and also in rat and lizard brain and uterine tissue (Skipper et
al, 1993).
70
Table 2.4 Oestrogen receptor splice variant expression in normal breast tissue
Domain
RNA Change
Protein Change
Reference
C
Out of frame exon 2
deletion
Truncated ER (152 aa)
Gotteland, e ta l, 1995;
Leygue et a l, 1996
Exon 2 and 3 deletion
Unknown
Leygue et a l, 1996
In frame exon 3
deletion
Internal truncated ER (556
aa)
Gotteland, et al, 1995;
Leygue etal., 1996
In flame exon 3 and 4
deletion
Internal truncated ER (44 aa)
Gotteland, e ta l, 1995
In frame exon 4
deletion
Internal truncated ER (483
aa)
Gotteland, et al, 1995
Out of frame exons 3-5
deleted
Out of frame exon 4
and 7
Out of frame exon 5
deletion
23.9 kDER
Pfeffer et al, 1995
39.3 kD ER
Pfeffer et a l, 1995
Truncated ER (371 aa)
Gotteland, et al, 1995;
Pfeffer et a l, 1995;
Leygue et a l, 1996
Pfeffer et a l, 1995
C/E/F
E/F
Out of frame exons 5-7
deleted
Out of frame exon 7
deletion
41.4 kDER
Truncated ER (466 aa)
71
Gotteland, et al, 1995;
Leygue et a l, 1996
2.9
Levels O f Oestrogen Receptor Variant Expression
A number of studies have addressed the issue regarding the relative levels of expression of
splice variants with respect to wt ER in both primary breast cancers, established breast
cancer cell lines and in healthy mammary gland tissue (Gotteland et al, 1995; Castles et al,
1995; Pfeffer et al, 1996)
Pfeffer et a l, (1996) examined the pattern of expression of nine different variants between a
tumour sample and normal breast and obtained nearly the same pattern of expression to that
of the breast cancer cell line MCF-7. Additionally, they also compared the relative levels of
variant expression and found no dramatic variation in the level of expression between the
three samples.
Gotteland et al, (1995) identified a series of six splice variants from primary breast cancers
and normal mammary gland tissue in all RNAs analysed, regardless of origin. In addition,
the relative amount of these different variants in ER positive tumours was comparable to
that measured in ER negative tumours and healthy mammary gland tissue (Table 2.5),
suggesting that tumour progression was not related to the emergence of any ER variants.
In contrast the these studies, Castles et al, (1995) demonstrated that several ER positive cell
lines (MCF-7, ZR-75-1, MDA-MB-134 and MDA-MB-361) as well as some ER negative
cell lines (BT-20, MDA-MB-330, BT-474, T47Dco) were found to contain varying levels of
both wt and exon 5 and 7 variant transcripts. In addition to confirming the co-expression of
variants and wt ER mRNA, they demonstrated that the ratio of expression of the exon 5 and
exon 7 variants to each other as well as to the wt ER differed among the cell lines (Table
2.6). In all ER positive cell lines, except MCF-7, and ER negative cell lines, except BT474, the exon 5 variant was the predominant transcript. The BT-20 cell line expressed the
exon 5 variant at the highest ratio (68%) with wt only accounting for 8% of the detectable
ER transcripts. In addition this was the only cell line which expressed the exon 7 variant
(28%) at a level which was equivalent to or slightly greater than the level of wt ER
transcript.
72
Table 2.5 Estimation of oestrogen receptor splice variant mRNA present in normal and
tumour cells (taken from Gotteland et al, 1995)
Variant
Relative amount of variants3 in :
Normal Tissue0
Primary breast tumours :
ER+/PgR+/pS2+c
ER-/PgR-/pS2-u
Exon 2
ND
ND
ND
Exon 3
0.09
0.06
0.09
Exon 4
0.13
0.12
0.11
Exon 5
0.06
0.05
0.04
Exon 7
0.69
0.55
0.43
Exons 3 + 4
0.11
0.06
0.08
Key ND, not determined (too low to be estimated), a Intensity adjusted to account for
differences in molecular weights over the sum of the intensity of bands specifically
amplified with relevant primers with: b two pools of control RNA, c three ER+/PgR+/pS2+
tumours, d three ER-/PgR-/pS2- tumours.
Table 2.6 Average (%) composition of total oestrogen receptor mRNA from a panel of
breast cancer cell lines (taken from Castles et al, 1995)
Cell line
Hormone status
Wild-type (%)
MCF-7
ER+/PgR+
57
39
4
ZR-75-1
ER+/
41
47
12
MDA-MB-361
ER+/
41
51
7
MDA-MB-134
ER+/PgR-
34
41
25
BT-474
ER-/PgR+
41
44
15
T47Dco
ER-/PgR+
39
47
14
BT-20
ER-/
8
68
24
MDA-MB-330
ER-/
27
58
15
A5 variant (%)
A7 variant (%)
Key A5 denotes the exon 5 deletion variant with A7 denoting the exon 7 deletion variant
73
The cell line MDA-MB-134 expressed the exon 7 variant at a higher level than any of the
other ER positive cell lines. Although this cell line was ER positive by ligand binding, PgR
expression is not induced in response to oestrogen stimulation (Reiner et al, 1986). Given
the high levels of exon 7 variant expression, the authors speculated that elevated levels of
this variant could play a role in the lack of PgR induction by oestradiol, with the low
constitutive PgR expression in these cells being due to gene activation by the constitutively
active exon 5 variant.
Castles et al, (1995) demonstrated that the cell lines MCF-7 and ZR-75-1 had ratios of exon
5 to wt approaching 1:1 which differed from that of 0.5:1 reported by Zhang et al, (1993).
The authors suggested that this difference was possibly due to inter-assay variation,
however, Klotz et al, (1994) reported that different stocks of the MCF-7 cell line expressed
considerably different ratios of the exon 5 variant to wt transcripts, and that the mitogenic
effects of 17p-oestradiol in the various MCF-7 stocks significantly correlated with the
expression ratio of wt to variant ratio. This suggests that the differential expression of
variant to wt ER mRNA was possibly related to oestrogen responsiveness of these cells.
Castles et al, (1995) also demonstrated that the exon 5 variant was expressed at higher
levels than the exon 7 variant in all cell lines tested, suggesting that the two deletions did not
reside on the same mRNA transcript, but may be the result of alternative splicing and the
formation of separate transcripts.
Because of the high sensitivity of the methods employed in demonstrating the co-expression
of so many variants, this raises the question as to whether minor side-products of the
splicing process without any physiological meaning are detected. This situation has been
observed for the transcriptional process, where low levels of transcription of any gene in any
cell occur (Chelly et al, 1989). It is possible that the splicing machinery itself is altered due
to genomic alterations during tumour progression. Pfeffer et al, (1995) examined the
mRNA of genes closely related to the ER (GR, aRAR, yRAR) in MCF-7 cells expressing
ER splice variants, but did not find any splice variants, indicating that variant formation was
specific for the ER and not due to aberrations to splicing mechanisms.
Due to the presence of variant ER RNA’s in primary breast tumour cells, cell lines and
normal cells it has been proposed that the regulation of ER transcripts by alternative splicing
74
may be involved in the control of the expression of the ER gene independently from cellular
transformation (Gotteland et al., 1995; Pfeffer et al, 1996). The fact that some variants
result from the precise deletion of either a single exon (2, 3, 4, 5, or 7) or two exons (3 + 4)
strongly supports this proposition o f the involvement of alternative splicing mechanisms.
The precise deletion of exons from ER-specific mRNA depends neither on the state of the
cell (transformed or untransformed), nor on the hormonal status of the tumour or cell line,
since this mechanism is active not only in the ER+/PgR+ and ER+/PgR- states as well as in
the ER-/PgR- and ER-/PgR+ states of tumours, but is also in normal, healthy tissues.
Differential splicing may lead to a functional message, which is then translated into protein,
as well as exon skipping variants which are not functional. If this differential expression is
regulated then splicing could control the level of functional ER in the cell. Given the
potential biological action of these variants, it is probable that their differential expression
may be an important factor in modulating the overall oestrogen responsive tissues as well as
the expression of oestrogen responsive genes. It is possible that overexpression or selection
of such variants in relation to the wt, due to environmental factors, might be related to the
development of a malignant phenotype or to hormone resistance in some breast tumours.
2.10
Differential Promoter Usage O f The Oestrogen Receptor
Several eukaryotic genes are reported to be transcribed from multiple promoters in a tissuespecific and developmental fashion (Leff et al, 1986; Schliber and Sierra, 1987) adding an
additional level of complexity to the regulation of genes. Members of the steroid hormone,
thyroid hormone, and retinoic acid receptor gene families have also been shown to be
transcribed from multiple promoters including the mouse glucocorticoid receptor (Strahle et
al., 1992), the human and chicken progesterone receptor (Kastner et al., 1990a; Jeltsch et
al., 1990), the Xenopus thyroid hormone receptor p (Shi et al., 1992), and the mouse retinoic
acid receptor a and p (Kastner et al., 1990b; Leroy et al., 1991; Zalent et al., 1991).
In addition several different isoforms of the Xenopus ER have been detected by northern
blotting (Weiler et al., 1987; Baker and Tata, 1990) and as already discussed the human
oestrogen receptor has been shown to be transcribed from two promoters that are separated
by a 2 kb intron. The transcripts initiating from the distal and proximal promoters,
respectively are almost identical in size due to differential splicing, with the only difference
75
between the two being the most 5’ untranslated 120-164 bases, which are unique for each
transcript.
The feature of alternative promoter usage may therefore allow the control of specific subsets
of target genes in different cell types. This idea is supported by the demonstration that the
two chicken and human progesterone receptor isoforms, differentially activate target genes
(Kastner et al., 1990b; Tora et al, 1988; Turcotte et al., 1990).
Analysis of promoter usage in hormonally responsive tissue (Weigel et al., 1995)
demonstrated that activity at P0 was characteristic of endometrium but not readily observed
in normal breast epithelium supporting the possibility that use of P0 might be tissue specific.
Analysis of promoter usage in human breast carcinoma cell lines demonstrated that four of
these, MCF-7, T47-D, BT-20 and MDA-MB-361, utilised the P0 promoter indicating a
change from the tissue specific characteristics of normal breast epithelium. In light of these
results they predicted that an endometrial-specific factor was involved in P0 activity and that
this factor becomes activated in certain breast cancer cell lines.
Functional analysis of PI was recently carried out in a breast carcinoma cell line model and
identified a region of transcriptional activity near the PI site. This region contained two
binding sites for a cellular factor ERF-1 (deConnock et al., 1995) and was found to be
abundantly expressed in both breast and endometrial cell lines.
The two transcripts generated from the two promoters differ only in the 5’ untranslated
regions and therefore code for the same proteins. Therefore these differences may represent
differences in translational efficiency and/or mRNA stability. In addition, it is possible that
the transcription of the two promoters is differentially regulated, as the nucleotide
composition of the two promoters is markedly different (Grandien et al., 1993).
2.11
Methylation O f Oestrogen Receptor CpG Islands
Approximately two thirds of breast cancers express the ER and their growth is stimulated by
oestrogen. For these tumours, therapeutic strategies include oestrogen ablation or
antiestrogens. However, the remaining fraction of primary breast cancers lack detectable ER
protein and are rarely responsive to hormonal treatment (Katzenellebogen, 1991).
Frequently these ER negative tumours lack ER gene expression (Barrett-Lee et al., 1987;
76
Ottaviano et al., 1994), yet this is not due to mutations within the ER gene (Roodi, et al,
1995). Consequently, acquired loss of ER transcription is a potential mechanism for
hormone resistance.
One mechanism that has been proposed to block transcription of the ER gene in ER negative
breast cancers, without structural alterations to the gene, is methylation of cystein-rich areas
termed CpG islands in the 5’ regulatory region of the gene (Bird, 1986). CpG islands are
always unmethylated in normal adult tissue, with the exception of transcriptionally silent
genes on the inactive X chromosome (Bird, 1986) and selected genes which are parentally
imprinted to silence expression of one allele. The importance of CpG island methylation in
silencing such imprinted genes has been recently shown in mutant mice that are deficient in
DNA methyltransferase activity (Li et al., 1993)
DNA methylation at the 5’ position of cytosine residues has been shown to be an important
mechanism which may regulate gene expression (Ceder and Razin, 1990) with an inverse
relationship between methylation and expression demonstrated for many genes, indicating
an important role in the repression of transcription (Doerfler, 1983). A fundamental role for
DNA methylation in neoplastic transformation has been suggested by Goelz, et al., (1985)
who demonstrated that DNA from both benign polyps and malignant carcinomas was
hypomethylated, suggesting that alterations to DNA methylation could be a key event in the
initiation of neoplasia. This has prompted a number of researchers to examine the
methylation state o f the ER gene in both ER positive and negative primary breast tumours,
breast cancer cell lines, and normal mammary epithelium.
Ottaviano et al., (1994) examined the methylation state of the ER CpG island in a panel of
ER positive and negative cell lines. They demonstrated that ER negative cell lines displayed
extensive CpG island methylation and did not express ER mRNA supporting the hypothesis
that loss of ER gene transcription could account for loss of ER expression. Given that
transformed human cells have been shown to have increased DNA methyltransferase
activity, and that this activity has been shown to increase further during the progression of
tumours like colon cancer (Issa et al., 1993), the expression and activity of DNA
methyltransferase mRNA was evaluated. Northern analysis demonstrated a 2-10 fold
increase in DNA methyltransferase mRNA in ER negative cell lines compared with ER
77
positive cell lines and with a corresponding elevation in activity in most of the ER negative
cell lines suggesting an increased overall capacity to methylate DNA .
Additional evidence for a role for DNA methylation of the ER CpG island in suppression of
ER gene expression in ER negative breast cancer cells was demonstrated by Ferguson et al,
(1995). They demonstrated that partial demethylation of the ER CpG island in the negative
cell line MDA-MB-231 by two inhibitors of DNA methylation (5-azacytidine and 5-aza-2’deoxycytidine) corresponded with reexpression of the ER gene as detected by RT-PCR and
Western blot analysis which was functionally active as demonstrated by its ability to
activate transcription of oestrogen-responsive genes.
Lapidus et al, (1996) examined the methylation status of the ER CpG island in both ER
positive and negative primary human breast tumours. The ER CpG island was demonstrated
to be methylated in 25% o f ER negative tumours but remains unmethylated in 100% of ER
positive tumours and normal breast tissue. In addition, the PgR CpG island was
unmethylated in normal breast tissue and PgR positive tumours but was hypermethylated in
40% of PgR negative human breast tumours. Consequently the findings from this study
support the findings in breast cancer cell lines that methylation of the ER gene CpG islands
is associated with lack of ER gene expression in a significant fraction of human breast
tumours.
Analysis of ER CpG island methylation of breast tumours and the surrounding normal
breast by Petrangeli et al, (1995) and found similar results in as much that ER gene
methylation was inversely correlated with protein levels. They also demonstrated that the
ER gene was significantly hypomethylated in breast tumour tissue with respect to the
surrounding normal tissue and suggested that hypomethylation and subsequent protein
overexpression could play a role in neoplastic transformation, leading to enhanced
responsiveness to oestrogen stimuli. The hypermethylation present in surrounding normal
tissue agrees with previous analysis by Petrangeli et al., (1994) who demonstrated that non­
neoplastic breast tissue expressed very low levels of ER, yet was responsive to oestrogenic
stimuli.
A number of reports have also implicated ER CpG island methylation in other malignancies.
Issa et al, (1994) examined the methylation status of the ER CpG island in normal colonic
78
epithelium, benign adenomatous polyps and colorectal carcinomas. They demonstrated
significant methylation in all groups with a progressive age-related CpG island methylation
in normal colonic mucosa, which appears to be one of the earliest molecular changes in the
development of colorectal neoplasia. Issa et al, (1996) also examined the methylation status
of the ER CpG island of normal and neoplastic bone marrow cells and observed that it was
methylated in almost 90% of human leukaemia’s and lymphomas with minimum
methylation in normal bone marrow cells.
Aims O f Chapter
The objectives of this chapter were to screen a large group of mammographically detected,
impalpable, early invasive breast carcinomas for alterations/mutations to the oestrogen
receptor. Initially, carcinomas were to be characterized with respect to steroid receptor
expression and proliferative activity using immunocytochemistry. The presence of altered
oestrogen receptor expressed in these carcinomas was to be determined by isolating RNA
from these cases and amplifying using optimized reverse transcriptase - polymerase chain
reaction. Variant/mutant screening was to be achieved using optimized single stranded
conformational polymorphism and non isotopic RNAse cleavage assay approaches. Cases
which demonstrated altered ER species were to be characterized by sequence analysis and
their expression correlated with pathological data for these carcinomas.
In addition, the methylation status of the oestrogen receptor gene promoter from these
carcinomas, and a selection of ER positive and negative cell lines, was to be investigated
using a novel PCR based approach. The relationship between aberrant methylation and both
pathological data and expression of variants/mutants of the carcinoma group was to be
assessed.
79
Mammographicaly Detected
Breast Carcinoma
J
I
I
RNA Isolation
Reverse Transcription
PCR Amplification
DNA Binding Domain
Proximal Hormone
Binding Domain
Distal Hormone
Binding Domain
Single Stranded Conformational
Analysis
I
Sequence Characterization
Flow D iagram 1 General overview o f the planned approach for ER mutational analysis. R N A was to be
extracted and first strand cD N A produced by reverse transcription, and used as template radioactive PCR at
the ER D N A binding domain and the proximal and distal half o f the hormone binding domain. Radioactive
products were to be screened for mutations by single stranded conformational polymorphism analysis with any
bandshifts to be characterized by sequence analysis.
80
2.12
Methods A n d Materials
2.12.1 Materials
2.12.1.1 Chemicals
All chemicals were 'Analar' grade, purchased from BDH, with a few exceptions: ethidium
bromide, diethyl pyrocarbonate, ammonium persulphate, 3-aminopropylethoxysilane,
bovine serum albumin (PCR grade), magnesium chloride (PCR grade), ammonium sulphate
(PCR grade), TEMED, TRIzol Reagent (Sigma); chloroform, isopropanol, phenol, acetone
and hydrogen peroxide (Fisons); Seakem and Nusieve agarose (ICN); Sequagel (40%
acrylamide: bisacrylamide, 19:1) National Diagnostics; Mutation Detection Enhancement
acrylamide gel matrix from Flowgen.
2.12.1.2 DNAs
Oestrogen receptor cDNA was purchased from American Type Culture Collection; Human
genomic DNA was prepared in house by Jon Naylor, Department of Pathology, University
of Leicester; PCR oligonucleotide primers, and oligo dT primers were synthesized on an
Applied Biosystems DNA Synthesizer; 5'-Biotin modified primer were supplied by Oswell
DNA Services; and random hexamer primers and deoxynucleotide triphosphates were
supplied by Pharmacia; Magnetic beads conjugated to oligo (dT)25 were supplied by Dynal.
2.12.1.3 Radioisotopes
(a,32p) dCTP (800Ci/ml) was supplied by Amersham and ( a ^ S ) dATP (500Ci/ml) was
supplied by ICN Flow and Dupont.
2.12.1.4 Enzymes, Kits and Miscellaneous
Proteinase K, Hhal and Sure Cut Buffer were supplied by Boehringer Mannheim; RNase A
was supplied by Sigma; Taq Polymerases were supplied by Promega and Life Sciences
International, Avian Myeloblastosis Virus Reverse Transcriptase, (AMV) 10 x reverse
transcription buffer, RNase inhibitor and Tth DNA Polymerase Kit, containing 10 x reverse
transcription buffer, and lOmM manganese chloride were supplied by Promega; Large
fragment of DNA polymerase I (Klenow fragment) was supplied by Life Sciences
International and Sequenase Version 2.0 DNA sequencing kit was supplied by United States
81
Biochemicals, Amersham. Bovine Serum Albumin was supplied by Boehringher Mannheim
and Mismatch Detect™ kit was supplied by Ambion.
2.12.1.5 Antibodies And Serum
Normal rabbit serum was obtained from Gibco; Monoclonal Mouse Anti-Estrogen Receptor,
1° antibody, Clone ID5, StreptABC Complex HRP Kit and Biotinylated Affinity Isolated
Rabbit Immunoglobulin antiserum to Mouse Immunoglobulin, 2° antibody were supplied by
DAKO; NCL-PgR Mouse Monoclonal, 1° antibody, was supplied by Novacastra;
Monoclonal Antibody MIB-1 (Ki-67 antigen), 1° antibody, was supplied by DINOVA.
2.12.1.6 Tissue Culture
MCF-7, T47-D oestrogen receptor positive (Keydar et al., 1979; Soule et al, 1973) and
MDA-MB-231 (Cailleau et al, 1974) oestrogen receptor negative cell lines were obtained
from American Type Culture Collection; DMEM (without phenol red), FCS, L-glutamine,
and Phosphate Buffered Saline were all supplied by Gibco Life Sciences; Trypsin and
Dimethylsulfoxide were supplied by Sigma. 75cm^ Culture flasks were supplied by
Nuncon.
2.12.1.7 Electrophoresis
Seakem agarose was supplied by ICN; NuSieve agarose was supplied by FMC.
(j)X1lA/H aelll DNA molecular weight markers were obtained from Promega. Accugel (40%,
19:1 acrylamideibisacrylamide) was supplied by National Diagnostics; 2 x MDE gel matrix
was supplied by AT Biochem. MetaPhore XR agarose and SYBER™ Green II were a kind
gift from David Haynes at Flowgen.
2.12.1.8 Photography and Autoradiography
Ethidium bromide stained gels were visualized under a UVP transilluminator and recorded
using a Polaroid MP-4 Land camera with Polaroid type 667 film or using a Sony
Documentation system 5000 with Sony UPP-11-HD 1 thermal paper. RX medical film from
82
Fuji was used for autoradiography and developed automatically using an RP X-OMAT
Kodak processor.
2.12.1.9 Tissues
For RT-PCR SSCP analysis, 44 fresh frozen, early, invasive breast carcinomas which were
impalpable, and detected by mammography, were studied. All were from the first round of
screening and were detected by the Leicestershire Breast Screening Service. In all instances
a malignant diagnosis had been made before surgery. Tissues were obtained within 30
minutes of surgery and samples 10 x 5 x 3mm frozen using liquid nitrogen and immediately
stored under liquid nitrogen. Parallel slices were fixed in 4% formaldehyde in saline for 18
hours, the processed through paraffin wax. Cases 15mm or less in maximum diameter were
examined. All had either axillary node sampling or axillary dissection. None of the tumours
were from women with a strong family history of breast cancer.
Carcinomas were reported according to the Royal College of Pathologists working party
guidelines (1990). Infiltrating ductal carcinomas were graded using the modified Bloom
and Richardson system (Elston and Ellis, 1991). All histology was undertaken by Dr R.A.
Walker. Clinicopathological features of the population studied are shown in Table 2.7 (data
shown in Appendix I).
Table 2.7 Clinicopathological features of 44 frozen “early”, sporadic, invasive carcinomas
Type
Grade
No. of Cases
No. of Cases
7
Tumour Size
(mm)
< 10
Tub
I
Lob/Tub
I
1
10
4
IDC/ILC
II
1
11
3
IDC
I
10
12
7
IDC
II
18
13
5
IDC
III
7
14
6
15
18
Total
44
1
44
K ey Tub (tubular carcinoma), Lob/Tub (lobular and tubular carcinoma), IDC/ILC infiltrating ductal with
infiltrating lobular carcinoma), IDC (infiltrating ductal carcinoma), numbers in brackets, node positive
83
2.12.1.10 Commonly Used Solutions
Ultrapure Water
Distilled and sterile throughout
3-Aminopropyltriethoxysilane Soln.
1% 3-aminopropyltriethoxysilane in dry
acetone
Mayer's Haematoxylin Soln.
0.1% Haematoxylin, 5% ammonium or
potassium alum, 0.02% sodium iodide,
0.1% acid, 5% chloral hydrate
Eosin Soln.
1% Aqueous water soluble eosin
Citric Acid Buffer
lOmM Citric acid, pH 6.0
TBS
0.05M Tris, 0.15M sodium chloride,
0.1% (w/v) BSA
Diaminobenzidine Soln.
0.5mg/ml Diaminobenzidine, 0.03%
hydrogen peroxide in TBS
Solution D
4M Guanidinium thiocyanate, 25mM
sodium citrate, pH 7.0, 0.5% Sarkosyl
and 0.1M 2-p-mercaptoethanol
Agarose Gel Loading Buffer
0.2% Bromophenol blue, 0.2% xylene
cyanol FF, 0.04M Tris-acetate, 50%
glycerol, 0.02M EDTA, pH7.5
Denaturing Gel Loading Buffer
95% Formamide, 0.05% (w/v) xylene
cyanol FF, 0.05% (w/v) bromophenol
blue
SSCP Gel Loading Buffer
95% Formamide, 0.05% (w/v) xylene
cyanol FF, 0.05% (w/v) bromophenol
blue, lOmM EDTA, 0.1% SDS
MDE SSCP Gel loading buffer
95% Formamide, lOmM sodium
hydroxide, 0.25% bromophenol blue,
0.25% xylene cyanol FF
84
Non-isotopic SSCP denaturing
0.5M Sodium hydroxide, lOmM EDTA
solution
Non-isotopic SSCP loading buffer
100% Formamide, 0.5% bromophenol
blue, 0.5% xylene cyanol FF
Metaphor XR Agarose SSCP gel
95% Formamide, 20mM EDTA, 0.05%
Loading Buffer
bromophenol blue, 0.05% xylene cyanol FF
Ethidium Bromide Soln.
1Omg/ml Ethidium bromide
Standard PCR Buffer
45mM Tris-HCl, pH 8.8, 1 ImM
Ammonium sulphate, 45mM
magnesium chloride, 6.7mM 2-fmercaptoethanol, 113pg/ml BSA,
200pM each dNTP, 4.4pM EDTA
35S PCR Buffer
45mM Tris-HCl, pH 8.8, 1 ImM
ammonium sulphate, 4.5mM
magnesium chloride, 6.7mM 2-(3mercaptoethanol, 113pg BSA, 200 pM
dCTP, dTTP, dGTP, 25mM dATP, 4.4pM
EDTA
10 x TBE Buffer
108g/l Tris, 55g/l boric acid, 5mM
EDTA, pH 8.0
10 x TAE Buffer
0.4M Tris-acetate, 0.2M EDTA (pH
7.5)
Denaturing Soln.
1.5M Sodium chloride, 0.5M sodium
hydroxide
Neutralizing Soln.
1.5M Sodium hydroxide, 1M Tris
20 x SSC
3M Sodium chloride, 0.3M tri-sodium
citrate
Church Buffer
0.5M Phosphate, pH 7.2, ImM EDTA,
pH 8.0, 7% SDS
85
Oligolabelling Buffer
50mM Tris-Cl, pH 8.0, 5mM chloride,
ImM p-mercaptoethanol, each dATP,
dGTP, dTTP, 0.2M HEPES, pH 6.6,
250 mg/ml random hexamers
2 x Binding and Washing Buffer
lOmM Tris-HCl, pH 7.5, 1.0 mM
(2 x B&W)
EDTA, 2.0M sodium chloride, 0.1%
Tween-20
2 x (mRNA) Binding and Washing
20mM Tris-HCl, pH 7.5, 1.0M LiCl,
Buffer (2 x B&W)
2mM EDTA
86
2.12.2 Methods
2.12.2.1 Silane Coating O f Microscope Slides
Silane coating of microscope slides was carried out to increase adherence of sections to
glass surfaces and to decrease non-specific background staining. Silane coating of
microscope slides was achieved using the method described by van Prooijen-Knegt et al.,
(1982) in which detergent cleaned microscope slides were washed in 96% IMS, allowed to
air
dry
and
coated
by
immersion
in
a
freshly
prepared
solution
of
3-
aminopropyltriethoxysilane for 5 seconds. Slides were then rinsed briefly in dry acetone,
then in ultrapure water and allowed to air dry overnight at 42°C.
2.12.2.2 Haematoxylin A nd Eosin Staining O f Tissue Sections
Dewaxed and dehydrated sections were immersed in Mayer's haematoxylin solution for
approximately 5 seconds, rinsed in running tap water, briefly immersed in eosin solution
and rinsed in running tap water. Stained sections were then dehydrated, transferred to xylene
and mounted in resinous mountant (XAM, BDH).
2.12.2.3 lmmunocytochemistry
2.12.2.3.1 Antigen Unmasking O f Formalin-Fixed Paraffin Embedded Breast Tissue
Sections
The method used for antigen retrieval was based on those described by Shi et al., (1991) and
Cattoretti et al., (1993). In brief, dewaxed, dehydrated sections were placed in a microwave
proof coplin jar and filled with lOmM citric acid buffer. The jar was then irradiated in a 700800W microwave oven for 5 minutes at high power, with occasional re-filling to ensure the
buffer level was above the sections. The procedure was then repeated for ER and PgR
staining and twice more for MIB-1. Sections were allowed to cool for a minimum of 15
minutes prior to processing for lmmunocytochemistry.
2.12.2.3.2 Immunocytochemical Staining For ER, PgR and MIB-1
Avidin-Biotin Complex peroxidase immunocytochemistry was carried out based on the
method described by Hsu et al., (1981). Non-specific staining was blocked by covering
sections with lOOpl of 20% normal rabbit serum diluted in TBS for 10 minutes in a
humidity chamber. Slides were drained, excess serum removed from around the section, and
100ml of a 1:100 dilution of the respective primary antibody applied to each section and
87
allowed to incubate either at room temperature for 2 hours or overnight at 4°C. Primary
antibody was washed off with TBS and lOOpl of a 1:400 dilution of biotinylated secondary
antibody (biotinylated affinity isolated rabbit anti-mouse antiserum immunoglobulin to
mouse immunoglobulin) applied to each section and incubated at room temperature for 30
minutes. At this point an Avidin-Biotin complex was prepared using a StreptABC Complex
HRP Kit by the mixing of 1ml TBS, lp l Streptavidin and lpl biotinylated Horseradish
Peroxidase and allowed to stand for 30 minutes. Secondary antibody was washed off with
TBS, lOOpl of ABC solution applied to each section and allowed to incubate at room
temperature for 30 minutes. Excess ABC solution was then washed off sections with TBS
and 100ml of diaminobenzidine solution applied to each section and allowed to incubate for
approximately 15 minutes until colour formation when viewed under a microscope, at which
point sections were washed in TBS, running tap water and counterstained with Mayers
Haematoxylin. Sections were then dehydrated to xylene and mounted in resinous mountant
(XAM).
2.12.2.3.3 Assessment O f ER And PgR lmmunocytochemistry
Assessment of ER and PgR immunocytochemistry was achieved using the H Score system.
Approximately 1000 cells per tumour were counted and the percentage of tumour cells
showing no (0), weak (1), moderate (2), and strong (3) positivity was estimated. The H
Score was then calculated as follows:
H Score = (1 x percentage o f weak cells) + (2 x percentage o f moderate cells)
+ (3 x percentage strong positive cells)
A range of H Score value o f 0 - 300 therefore results. Tumours were scored as positive if the
calculated value fell within the range 50 - 300.
2.12.2.3.4 Assessment O f MIB-1 Immunocytochemistry
Assessment of MIB-1 immunocytochemistry was achieved by counting at approximately
1000 cells within an individual tumour section and calculating the percentage of cells which
demonstrated nuclear staining.
88
2.12.2.4 Cell Culture
2.12.2.4.1 Maintenance O f Breast Cancer Cell Lines
Oestrogen receptor positive T47-D and MCF-7 and ER negative MDA-MB-231 cell lines
*■>
were maintained in 75cm J culture flasks in DMEM without phenol red, 10% FCS and 1%
L-glutamine, at 37°C in a 5% CO 2 , humidified atmosphere.
2.12.2.4.2 Passage
Culture medium was aspirated, cell monolayers were washed with phosphate buffered saline
and cells removed by treating monolayers with trypsin. The resultant suspension was
centrifuged and the pellet resuspended in approximately 6mls of prewarmed, fresh culture
media. 2mls o f this cell suspension was then used to seed 3 new flasks containing lOmls
fresh, prewarmed culture media. Flasks were then maintained as described above.
2.12.2.4.3 Freezing Down
Culture medium was aspirated, cell monolayer washed with phosphate buffered saline and
cells removed by treating monolayer with trypsin. The resultant suspension was centrifuged
and the resultant pellet resuspended in 0.5 ml FCS, 0.5 ml freezing solution (20%
dimethylsulfoxide, 80% DMEM), placed at -70°C for 24 hours and then after stored
indefinitely under liquid nitrogen.
2.12.2.5 RNA Isolation
2.12.2.5.1 Cell Lines
Two methods were used to extract RNA. In the first, RNA was extracted according to the
method described by Chomczynski and Sacchi (1987). In brief, 1ml of Solution D was
added to each cell pellet containing approximately 10^ cells. The following solutions were
then sequentially added: 0.1ml of 2M sodium acetate, 1ml of water saturated phenol, and
0.2ml chloroform : isoamyl alcohol mixture (24:1) with mixing by inversion after each
addition. The final suspension was then shaken vigorously, left on ice for 15 minutes and
centrifuged at 4°C for 20 minutes. The aqueous phase was removed and RNA precipitated
in isopropanol for one hour at -20°C. RNA was sedimented, and resuspended in Solution D,
followed by re-precipitation in isopropanol at -20°C for one hour. The pellet was then
washed with 75% ethanol and dissolved in DEPC treated water. Efficiency of extraction and
89
integrity o f RNA was assessed by fractionation of RNA total extract on a 1% Seakem
agarose, and by spectrophotometry at 260 and 280nm.
The second method for RNA extracted utilized TRIzolTM reagent (Chomzynski, 1993) and
was carried out to the manufacturer's protocol. In brief, approximately 0.5 x 106 T47-D and
MCF-7 monolayer cell cultures were directly lysed in culture flasks by the addition of
1.5mls o f reagent, passing the resultant cell lysate through a 10ml pipette and allowing to
stand for 5 minutes at room temperature to allow complete dissociation of nucleoprotein
complexes. The lysate was mixed with 0.3mls of chloroform, shaken vigorously by hand for
15 seconds, incubated at room temperature for 2 minutes followed by centrifugation at
12,000 x g for 15 minutes at 4°C. The aqueous layer was removed and RNA precipitated by
the addition of 0.75mls o f isopropanol. Samples were incubated at room temperature for 10
minutes then RNA sedimented at 12,000 x g for 10 minutes at 4°C. The supernatant was
removed, RNA washed with 1ml 75% ethanol, sedimented at 7,500 x g at 4°C, air dried
and resuspended in lOOpl DEPC treated water with heating to 55°C to 60°C for 10 minutes.
Optical densities were determined spectrophotometrically at 260nm and 280nm to determine
quality and quantity of RNA isolates.
2.12.2.5.2 Frozen Tissue
The optimum volume o f reagent and the number of sections required to result in a good
yield of RNA were assessed. 1, 2 and 3 serial sections were cut on a cryostat and placed into
lm l, 0.4ml and 0.25ml of reagent. Samples were placed on ice for 20-30 minutes during
which time the tissue was passed through 15G, 19G and 21G needles to aid homogenization.
The protocol was then identical to that described in (2.12.2.5.1) but using correspondingly
reduced amounts of reagents, i.e., 0.2ml chloroform, 0.3ml isopropanol, 0.5ml 75% ethanol
and resuspending RNA in 20pl RNase-free water.
2.12.2.5.3 Oligo d(T)25 mRNA Isolation From Total RNA
The volume total RNA isolated from frozen tumours was adjusted to lOOpl with ultrapure
water, heated to 65°C for 2 minutes to disrupt any secondary structure and added to 100pl
oligo dT dynabeads (20pl stock oligo (dT)25 dynabeads prewashed and resuspended in lOOpl
2 x B&W (mRNA)), gently mixed and allowed to hybridize at room temperature for 5
minutes. Oligo dT dynabead/mRNA hybrids were isolated from the supernatant using a
90
Dynal Magnetic Particle Concentrator (MPC), washed twice with 200pl of 1 x B&W
solution, and 2 x washes in 20pl of 5 x reverse transcription buffer with resuspension in 5 pi
of 5 x reverse transcription buffer. Oligo dT/mRNA/5 x reverse transcription buffer
solutions were then used directly as template for Avian Myeloblastosis reverse transcription.
2.12.2.6 Reverse Transcription
2.12.2.6.1 Avian Myeloblastosis Virus Reverse Transcription
2.12.2.6.1.1 Cell Line RNA
RNA was used as template for reverse transcriptions for 1 hour at 42°C in the following
reaction mixtures and overlaid with light paraffin mineral oil (reaction volume 25pl): 5-8
units Avian myeloblastosis virus reverse transcriptase, reverse transcriptase buffer, 20 units
RNase inhibitor, 5pM dNTP's and lOmM dithiothreitol using lOOpmol or 25pmol random
hexamer; 0.1 mg oligo dT primer (T)25, lOpmol oestrogen receptor (ER Set 2R see Table 2.7)
or actin (5’-CTA GAA GCA ATT TGC GGT GGC AG-3’) downstream primers or already
primed oligo dT/mRNA (Section 2.12.2.5.3).
Initially, RNA, reverse transcription primers and RNase free water were mixed and heated
to 70°C for 5 minutes then allowed to cool to room temperature. The remaining reaction
components were then added. Once terminated, 2.5pi was used as template for PCR with the
remainder stored at -20°C.
In order to overcome subsequent problems of aerosol contamination additional precautions
were also taken. These included the use of dedicated, alcohol cleaned, UV irradiated pipettes
for reverse transcriptions using filter pipette tips, with all manipulations carried out in a
class II safety cabinet.
2.12.2.6.1.2 Frozen Tissue RNA
To optimize the method, RNA was isolated from sections taken from a large sample of
asymptomatic breast carcinoma (A) approximately 1cm x 3-4mm in size (as in section
2.12.2.5.2), and yield compared with RNA isolation from sections from a smaller sample of
asymptomatic breast carcinoma (B) approximately 3-4mm x 3-4mm in size. lOpl of total
RNA isolated from 1, 2, and 3 frozen tissue sections RNA isolations from the small and
large pieces of breast carcinomas and used as template for reverse transcription using
91
lOOpmol o f random hexamer primers in the reaction described in section (2.12.2.6.1.1).
Once terminated, 2.5pi was used as template for PCR with the remainder stored at -20°C.
2.12.2.6.2 Thermus thermophillus DNA Polymerase Reverse Transcription
Total RNA isolated from MCF-7 cell line was used as template for reverse transcription
(Ruttimann et al., 1985 and Myers et al, 1991) using a Tth reverse transcription kit
(Promega). In brief, lp g total RNA was reverse transcribed in the following reaction
mixture: reverse transcription buffer, ImM manganese chloride, and 0.4pM each dNTP.
Reaction mixture was denatured at 65°C for 10 minutes then placed on ice. 15 and 50pmol
oestrogen receptor (ER Set 2R) and actin downstream primer, 4-6 units Tth DNA
polymerase, and ultrapure water to give total reaction volume of 20ml was added and placed
at 50°C for 5 minutes then allowed to ramp to 70°C. Once at 70°C, reverse transcription
was allowed to proceed for 20 minutes. Reactions were then placed on ice and used as
template for PCR amplification.
2.12.2.7Polymerase Chain Reaction
2.12.2.7.1 Design O f Oestrogen Receptor PCR Primers
Oestrogen receptor PCR primer sequences ER set 2, ER Set 3, and ER Set 5 were designed
to amplify the DNA binding domain and proximal region of the hormone binding domain
respectively. This was achieved using the oestrogen receptor cDNA sequence described by
Ponglikitmorkol et al., (1988) and the publication by Green et al., (1986) which described
functional domains within the cDNA. Primers were selected using the primer program on
genome data base (GDB) with the following criteria: i) primers were approximately 20mers;
ii) approximately 50% G/C rich; iii) each primer being directed to separate exons; iv) each
primer flanked a region o f interest ER Set 4 primers used were those described by Fuqua et
al., (1990) and flanked the distal region of the DNA binding domain. Confirmation of
sequence specificity was obtained by running the sequences through GDB and EMBL.
Mismatch Detect™ analysis primers used were identical to ER Set 2 with T7 and SP6
recognition sequences 5’ to the primer sequence. Primers were also designed to amplify a
region encompassing a dinucleotide repeat microsatellite sequence and a H hal methylation
sensitive restriction site mapped within the 5’ promoter region of the ER gene using the ER
5’ sequence described by Keaveney et al. (1992). Oestrogen receptor primer sequences
selected are listed in Table 2.8 with position of annealing sites described in Appendix II.
92
Table 2.8 Oestrogen receptor PCR primer sequences
Key F denotes forward primer (sense); R denotes reverse primer (antisense); Underlined indicates T7 (Forward) and SP6 (Reverse) recognition sequences;
Numbers in brackets denote nucleotide number with reference to the sequence described by Green et al, 1986 (Appendix II). * Numbered brackets
denote nucleotide number with reference to the sequence described by Keaveney et al., 1992 (Appendix II).
Target
Domain
B Region
ER Set
Prim er Sequence
ER Target
ER Set 1
F 5’-CAG GTG CCC TAC TAC CTG GA-3’
(739)
R 5’- CGA GTC TCC TTG GCA GAT TC-3’
(908)
F 5’-TTC AGA TAA TCG ACG CCA GG-3’
(819)
R 5’-GGC GCT TGT GTT TCA ACA TT-3’
(1168)
F 5’-ACA AGC GCC AGA GAG ATG AT-3’
(1160)
R 5’-TCG AAG CTT CAC TGA AGG GT-3’
(1384)
F 5’-GGA GAC ATG AGA GCT GCC AAC-3’
(1210)
R 5’-CCA GCA GCA TGT CGA AGA TC-3’
(1648)
F 5’-GAT CTT CGA CAT GCT GCT GG-3’
(1629)
R 5’-AAG TGG CTT TGG TCC GTC TCC -3’
(2063)
ER Set 6
F 5’-TAA TAC GAC TCA CTA TAG GGT TCA GAT AAT CGA CGC CAG G-3’
(819)
(T7/SP6 ER Set 2)
R 5'-ATT TAG GTG ACA CTA TAG GAG GCG CTT GTG TTT CAA CAT T-3’
(1168)
ER Set 7
F 5’-TAA TAC GAC TCA CTA TAG GGG GAG ACA TGA GAG CTG CCA AC-3’
(1210)
(T7/SP6)
R 5’-ATT TAG GTG ACA CTA TAG GAA AGT GGC TTT GGT CCG TCT CC-3’
(2063)
ER 2092
5’-AAT GCG ATG AAG TAG AGC CC-3’
(2092)
F Region
ER Meth 1 +2
F 5’-GAC GCA TGA TAT ACT TCA CC-3’
(1580)*
5’ UTR
R 5’ GAG ATT CTC TCA GCC TGA C-3’
(2091)*
ER Set 2
ER Set 3
ER Set 4
ER Set 5
DNA Binding
Hormone
Binding
Hormone
Binding
Hormone
Binding
DNA Binding
Hormone
Binding
2.12.2.7.2 Optimization O f Oestrogen Receptor PCR
For PCR optimization lng of ER cDNA was used as template with PCR carried out based
on the method described by Saiki et al., (1988), in standard PCR buffers. The concentration
of each primer were typically lOpmol with the final volume made up to 50pl with PCR
grade water. 0.5 Units Taq Polymerase per reaction was added, with primers, after an initial
denaturation step (hot start PCR). PCR cycling was performed as follows for 30-35 cycles
using either a Hybaid Omnigene or Perkin-Elmer Cetus (Version 2.2) DNA Thermal
Cycler:
Initial denaturation
94°C
Primer Annealing
(E A T )lm in
Primer Annealing
(EAT) lmin
Primer Extension
72°C
lmin
Denaturation
94°C
lmin
Final Primer Extension
72°C 7 min 1 Cycle
lOmin
1 Cycle;
30-35 Cycles;
In all cases, the estimated annealing temperature (EAT) was calculated using the following
formula:
(EAT) = 4(G+C) + 2(T+A) - 5 °C
PCR products were fractionated through 3% NuSieve / 1% Seakem agarose gels in 1 x TAE
running buffer at a voltage o f 60-80V for 4-6 hours. PCR product was first added to 5 x gel
loading buffer prior to loading. Gels were stained with ethidium bromide (0.5mg/ml), and
visualized on a UVP transilluminator. If the estimated annealing temperature did not result
in specific products when run on agarose gels, the annealing temperature was increased until
specific products were obtained.
2.12.2.7.3 Optimization O f Nested PCR
First round PCR was carried out using external primers 739/1384 which flank the DNA
binding domain o f the ER gene as described in section (2.12.2.7.1). Cycling was carried out
94
as previously described in section (2.13.2.6.2) with ER cDNA and reverse transcriptase
product from an asymptomatic breast tumour as template. Second round PCR was carried
out using 1jj.1 of: neat, 1/10, 1/100, 1/1000 dilution of first round product as template with
ER Set 6 PCR primers.
2.12.2.7.4 Agarose Gel Analysis O f PCR Products
Products fractionated on 2% Seakem Agarose gels in 1 x TAE at 4-5 V/cm. Prior to loading,
products were added to 5 x agarose gel loading buffer. Gels were then stained with ethidium
bromide (0.5pg/ml) and visualized on a UVP transilluminator.
2.12.2.7.5 PCR Amplification O f Reverse Transcription Products
First strand cDNA (5 and lOpl of AMV RT product), 20pl Tth RT-product, was directly
used as template for 35 cycles o f PCR amplification. PCR was carried out in standard PCR
buffer using typically lOpmol o f primers. Final volumes were made up to 50pl with PCR
grade water overlaid with mineral oil and 0.5 Units Taq Polymerase per reaction was added,
with primers, after an initial denaturation step. PCR cycling and agarose gel analysis of
PCR products was performed as already described.
2.12.2.7.6 Labelled PCR
50ng o f high molecular weight DNA, 2.5pl first strand cDNA was used directly as template
for 35 cycles o f PCR amplification. PCR was carried out in 35S PCR buffer with 1 unit Taq
polymerase and 0.3pi 35S a-dA TP (500Ci/ml) per reaction, with the concentration of each
primer typically lOpmol and a final volume of 25pl with PCR grade water. Primer
sequences, cycling conditions and controls were identical to those described in section
(2.16.10.1 -2.16.10.3).
2.12.2.7.7 PCR Methylation Analysis O f Oestrogen Receptor CpG Islands
The presence of a polymorphic dinucleotide microsatellite repeat sequence and a
methylation sensitive restriction site (.Hhal) within close proximity of each other in the 5’
UTR of the ER was taken advantage of to assess the methylation status of a group of early,
invasive, asymptomatic breast tumours. Approximately lpg of high molecular weight DNA
from the breast cancer cell lines T47-D and MDA-MB-231 (methylation positive and
negative respectively) or lOpl of asymptomatic tumour DNA were digested with 25 units of
95
Hha I in 1 X Sure Cut buffer, 0.05 mM spermidine, 50ng/ml BSA, in a final volume of
25pl, and incubated at 37°C for approximately 16 to 18 hours followed by heat inactivation
of the enzyme at 99°C. A no digest control was performed identical to the digest already
described but omitting Hha I. 2.5pi and 5pi of
+/- Hha I digests for cell line and
asymptomatic tumours respectively were used as template for 35 amplification cycles using
a semi-multiplex 35S PCR with ESR 1 + 2 primers which flank the microsatellite repeat and
ESR 1 and ER Meth 2 which flank the microsatellite repeat and the Hha I restriction site.
Cycling conditions used were as described in Appendix II.
2.12.2.7.8 Polyacrylamide Gel Electrophoresis Analysis Of Labelled PCR Products
35S PCR products were fractionated on 5% denaturing polyacrylamide gels prepared using
the Sequagel™ Sequencing system according to the manufacturers instructions, in 1 x TBE
at 50°C for 2-4 hours. The PCR products were first added to denaturing gel loading buffer
prior to loading, denatured at 94°C for 2.5 minutes and quenched on ice. The gel was then
dried and then exposed to blue sensitive X-ray film at room temperature for 1-5 days.
2.12.2.7.9 Southern Analysis OF RT-PCR Products
Southern analysis of RT-PCR generated DNA fragments was carried out based on the
method described by Southern (1975).
50pl o f RT-PCR product were fractionated on 2% Seakem agarose gels in 1 x TAE at 6070V for 2-3 hours, stained with ethidium bromide (0.5mg/ml), and photographed. Gels were
pre-treated prior to transfer in denaturing solution and then neutralizing solution for 1 hour
at room temperature with agitation. After rinsing in ultrapure water, DNA was transferred to
a nylon membrane (Hybond-N, Amersham) overnight by capillary transfer in 20 x SSC
buffer. DNA was then immobilized by baking the filter at 80°C for 2 hours. Efficiency of
transfer was assessed by re-staining the gel with ethidium bromide (5mg/ml) with
visualization under UV transillumination.
Approximately 50ng PCR generated cDNA fragment was radiolabelled by random
hexadeoxyribonucleotide priming (Feinberg and Vogelstein, 1983) with cc32P-dCTP using
Klenow fragment of DNA Polymerase I. After initial denaturing at 98°C for 10 minutes the
DNA was labeled in the following reaction: 10 units Klenow Fragment; 3 pi a 32P-dCTP
96
(800Ci/ml); 2.2 mg/ml (BSA) in 1 x oligonucleotide-labelling buffer in a final volume of
50pl.. The reaction was allowed to proceed at 37°C for 3 hours or at room temperature
overnight.
Radiolabelled probe was then separated from unincorporated dNTP's by centrifugation
through a calibrated 1ml Sephadex G-50 (Pharmacia) column prepared as described by
Maniatis et al, (1989). Percentage incorporation of radionucleotide into the probe was
calculated by increasing the volume of labelling reaction to 1OOpl with ultrapure water and
counting on an Oncor Incorporated probe counter before and after the centrifugation through
the column. The following formula was used to calculate the percentage incorporation :
% Incorporation = (Count after column / Count before column) x 100
Nylon filters were pre-wetted in 6 x SSC then placed into orbital hybridization chambers
(Hybaid) between two nylon meshes. Filters were then prehybridized in 20ml of Church
Buffer (Church and Gilbert, 1984), with 400pl denatured salmon sperm DNA (lOmg/ml) at
65°C with orbital action in a Hybaid hybridization oven for approximately 1 hour. 400pl of
denatured salmon sperm DNA was added to radio labeled probes and denatured at 98°C
before being introduced into the hybridization solution. Hybridization’s were allowed to
proceed overnight at 65°C. Post-hybridization washes were carried out at 65°C with 2 x
SSC / 0.1% SDS for 2 x 10 minutes and 1 x SSC / 0.1% SDS for 2 x 10 minutes. Filters
were then autoradiographed at -70°C for 1-5 days using Fuji X-ray film with intensifying
screens.
2.12.2.8 Single Stranded Conformational Polymorphism Analysis (SSCP)
2.12.2.8.1 Isotopic SSCP
SSCP analysis was carried out based on the method described by Orita et al., (1989a,b). 35S
PCR products were generated as described (2.12.2.7.6) using genomic DNAs, which were
tumour/normal pairs from the same individuals prepared by Gilbert Evans, Department of
Pathology, University o f Leicester, with ER Set 3 primers, and RT-PCR products generated
using all other primer sets.
97
Isotopic SSCP was carried out using two different methods. In the first method, 35S PCR
products were fractionated on 5-6% non-denaturing polyacrylamide gels prepared using
Accugel (40% acrylamide, National Diagnostics) in 1 x TBE, 0.1% SDS, 0.08 %
ammonium persulphate and 40pl TEMED, made to a final volume of 100ml in ultrapure
water, at 30V for 2-4 hours with cooling. Prior to loading, PCR products were first added to
SSCP gel loading buffer, denatured at 94°C for 2.5 minutes and quenched on ice for at least
2.5 minutes. Gels were then dried and then exposed to blue sensitive X-ray film without
intensifying screens for 1-5 days.
The second method utilized a unique polyacrylamide-derived matrix, Mutation Detection
Enhancement gel (MDE), which significantly improves resolution of conformationally
different DNA molecules. SSCP analysis using MDE differed little from the previous
method using conventional polyacrylamide. Labeled products were fractionated on 0.5 - 0.7
non-denaturing MDE gels in 0.6 x TBE at 200-400V for 15-18 hours. Prior to loading PCR
products were mixed 2:1 with the MDE gel loading buffer and denatured at 94°C for 2.5
minutes and quenched on ice for 2.5 minutes to allow adequate secondary structure to form.
Once sufficiently fractionated the gel was dried at 80°C and exposed to x-ray film for 1-5
days. p53 point mutant/wt cDNA PCR products (supplied in Mismatch Detect™) Kit were
co-fractionated as SSCP positive controls.
2.12.2.8.2 Non Isotopic SSCP
A number o f methods were attempted to carry out non-isotopic SSCP. The first was based
on the method of Yap and McGee, (1992). In brief, lOOng of high molecular weight DNA
was used as template for lOpl to 25 j l x 1 PCR using ER set 3, and (A508) primers, cycling
conditions given in Appendix II. 5pl-15pl of chloroform extracted PCR product was loaded
onto 10%, 8%, and 6% non-denaturing polyacrylamide gels (Accugel, 0.4mm thick) and
fractionated at 20-30V/cm in 0.5 x TBE for 2-3 hours. Prior to loading, l-2pl denaturing
solution was added to sample and heated to 80°C for 3-5 minutes, quenched on ice for
approximately 2 minutes and l-2pl loading buffer added. Once sufficiently fractionated,
gels were then stained with ethidium bromide (0.5mg/ml) and visualized under UV
illumination.
98
An alternative method for non-isotopic SSCP utilized MetaPhor XR agarose (Flowgen). 4%
horizontal and vertical gels (0.3mm and 1mm thick respectively), with and without 5%
glycerol,
were
prepared
in
1 x
TBE
buffer according
to the
manufacturer’s
recommendations. Prior to loading, loading buffer was added to sample in a 1:1 ratio and
heat denatured for 5 minutes at 95°C, quenched on ice, and loaded immediately. Samples
were then fractionated in 1 x TBE with and without glycerol at 5V/cm for 3-4 hours then
stained with either ethidium bromide (1 mg/ml) for 20 minutes or SYBR Green II dye
(1:10,00 dilution) for 40 minutes and visualized under UV illumination.
The third method utilized the MDE system and was identical to that described in section
(2.12.2.8.1) for isotopic SSCP usage. Once samples had fractionated gels were stained in
5mg/ml ethidium bromide and visualized under UV illumination. Again, p53 point
mutant/wf cDNA PCR products (supplied in Mismatch Detect™) Kit were co-ffactionated as
SSCP positive controls.
2.12.2.9 Mismatch Detect II™ Analysis - Non Isotopic RNAse Cleavage Assay (NIRCA)
This mutational screening method is based upon that described by Myers et al, (1985) and
Winter et al., 1985. The method relies on the ability of RNase A to cleave single unpaired
bases (mismatches) in an RNA probe hybridized to an experimental target containing a
mutation. The method involves amplification of a target region and wild type control by
PCR, incorporating opposable T7 and SP6 phage promoters into the product. These products
are used as templates for in vitro transcription of sense and antisense strands of the
experimental and wild type PCR products. Hybridization of sense and antisense
experimental RNA transcripts with the complementary wt transcripts produces RNA duplex
targets for RNase digestion, the products of which are analyzed on agarose gels.
2.12.2.9.1 PCR And Nested PCR
25pl first round PCR carried out using external primers (739 and 1384), and (1160 and
2092) flanking the DNA binding and hormone binding domains of the ER gene respectively,
using cycling conditions described in Appendix II. Template for the first round PCR
included 2.5 pi AMV reverse transcription product, lng ER cDNA, and p53 wild type and
mutant control plasmids. The template for the nested reaction consisted o f 1pi o f a 1 in 100
dilution of the corresponding first round PCR. ER Set 6 and ER Set 7 which contain T7 and
SP6 phage consensus sequences at the 5’ ends and flank the DNA binding and hormone
99
binding domains respectively, were used in the 50pl nested PCR reaction with cycling
conditions described in Appendix II.
2.12.2.9.2 Production O f Sense And Antisense RNA Probes
RNA probes were transcribed for both experimental and wt templates in the following
reaction: 2pl nested PCR product, 1 x Transcription buffer, 0.5pM rNTP mix, 20 units T7
polymerase (for sense RNA probes production) or 20 units SP6 RNA polymerase (for
antisense RNA probe production), ultrapure water to lOpl, and incubated at 37°C for 1 hour.
An equal volume of Mismatch Hybridization Buffer added to each reaction, and heated to
95°C for 3 minutes to inactivated.
2.12.2.9.3 Hybridization O f Experimental And Control Transcripts
Equal volumes of: w//p53 wt T7 transcript and experimental/p53 mutant SP6 transcripts; wt/
p53 wt SP6 transcript and experimental/p53 mutant T7 transcript; wt/p53 wt T7 transcript
and wt/p53 wt SP6 transcript (no-mismatch control duplex) mixed, heated to 95°C for 3
minutes and allowed to cool to room temperature.
2.12.2.9.4 RNAse Treatment of Experimental -Wild Type Hybrids
4pl of each hybridized sample digested with 16pl of RNase solution at 37°C for 45 minutes.
Reaction terminated by the addition of 4.5pi Mismatch Gel Loading Solution. Choice of and
optimum concentration o f RNases determined empirically.
2.12.2.9.5 Agarose Gel Analysis O f RNAse Cleavage Products
Products fractionated on 2% horizontal agarose gels in 1 x TBE, and ran at ~5 V/cm until
cleavage products had separated sufficiently when analyzed using a UVP transillumination.
2.12.2.10 Recovery O f Variant Oestrogen Receptor Species From Gels
Aberrant ER PCR products showing either gross molecular weight alterations (splice
variants) identified on agarose gels or exhibiting mobility shifts on SSCP gels were excised
for sequence analysis.
Recovery o f variant ER species showing gross molecular weight differences identified on
agarose gels was achieved using the method described by Zhen and Swank, (1993). In brief,
species of interest were identified on agarose gel by ethidium bromide staining and UV
illumination. Using a scalpel blade a trough was cut directly in front of the leading edge of
the band o f interest and the gel slice removed. The trough was filled with 15% Polyethylene
glycol (PEG)/TAE and electrophoresis continued at 20 V/cm for 2-3 minutes. The
PEG/TAE/DNA was removed and subjected to conventional phenol and chloroform
extraction followed by 1/10 volume 3M sodium acetate with 2 volumes of ethanol. After
centrifugation DNA was washed with 70% ethanol and resuspended in 40pl ultrapure water.
I jliI, 5pi volumes o f the isolate was then used then used as template for PCR amplification
using primer used in the original amplification.
Recovery o f variant ER species exhibiting mobility shifts on SSCP was achieved using the
method described by Berx et al., (1995a). In brief, using a scalpel, the area containing the
band of interest was excised from polyacrylamide gel and placed into 150pl ultrapure water
and incubated at 60°C for 30 minutes then placed immediately at -70°C for 15 minutes.
Cycles o f heating and freezing was repeated 3 times and lOpl used as template for PCR
amplification using primers used in the original amplification.
2.12.2.11 Subcloning O f Recovered Variant Oestrogen Receptor Species
Subcloning o f variant ER species recovered from agarose and SSCP gels was carried out
using the pCR-Script™ Amp SK(+) Cloning Kit.
2.12.2.11.1 Pfu PCR Product Generation
lp l of agarose gel isolated, lOpl of polyacrylamide gel isolated ER variants were used as
template for 30-35 cycles o f Pfu PCR in the following 25pl reaction: 20mM Tris-HCl, pH
8.75, lOmM KC1, lOmM (NH4)2S 0 4, 2mM M gS04, 0.1% Triton X-100, O.lmg/ml BSA,
200pM each dNTP, lOpmol each primer, 1 unit Pfu DNA polymerase, and ultrapure water
to 25pl. Hot start PCR carried out as already described with cycling conditions as described
in Appendix II.
2.12.2.11.2 Cloning Reaction
The optimum ratio o f insert to vector for the pCR-Script Amp SK (+) kit is 40:1 to 100:1.
Prior to preparing the cloning reaction the optimum amount of PCR product to produce such
a ratio was calculated. Insertion of Pfu PCR products into pCR-Script Amp SK (+) cloning
vector was carried out at room temperature for 1 hour in the following reaction: 1Ong pCRScript Amp SK (+) cloning vector, 1 x pCR-Script reaction buffer, 0.5mM rATP, 2-4pl
101
blunt-ended Pfu PCR product, 5 units S r f I, 5 units T4 DNA ligase, and ultrapure water to
lOpl. Once completed the reaction was terminated by heating at 65°C for 10 minutes.
2.12.2.11.3 Transformation
40pl Epicurian Coli XL-1 Blue M RF’ Kan supercompetant cells were placed into a 15ml
Falcon 2059 polypropylene tube and p-mercaptoethanol added to a final concentration of
25mM and incubated on ice for 10 minutes. 2pl of cloning reaction was added and mixed to
transformation reaction and incubated on ice for 30 minutes. Transformation reactions were
then heat pulsed for 45 seconds at 42°C and placed on ice for 2 minutes. 450pl of 42°C
prewarmed SOC medium (2% tryptone, 0.5% yeast extract, 0.05% NaCl, 0.1M MgCl2,
0.1M M gS04, 0.4% filter-sterilized glucose) was added and incubated at 37°C for 1 hour
with shaking at 225-250 rpm.
lOOpl of neat, transformation reaction was plated onto LB-ampicillin (1% NaCl, 1%
Tryptone, 0.5% yeast extract, 2 % agar, pH adjusted to 7.0 with 2.5 N NaOH, lOOmg
Ampicillin)
agar
plates
containing
l.lg
X-gal
(5-bromo-4-chloro-3-indoyl-p-D-
galactopyranoside) (mw 408.64) and 0.12g IPTG (isopropyl-1-thio-p-D-galactopyranoside
(mw 238.31) diluted in 20ml molecular biology grade dimethylformamide) and incubated
overnight at 37°C. In addition, the remaining transformation reaction was sedimented at
13,000rpm in an MSE bench top centrifuge, the cells resuspended in 100pi 42°C prewarmed
SOC and plated onto LB-ampicillin plates and incubated overnight at 37°C. White colonies
were restreaked onto fresh LB-Ampicillin plates and incubated overnight at 37°C.
2.12.2.11.4 Culturing O f Variant Oestrogen Receptor Subcloning Reactions
White colonies were removed from LB-Ampicillin agar plates and placed into 5ml LBAmpicillin broth (1% NaCl, 1% Tryptone, 0.5% yeast extract, pH adjusted to 7.0 with 2.5 N
NaOH, lOOmg Ampicillin) and incubated overnight at 37°C with shaking at 225-250 rpm.
2.12.2.11.5 Plasmid Isolation
An alkaline-lysis mini-preparation method of isolating plasmid DNA was carried out using
that described by Stephen et al., (1990). In brief, 5mls of overnight culture were centrifuged
on a bench top centrifuge at 13,000 rpm for 1 minute and the supernatant removed. The
pellet was resuspended in 400pl PI (lOOpg/ml RNase A, 50mM Tris-HCl, pH 8.0, lOmM
EDTA, pH 8.0) and 400pl freshly prepared PII (0.2M NaOH, 1% SDS) added, inverted
102
several times and placed on ice for 5 minutes. 300pl of 3M potassium acetate, pH 4.8 was
added, inverted several times and placed on ice for 5 minutes, followed by centrifugation at
13,000 rpm for 5 minutes at 4°C. The supernatants were removed, 1 volume of absolute
ethanol added and centrifuged at 13,000 rpm for 5 minutes at 4°C. The supernatants were
again removed, pellets allowed to air dry then resuspended in 40pl sterile TE.
2.12.2.12 Sequencing
Isolated DNA was used as template for manual sequencing based on the dideoxy chain
termination method described by Sanger et al., (1977) using the Sequenase™ Version 2.0
DNA Sequencing Kit according to the manufacturers instructions.
2.12.2.12.1 Direct DNA Sequencing O f PCR Products In Agarose Gel Slices
The method used for direct DNA sequencing o f amplified PCR products was based on that
o f Khorana et al., (1994). In brief, PCR products (2-3fig) were gel purified on 0.8%
NuSieve agarose gels with electrophoresis performed in 1 x TAE buffer. Following
electrophoresis, products were briefly stained using ethidium bromide (0.5mg/ml) and bands
o f interest were excised, removing as much excess agarose as possible. Slices were weighed,
an equal volume of deionised water added, and
melted at 75°C for 15 minutes. The
annealing reaction was prepared by combining 5-10pl of melted gel containing template
DNA, lp l (2 pmol/pl) o f sequencing primer, and deionised water to 1 lpl. This mixture was
denatured at 100°C for 3 minutes and annealing was performed by immediately placing the
tube on ice for 5 minutes. Dideoxy DNA sequencing was then performed on lOpl of
annealed template using SequenaseR version 2 (2.16.10.11.4) with minor modifications; 2pl
of 5 x reaction buffer was added to the labelling mix and 4pl of labelling mix was added to
the termination tubes.
2.12.2.12.2 Isolation O f Single Stranded DNA For Sequence Analysis By Streptavadin
Linked Dynabeads
Target sequence o f interest was subjected to PCR amplification as described in section
(2.12.2.7.2) using a lOpmol 5’-biotin modified downstream primer at the appropriate
conditions described in Appendix II. Biotinylated strands were immobilized by the addition
of 10p.l 2 x B&W washed Dynabeads M-280 Streptavidin, incubated at 28°C or room
temperature, and separated using a Dynal magnetic particle concentrator (MPC). The
103
Dynabeads/DNA were washed with 1 x B&W buffer and the DNA duplex melted by
resuspension in fresh 0.1M sodium hydroxide. The biotin strand was isolated using the MPC
followed by washing the Dynabeads/ssDNA sequentially with 0.1M sodium hydroxide, 1 x
B&W buffer and with 1 x TE. Dynabeads/ssDNA were then resuspended in 7pl deionised
water. The resultant solution was then used as template for sequencing using SequenaseR
version 2.0.
2.12.2.12.3 Plasmid Sequencing
The alkaline denaturation method used to denature double stranded plasmid DNA was that
described in the Sequenase
version 2.0 manual. In brief, approximately 5fig of plasmid
isolate was denatured by the addition o f 1710th volume of 2M NaOH, 2mM EDTA followed
by incubation at 37°C for 30 minutes. The mixture was then neutralized by the addition of
0.1 volume 3M Na acetate, pH 4.5, and DNA precipitated with 3 volumes of absolute
ethanol at -70°C for 15 minutes. After sedimentation at 13,000 rpm on a bench top
centrifuge, the pellet was allowed to air dry, resuspended in 7fil sequencing primer (~10ng).
The resultant solution was then used as template for sequencing using SequenaseR version
2 .0 .
2.12.2.12.4 Sequencing Protocol
2fil o f sequenase reaction buffer (final concentration 40mM Tris-HCl, pH 7.5, 20mM
MgCl2, 250mM NaCl) and 1 pmol (for streptavidin isolated DNA) or 10 pmol (for plasmid
DNA) o f the appropriate sequencing primer were added to 7|il of single stranded DNA and
incubated at 65°C for 2 minutes, and allowed to anneal by incubating the mixture at room
temperature for 10 minutes and on ice for 5 minutes.
The labelling reaction then proceeded by the addition o f the following on ice: 15mM
dithiothreitol; 0.5fiM dGTP, dCTP, dTTP; 25mM sodium isocitrate, 15mM MnCl2; 3.7mCi
a - 35S dATP; 3.25 units o f Sequenase version 2.0 T7 DNA polymerase prediluted in 20mM
potassium phosphate, pH 7.4, 4.5 mM dithiothreitol, 0.0125mM EDTA, 6.25% glycerol,
8.7mM Tris-HCL, pH 7.5, and 0.4375|ig/ml BSA, and incubated at 19°C for 5 minutes.
3. 5 (0.1 portions o f labelling reaction were then transferred to each of four termination
reactions containing 2.5pl o f 80|iM each dNTP, 50mM NaCl and 8(iM of the appropriate
ddNTP, preheated at 37°C. The termination reaction was allowed to proceed at 40°C for 3-5
104
minutes with the reaction terminated by the addition of 4pl of stop solution (95%
formamide, 0.05% bromophenol blue, 0.05% xylene cyanol FF).
Sequencing reactions were fractionated on a 6% denaturing polyacrylamide gel in 1 x TBE
at 70W for 2-4 hours. Gels were then dried and exposed to autoradiography 1- 5 days.
2.12.2.13 Biotinylation O f Oligonucleotide Primers
Biotinylation o f oligonucleotide primers was carried out using a commercial kit according to
the manufacturers instructions with minor variations. The method is based on two steps.
Firstly a phosphorothioate group is transferred to the 5’ hydroxyl group of the
oligonucleotide in a reaction catalyzed by T4 polynucleotide kinase using adenosine-5’-O(3-thiotriphosphate) (ATPyS) as a substrate. The second step involves a reaction with the
phosphorothioate oligonucleotide with N-iodacetyl-N’-biotinylhexylenediamine (NIBH)
which transfers a biotin residue, along with the linker, to the sulphur atom of the
phosphorothioate. The biotinylated primer is then purified.
The first step involved the following reaction: 500 pmol o f oligonucleotide, 1 x T4
polynucleotide kinase reaction buffer, 0.1 mM ATPyS , 30 units T4 polynucleotide kinase in
a final volume o f 50pl, incubated at 37°C for 1 hour, and heat inactivated at 70°C for 10
minutes. DNA was then ethanol precipitated and the DNA resuspended in the following
second stage reaction: 90 mM potassium phosphate, pFl 8.0, ImM NIBH in a final volume
o f 50pl and incubated at 50°C for 1 hour. The reaction volume was adjusted to lOOpl with 1
x TE and the biotinylated primer purified on a Sephadex G50 column equilibrated with 1 x
TE. 5pmols o f biotinylated primer then used in a conventional PCR reaction.
2.12.2.14 DNA Isolation From Frozen Tissue Using Trizol Reagent
The remaining aqueous phase removed from RNA extraction procedure, section
(2.13.2.4.2), and DNA precipitated from the interphase and organic phase by addition of
150pl o f 100% ethanol, mixing by inversion, incubation at room temperature for 2-3
minutes and centrifuged at 2,000 x g for 5 minutes at 4°C. The supernatant was removed and
the DNA pellet washed twice in 1ml of a 0.1 M sodium citrate/10% ethanol solution. During
each wash the DNA pellet was allowed to stand for 30 minutes with occasional mixing. The
DNA pellet was sedimented at 2,000 x g for 5 minutes at 4°C, resuspended in 1.5 ml of 75%
ethanol and allowed to stand at room temperature for 20 minutes. The DNA pellet was again
105
sedimented at 2,000 x g for 5 minutes at 4°C, the supernatant removed, allowed to air dry
and resuspended in 30pl ultrapure water.
106
2.13
Results - Optimization O f Methodologies
2.13.1 RNA Isolation
2.13.1.1 Cell Line RNA
RNA preparations were carried out on breast cancer cell lines T47-D, MCF-7 and MDA MB
231 for use as template for the optimization of methodologies. Typically, RNA yields were
in the range 600-1600pg/ml from 106 cell as, determined by spectrophotometric
measurement at 260nm. Figure 2.6 illustrates a successful total RNA preparation from the
T47-D cell line. Strong ethidium bromide staining indicates the presence of 28S and 18S
ribosomal RNAs. The low molecular weight smear visualized may represent a low degree of
degradation. Alternatively, this RNA may represent low molecular weight RNA (4-5S) as
described by Chomczynski and Sacchi (1987).
28S
18S
4 -5 S
F igure 2.6 Agarose electrophoresis o f RNA isolated using TRIzol reagent from T47-D breast carcinoma cell
line stained with ethidium bromide
2.13.1.2 Frozen Breast Tumour RNA
RNA isolations were carried out using TRIzol reagent on frozen breast cancer tumours, the
first on a large sample of asymptomatic breast carcinoma (A) of approximately 1cm in
length and 3-4mm in width and a smaller sample of asymptomatic breast carcinoma (B) of
approximately 3-4mm in length and 3-4mm in width. In addition, 1ml, 0.4ml and 0.25ml of
TRIzol reagent and 1, 2, and 3 serial tumour sections were used in the extraction to
determine the optimum volume of reagent and the optimum amount of tissue to obtain a
reasonable RNA yield. In all cases quality and quantity of isolated RNA was assessed by
spectrophotometric measurement at 260nm and 280nm the results of which are summarized
in Table 2.9.
Both (A) and (B) gave good RNA yields (48-112pg/ml for tumour A and 32-72pg/ml for
tumour B) using 1ml TRIzol reagent with 1, 2 and 3 serial sections. As expected, a greater
107
yield was obtained from tumour A. Decreasing the amount of TRIzol did not appear to
affect the yield o f RNA. In all cases, the 260nm/280nm ratio were in the range of 1.0 to 1.5.
Table 2.9 Effect o f tumour sample size and quantity (ml) of TRIzol reagent on quantity and
quality of RNA isolates from fresh frozen breast tumours
ml Trizol
1
1
Tumour
(A)
(B)
No. of
sections
1
OD
260nm
0.006
260/280
ratio
1.2
[pg/ml]
2
0.008
1.2
64
3
0.014
1.3
112
1
0.004
1.0
32
2
0.005
1.1
40
3
0.009
1.2
72
48
0.4
(B)
3
0.01
1.5
80
0.25
(B)
3
0.01
1.0
80
RNA was isolated from 3 serial sections from the study group consisting of 44
asymptomatic breast cancers using 0.4ml TRIzol reagent. Yields were in the range 16336pg/ml with 260nm/280nm ratios in the range 1.3-3.3 A high proportion had ratios within
the range 1.4-1.9.
2.13.2 PCR Optimization
In order to optimize the amplification conditions for ER primer sets, PCR was carried out in
PCR buffer using either ER cDNA or reverse transcription product as template at the
annealing temperatures described in Table 2.10. PCR products were then fractionated on 2%
Seakem agarose gels, stained with ethidium bromide and visualized under UV illumination,
the results o f which are shown in figure 2.7(a-f).
PCR amplification generated fragments ranging from 224 bp to 932 bp. Amplification using
1092/2024 primers would theoretically produce a product of approximately (932 bp) but
amplification with ER cDNA as template yielded a product o f approximately 600 bp (Figure
108
F igu re 2 .7 (a -f) A garose electrophoresis o f oestrogen receptor PCR products. PCR was carried out using ER
cD N A (Figure 2.7(a) lanes 1-3; Figure 2.7(c) lane 1; Figure 2.7(e) lane I; Figure 2.7 (f) lane 1) or asymptomatic
breast carcinom a reverse transcription product (Figure 2.7(b) lane 1; Figure 2.7(d) lane 1; Figure 2 .7 (f) lane 2).
Products w ere observed which corresponded to their theoretical m olecular size (bp) when compared to molecular
w eight markers as indicated.
o
-a
—
-
o.
JD
603
310
234
194
F ig 2 .7 (a )
F ig 2 .7 (b )
r£
1078
1.
—
853
389
F ig 2 .7 (d )
F ig 2 .7 (c )
1.
2.
% *
1078 —
932
872 —
hM Ji
603
550-600
645
603
310
F ig 2 .7 (e )
F ig 2 .7 ( f )
109
2.7(f) lane 1).
In addition ER Set 5 and ER Set 7 primers failed to amplify ER cDNA
Amplification for these primer sets was re-optimized using reverse transcription product
from a randomly selected asymptomatic breast tumour from the study group. Amplification
resulted in products which were approximate in size (bp) to their theoretical values the
results o f which are shown in Figure 2.7(f) lane 2 for 1160/2092; Figure 2.7(b) for ER Set 5
and Figure 2.7(d) for ER Set 7. Amplification using the remaining primer sets yielded
products approximate to their theoretical sizes the results of which are shown in Figure
2.7(a) for ER Set 2. ER Set 3, ER Set 4 (lanes 1, 2, 3 respectively); Figure 2.7(c) for ER Set
6 and Figure 2.7(d) for 639/1384.
Table 2.10 Summary o f annealing temperatures and sizes of PCR products
Primer Set
Annealing Temp
Product Size
Template
(°C)
(bp)
ER Set 2
56
349
ER cDNA
ER Set 3
58
224
ER cDNA
ER Set 4
68
438
ER cDNA
ER Set 5
68
415
RT Product
ER Set 6
56
389
ER cDNA
ER Set 7
68
853
RT Product
739/1384
59
645
ER cDNA
1160/2092
64
932
RT Product
2.13.3 Nested PCR Optimization
To increase the sensitivity o f detection o f ER transcripts, a nested PCR approach was
undertaken. Optimization was carried out with first round PCR using external primers (739
and 1384), which flank the DNA binding domain of the ER gene using both ER cDNA and
reverse transcriptase product from a randomly selected asymptomatic breast tumour from
the study group as template. Second round PCR was carried out using lp l of: neat, 1/10,
1/100, 1/1000 dilution o f first round product as template with ER Set 6 PCR primers, the
results o f which are shown in Figure 2.8. Nested PCR using ER cDNA (lanes 1-4) and RT
(lanes 6-10) as template resulted in observable products at the correct size of approximately
350 bp. In both cases, smearing was also observed, which was more severe for ER cDNA,
but which decreased with increasing dilution of first round product. The most specific
product with the least amount o f smearing was observed at the 1/1000 dilution of first round
110
Tf
1.
310
2.
3.
4.
5.
6.
7.
8.
9. 10.
—
F igure 2.8 Agarose electrophoresis o f ER S et 2 nested PCR products using 739/1384 first round products on
ER cD N A (lanes 1-4) and asymptomatic carcinoma reverse transcription product (lanes 7-8). Second round
amplification was carried out using neat (lanes 1 and 7), 1/10 (lanes 2 and 8), 1/100 (lanes 3 and 9) and 1/1000
dilution’s (lanes 4 and 10) o f first round product. An ER cD N A and a no template reaction were carried out as
positive and negative controls respectively.
product with the least amount of smearing was observed at the 1/1000 dilution of first round
RT-PCR product (Figure 2.8 (lane 10)).
2.1.4 Reverse Transcriptase - PCR (RT-PCR) Optimization
2.13.4.1 Cell Line RNA
In order to optimize the RT-PCR method varying quantities (0, lpg and 2pg) of T47-D cell
line RNA were used as template for reverse transcription using AMV reverse transcriptase
at 42°C with the following primers for first strand cDNA synthesis: random hexamers, oligo
dT and gene specific primers for ER Set 2 and actin. First strand cDNA were then used as
template for PCR using ER set 2 and actin primer sets. The products were fractionated on a
2% agarose gel, stained with ethidium bromide and visualized under UV illumination.
Amplification products were only just observable using lpg of template irrespective of the
reverse transcriptase primer used with no obvious amplification products from the other
reactions (results not shown).
Due to the low sensitivity of ethidium bromide staining, products were blotted onto a nylon
membrane and probed using a
12
P labeled-PCR generated cDNA fragment of the ER DBD.
After overnight hybridization the membrane was exposed to x-ray film for 1 to 5 days
(results not shown).
To overcome problems of poor sensitivity and probe labelling, an alternative approach was
undertaken in which 35S dATP was incorporated into PCR product during the amplification
111
reaction using the 1ng total RNA reverse transcriptase reactions described above as template
with ER Set 2 and actin primer pairs. PCR products were fractionated on 6% polyacrylamide
gels and exposed to X-ray film. Figure 2.9 (lanes 1-10) shows an autoradiographic image of
35S RT-PCR in which products for both ER Set 2 and actin can be observed using random
hexamer primer (lanes 1 and 2), for actin using oligo dT primer (lanes 3 and 4), with no
observable products using gene specific downstream primers (lanes 5 and 6).
RT-PCR was repeated using random hexamers and ER Set 2 to validate these results and
included reverse transcriptase negative controls, i.e. a no RNA template control (-RNA), and
a reaction with lp g total RNA with no reverse transcriptase enzyme (-RT), the results of
which are shown in Figure 2.10(lanes 1-4). Products were observed for lp g and 500ng RTPCR reactions (lanes 1 and 2) but not for the -RNA control (lane 3). Product was also
observed for the -RT control (lane 4). The experiment was repeated using ER Set 2 and actin
primers taking additional precautions as outlined in section (2.13.2.5.1.1) with varying
concentrations o f RNA as template for reverse transcription (lOOng-lpg). Figure 2.11 (lanes
1-12) shows the resultant autoradiographic image where amplification products can be seen
for both ER Set 2 (lanes 1-4) and actin (lanes 7-10). No observable products were seen in the
-RNA (lanes 5 and 11), and the -RT (lanes 6 and 12 ) negative controls.
2.13.4.2 Frozen Breast Tissue RNA
Labeled RT-PCR analysis was carried out using RNA isolated from frozen breast tumours A
and B section (2.12.2.6.1.2). lOpl o f each RNA isolation (i.e., the x 1, 2 and 3 section
isolations) was used as template for reverse transcription at 42°C using random hexamer
primers. First strand cDNA was then used as template for labeled PCR using ER Set 2, with
resulting labeled products fractionated on a 6% denaturing polyacrylamide gel, dried and
exposed to X-ray film. Figure 2.12 (lanes 1-10) shows the resultant autoradiographic image.
Amplification was not observed for tumour sample (A) (lanes 1-3) but observed for the
larger tumour sample (B) (lanes 4-6) with the no amplification for the -RNA and -RT
negative controls (lanes 7 and 8).
Asymptomatic tumour, oligo d(T)25 isolated, mRNA-RT products were amplified using ER
Set 2, ER Set 3 and actin primers with the products fractionated on 2% Seakem agarose gels,
stained with ethidium bromide and visualized under UV illumination (Figure 2.13(a-c)).
112
F igure 2.9 Polyacrylam ide gel electrophoresis o f ER Set 2 and A ctin RT-PCR products. 1pg T 47-D total RNA
was reverse transcribed using random hexamers (lanes 1 and 2), oligo dT(25), and downstream primers and
used as template for PCR using ER Set 2 (odd lanes) and A ctin (even lanes) primers. Products were
fractionated alongside PCR positive (ER cD N A (lane 7) and genom ic D N A (lane 8) for ER Set 2 and Actin
respectively) and negative (no tem plate) controls (lanes 9 and 10). A m plification products were observed for
reverse transcriptions primed with random hexamers (ER S et 2 and Actin (lanes 1 and 2 respectively) and for
oligo dT(25) (Actin lane 4). N o am plification products w ere observed for reverse transcriptions primed with
downstream primers (lanes 5 and 6).
F igu re 2.10 Polyacrylam ide gel electrophoresis o f ER S et 2 RT-PCR products including a reverse
transcription negative control, lp g , 0.5 p g , Ofig T 47-D total R NA w as reverse transcribed using random
hexamers and used as tem plate for PCR am plification using ER S et 2 primers. A reverse transcription reaction
containing lp g total R NA with no reverse transcriptase added (-R T negative control) w as also included (lane
4). Products were observed for lp g , 0 .5 p g and -RT controls (lanes 1, 2, and 4 respectively)
F igu re 2.11 Polyacrylam ide gel electrophoresis o f ER S et 2 and A ctin RT-PCR products with varying
concentrations o f T 47-D total RNA and after additional precautions to m inim ize aerosol contam ination, lp g ,
0.5p g, 0.2p g and 0.1 fig total RNA w as reverse transcribed using random hexamers and used as tem plate for
PCR am plification using ER S et 2 and Actin primers. In addition reverse transcriptions containing no R N A
(lanes 5 and 11) and R NA with no reverse transcriptase enzym e (lanes 6 and 12) were also included. Products
were observed for all concentrations for both ER S et 2 and Actin (lanes 1-4 and lanes 7-10 respectively) with
no observable product in the no RNA (lanes 5 and 11) and -RT (lanes 6 and 12) negative controls.
F igu re 2.12 Polyacrylam ide gel electrophoresis o f ER S et 2 RT-PCR products using RNA isolated from a
sm all and large sam ple o f asym ptom atic breast carcinomas. lOpl o f RNA isolation using 1, 2, and 3 10pm
frozen tumour section from a sm all and large asym ptom atic breast carcinom a sam ples was reverse transcribed
using random hexam ers and used as tem plate for PCR using ER S et 2. In addition reverse transcriptions
containing no RNA (lane 7) and R N A with no reverse transcriptase enzym e (lane 8) were also included.
Products w ere fractionated alongside PCR positive (ER cD N A (lane 9)) and negative (no tem plate (lane 10))
controls. Products w ere only observed using R N A from the larger carcinoma sample (A ) isolated from 2 and 3
serial sections (lanes 5 and 6 respectively) with no observable products for the smaller tumour sam ple (B )
regardless o f the number o f sections utilized (lanes 1-3) or for the larger tumour RNA isolation from a sin gle
section. In addition no observable product w as observed for the no R NA (lane 7) or -RT (lane 8 ) negative
controls.
1. 2 .
3. 4.
5. 6.
7.
8.
9. 10.
ft* —
F ig 2 .9
1.
2.
3.
4.
F ig 2 .1 0
1. 2.
3. 4.
5.
6.
7. 8.
9. 10.
11. 12.
F ig 2.11
1.
2.
3.
4 . 5.
6.
F ig 2 .1 2
113
7.
8.
9. 10.
1.
2.
3.
4.
■I
Figure 2.13(a) Agarose electrophoresis o f ER S et 2 RT-PCR products o f Dynabead oligo d (T ) 2 5 isolated
RNA from frozen, asymptomatic breast carcinoma. Products were observed in lanes 1 and 3 (+RT reaction
and PCR positive control respectively) but absent in lanes 2 and 4 (-RT enzyme reaction and no template PCR
controls respectively).
2.
3.
4.
Figure 2.13(b ) A garose electrophoresis o f ER S et 4 RT-PCR products o f Dynabead oligo d(T)25 isolated RNA
from frozen, asymptomatic breast carcinom a. Products were observed in lanes 1 and 3 (+RT reaction and PCR
positive control respectively) but absent in lanes 2 and 4 (-RT enzym e reaction and no template PCR controls
respectively).
1.
2.
3.
4.
Figure 2.13(c) A garose electrophoresis o f Actin RT-PCR products o f Dynabead oligo d(T)25 isolated RNA
from frozen, asym ptom atic breast carcinoma. Products were observed in lanes 1 and 3 (+RT reaction and PCR
positive control respectively) but absent in lanes 2 and 4 (-RT enzyme reaction and no template PCR controls
respectively).
114
RT-PCR products could be observed for all primers sets (lanes 1 in all cases) with no
observable products in the -RT negative controls (lane 2 in all cases).
2.13.5 Thermus therm ophilus Reverse T ran scrip tase - PC R
Reverse transcription was also carried out using the enzyme Tth DNA polymerase, a
thermostable enzyme isolated from Thermus thermophilus HB-8 (Ruttimann et al., 1985).
Tth DNA polymerase has DNA dependent polymerase activity in the presence of
magnesium chloride, but can catalyze the polymerization o f DNA using an RNA template in
the presence o f manganese chloride. Tth has a number o f advantages over AMV reverse
transcriptase as a result o f the higher temperature at which reverse transcription is carried
out. Tth reverse transcribes at 70°C and consequently strong secondary structural problems
encountered with RNA are minimized, since these structures are unstable at this
temperature. In addition, higher temperature also results in increased specificity of primer
hybridization and extension.
lp g o f the MCF-7 cell line RNA was used as template for reverse transcription with Tth at
70°C using oligo dT and gene specific primers. The products were fractionated on agarose
gels, stained with ethidium bromide and visualized under UV illumination. In all case there
was no evidence o f any RT-PCR product by ethidium bromide staining. In addition, PCR
positive controls consisting o f a ER cDNA template, for ER, and human genomic tonsil
DNA, for Actin were also negative.
2.13.6 SSCP O ptim ization
2.13.6.1 Isotopic SSCP
2.13.6.1.1 Genomic DNA (Accugel Gels)
Genomic DNAs isolated from tumours and peripheral blood lymphocytes from the same
individuals (tumour/normal pairs) were used as template for 35S PCR using ER Set 3, and
fractionated through 6% nondenaturing polyacrylamide gels. Figure 2.14(1-21) is a
representative SSCP autoradiograph showing tumour normal pairs (lanes 1-20) fractionated
alongside reference ER cDNA PCR products (lane 21). A complex banding pattern was
observed with no apparent shift observed for tumour products when compared to their
respective normal products.
115
F igure 2.14 Single stranded conformational polymorphism analysis o f ER Set 3 35S labeled PCR products
using, high molecular weight DNA extracted from breast carcinomas (odd lanes), and peripheral blood
lym phocytes (even lanes) from 10 patients as template. Tumour and corresponding normal PCR products were
heat denatured, allow ed to cool and co-fractionated on 6% non denaturing polyacrylam ide gels alongside an
ER cD N A control.
2.13.6.1.2 Frozen Tumour Tissue Total RNA RT-PCR ( MDE Gels)
lOOng total RNA isolated from the asymptomatic breast carcinoma group was used as
template for 35S RT-PCR using ER Set 2 and fractionated through 0.5 x nondenaturing
MDE Figure 2.15(1-10) shows a representative SSCP autoradiograph in which tumour
products (lanes 1-7) fractionated alongside an ER cDNA wt control (lane 8). Co-migration
of tumour samples alongside wt controls was observed with no apparent bandshifts detected.
2.13.6.1.3 Frozen Tissue Dynabead-Oligo dT RT-PCR (MDE Gels)
Dynabead-oligo d(T)25 isolated mRNAs from asymptomatic, invasive breast carcinomas
were reverse transcribed and used as template for 35S RT-PCR using ER Set 2, ER Set 4 and
ER Set 5. Products were then fractionated through 0.5 x (ER Set 2 and ER Set 4) or 0.7 x(ER
Set 5) non denaturing MDE gels. Figure 2.16(a-c) shows representative SSCP
autoradiographs where tumour RT-PCR products using the three primer sets were co­
fractionated alongside wt ER cDNA PCR products (for ER Set 2 and ER Set 4 only).
2.13.6.2 Non-Isotopic SSCP
2.16.6.2.1 Genomic DNA (Accugel And Metaphor Agarose Gels)
The non-isotopic SSCP method described by Yap and McGee, (1992) was carried out.
1OOng of tumour and normal genomic DNA from the same individuals were amplified using
116
1. 2.
3.
4. 5.
6.
7.
8.
1
Figure 2.15 Single stranded conformational polymorphism analysis o f ER Set 2 35S labeled RT-PCR products
using lOOng o f total RNA extracted from asymptomatic breast carcinomas as template. Tumour PCR products
(lanes 1-7) were heat denatured, allow ed to cool and co-fractionated on a 0.5 x non denaturing MDE
polyacrylamide gel alongside an ER cD N A control (lanes 8).
1. 2.
3. 4.
5. 6.
7. 8. 9. 10. 11.
F igure 2.16(a) Single stranded conform ational polymorphism analysis o f ER Set 2 35S labeled RT-PCR
products using Dynabead oligo d(T)25 mRNA extracted from asymptomatic breast carcinomas as template.
Tumour PCR products (lanes 1-7) were heat denatured, allowed to cool and co-fractionated on a 0.5 x non
denaturing MDE polyacrylamide gel alongside an ER cD N A control (lane 8).
1. 2 3. 4. 5. 6.
F igure 2.16(b) Single stranded conformational polymorphism analysis o f ER Set 4 35S labeled RT-PCR
products using Dynabead oligo d(T)25 mRNA extracted from asymptomatic breast carcinomas as template.
Tumour PCR products (lanes 1-5) were heat denatured, allowed to cool and co-fractionated on a 0.5 x non
denaturing MDE polyacrylamide gel alongside an ER cD N A control (lane 6).
1. 2. 3. 4. 5.
6. 7.
Figure 2.16(c) Single stranded conformational polymorphism analysis o f ER Set 5 35S labeled RT-PCR
products using Dynabead oligo d(T)25 mRNA extracted from asymptomatic breast carcinomas as template.
Tumour PCR products (lanes 1-7) were heat denatured, allowed to cool and fractionated on a 0.7 x non
denaturing MDE polyacrylam ide gel.
117
ER Set 3 and lOpl and 15pl of PCR product loaded onto 6%, 8%. and 10% non-denaturing
gels. Prior to loading samples were denatured either by alkali treatment or heating. A Cystic
fibrosis A508 3 base pair deletion carrier and a non-carrier DNA was also amplified to serve
as a positive SSCP control.
Only 8% gels with 15pl o f product gave any interpretable results for both patient and cystic
fibrosis A508 products, with alkali denaturation less efficient than heat denaturation (results
not shown). 6% gels gave rise to extensive smearing, 10% gels retarded samples and lOjal
PCR product was difficult to visualize (results not shown).
An alternative method utilized MetaPhor XR high resolution agarose in a horizontal and
vertical format. Samples described above were loaded on 4% gels after heat denaturation.
After fractionation gels were stained with both ethidium bromide or Syber Green II dye.
Resolution for all templates using both staining methods was poor (results not shown).
Analysis in a vertical format was impossible due to difficulty in pouring the very viscous
agarose.
2.13.6.2.2 p53 wt/mt controls / Frozen Tissue Dynabead Oligo d(T);5RT-PCR (MDE Gels)
A non isotopic SSCP method was carried out as described in section (2.15.5.1) using
Dynabead oligo d(T);5 reverse transcription product as template for non isotopic PCR using
ER Set 4. In addition, wt and mt p53 control plasmids were also amplified. Products were
then fractionated on 0.5 X non denaturing MDE gels. Once fractionated, gels were stained
with ethidium bromide and visualized under UV illumination. Figures 2.17(a) and 2.17(b)
show representative non isotopic SSCP for ER Set 4 and p53 respectively. In both cases
bandshifts can be seen in test or mutant sample when compared to wt control.
2.13.7 Recovery Of Aberrant Bands From Gels
Aberrant ER PCR products showing either gross molecular weight alterations identified on
agarose gels or exhibiting mobility shifts on SSCP gels were isolated as described in section
(2.12.2.10). Isolation was assessed by PCR reamplification using the primer sets used in
the original PCR amplification, the results of which are shown in Figure 2.18(a-b). Figure
2.18(a) (lanes 1- 4) demonstrate reamplification of product recovered from agarose gels
using ER Set 2 while Figure 2.18(b) demonstrates reamplification of product isolated from
an agarose gel (lane 1) and from an MDE polyacrylamide gel (lane 2) using ER Set 4.
118
►
I
Fig 2.17(b)
Fig 2.17(a)
F igure 2.17(a-b ) Non isotopic single stranded conformational polymorphism analysis o f p53 controls and ER
S et 4 PCR products using p53 and ER cD N A control plasmids and Dynabead oligo d(T)25 mRNA isolated
reverse transcription products as template. P53 wt and m t products (Fig 2.17(a)) and ER cD N A and
asym ptom atic 43 products (Fig 2.17(b)) were heat denatured, allowed to cool and co-fractionated on 0.5 x non
denaturing MDE polyacrylamide gels.
603 —
310 —
F igure 2.18(a) Agarose gel analysis o f ER Set 2 reamplification PCR products. PCR products from
asym ptom atic breast carcinomas which demonstrated gross molecular weight alterations were gel purified on
agarose gels and aberrant species isolated. Determination o f isolation efficiency w as determined by
ream plifying isolated DNA by PCR and fractionating products on agarose gels. Reamplification products o f
altered molecular weight were observed (lanes 1-4) alongside a wt ER cD N A control.
F igure 2.18(b) Agarose gel analysis o f ER Set 4 reamplification PCR products. PCR products from
asym ptom atic breast carcinomas which demonstrated gross molecular weight alterations or MDE SSCP
bandshifts were gel purified and aberrant species isolated. Determination o f isolation efficiency was
determined by reamplifying isolated D N A by PCR and fractionating products on agarose gels.
Reamplification products o f altered molecular weight (lane 1) and from an SSCP bandshift (lane 2) were
observed.
119
2.13.8 Sequence Analysis O f O estrogen Receptor Variants
Sequencing o f ER cDNA was carried out using the di-deoxy chain termination method
described by Sanger (1977) with either agarose gel purified PCR product, biotinylated single
stranded DNA PCR product, or double stranded plasmid DNA as template.
2.13.8.1 Agarose Gels Slices
The method for sequencing gel purified PCR products was based on the method of Khorana
et al., (1994). PCR products were fractionated on low melt temperature agarose gels, the
bands o f interest excised, resuspended in water, and used directly in the annealing reaction.
Sequencing was then carried out using the Sequenase protocol. Limited success was
achieved, with poor quality sequence, i.e., a high degree of cross banding, requiring lengthy
exposure of the dried gel to x-ray film (results not shown).
2.13.8.2 Biotin PCR Method
An alternative method utilized sequencing single strands which had been isolated by
magnetic selection (Dynabeads). ER cDNA, RT products, agarose and polyacrylamide gel
purified RT-PCR products were used as template for PCR using a biotinylated primer.
Streptavidin labeled magnetic beads were introduced to biotin PCR products. After washing,
the salt concentration was increased, releasing single strands which were then used directly
in the annealing reaction. Sequencing was then carried out using the sequenase system.
Again, limited success was achieved, with poor quality sequence with a high degree o f cross
banding, requiring lengthy exposure o f the dried gel to x-ray film (results not shown).
2.13.8.3 Plasmid Sequencing
Pfu amplified PCR products using ER molecular weight or SSCP bandshift variants were
cloned into pCR-Script Amp SK (+) cloning vectors, transfected into Epicurian Coli XL-1
Blue M RF’ Kan supercompetant cells, cultured overnight and minipreped. Isolated, double
stranded clones were then alkali denatured, precipitated and subjected to sequencing using
the sequenase system. This method yielded sequence of excellent quality from an overnight
to 2 day exposure with limited crossbanding and length of read exceeding 400 nucleotides
when double loaded (results shown in section 2.14.2).
120
2.14
Results
Once all methodologies had been optimized the mammographically detected, invasive breast
carcinoma group were screened for alterations using the approach outlined in Flow Diagram
2.
3 x lOpM Sections o f
M ammographically Detected
Breast Carcinomas
1
T R lzol R N A Isolation
1
O ligo d(T )25 mRNA Isolation
1
A M V Reverse Transcription
1
1st Round PCR
1
2 nd Round Nested PCR
DNA Binding Domain
Proximal Hormone
Binding Domain
Distal Hormone
Binding Domain
Single Stranded Conformational
Polymorphism A nalysis
Agarose Gel
Analysis
Non-Isotopic RNase
Cleavage Assay Analysis
Subcloned U sing pCR-Script™
Am p SK(-i-) Cloning Kit
Sequence Characterization
F low D iagram 2 O verview o f approach undertaken for mutational analysis o f the oestrogen receptor gen e in
m am m ographically detected, invasive breast carcinom as. Total RNA w as isolated from 3 x lOpM sections
using T R lzol reagent and enriched for m R N A using oligo d(T)25 Dynabeads. m RN A w as then reverse
transcribed using A M V using o lig o d (T )25 prim ing to produce first strand cD N A . cD N A w as then used as
template for first round PCR using primers w hich flanked either the ER D N A or horm one binding dom ains
w hich w as then diluted and used as tem plate for PCR using nested primers to these areas. Isotopic PCR
products were analyzed for alterations by sin gle stranded conformational polymorphism analysis, with non
isotopic PCR analyzed by agarose gel analysis and non isotopic RNase cleavage analysis. Aberrant sp ecies
were then subcloned using a pCR -Script A m p SK (+) C loning kit and characterized by dideoxynucleotide
sequencing.
121
2.14.1 ER, PgR And MIB-1 Im m unocytochem ical Analyses
ER, PgR and MIB-1
immunocytochemistry was carried out to characterize new,
mammographically detected cases, selected as being no greater than 1.5 cm in diameter, and
to enable selection o f ER positive tumours for mutation analysis.
Figure 2.19(a) and 2.19(b) demonstrates a typical ER positive tumour with localization of
the signal limited to the nucleus with no cytoplasmic staining.
This pattern o f staining was observed also for PgR and MIB-1 (Figure 2.20(a) and 2.20(b)
respectively). As expected, staining was also observed in vascular endothelial cells and
erythrocytes which have been shown to be due to endogenous peroxidase activity present in
these cells (Graham et al., 1966). Although this endogenous peroxidase activity can be
quenched by pretreatment with hydrogen peroxide (Li et al., 1987), given that this staining
was easily distinguishable from that observed within tumour cell nuclei, such pretreatment
was deemed unnecessary.
Cases were scored for ER and PgR using the H score system and proliferation indices were
calculated (% stained cells). The data obtained for each case are summarized in Appendix I.
Thirty nine o f the 44 carcinomas (88%) were positive for ER. The H scores ranged from 50
to 261, with 9 having scores between 50 and 100, 23 between 100 and 200 and 7 greater
than 200. The mean ER H score for tubular carcinomas was 164.6, for well differentiated
infiltrating ductal carcinomas 154, for moderately differentiated infiltrating carcinomas 151,
and for poorly differentiated infiltrating carcinomas 67.9. All 5 of the carcinomas which
were ER negative were infiltrating ductal, one being well to moderately differentiated and
four poorly differentiated.
Twenty one o f the 44 carcinoma group (47.7%) were positive for PgR, 21 o f the 39 ER
positive carcinomas were positive for PgR (47.7%) and 18 were negative for PgR (40.9%).
Five carcinomas were negative for both
ER and PgR (11.36%). No ER negative PgR
positive tumours were characterized.
122
F igure 2 .19(a) x 250 Formalin-Fixed mammary carcinoma tissue (grade I IDC), microwave oven pretreated,
and im m unohistochem ically stained for oestrogen receptor by ABC-HRP diaminobenzidine H20 2 procedure.
N eoplastic cells are heavily stained showing nuclear localization o f sig n a l, whereas stromal cells are negative.
'
V-
|. M
L*
*
n
V»
4
&
»#£>►
i
F igure 2.19(b) x 400 photograph o f above case show ing nuclear localization
A
l
/
»*
...
A »
t
u
■/tt
V /
/V
y
--i-V
, .iy * :
^
iry
123
\
< > '
F igure 2.20(a) x 250 Formalin-Fixed mammary carcinoma tissue (same case), microwave pretreated and
stained by immunohistochemistry for progesterone receptor by ABC-HRP diaminobenzine.
Figure 2.20(b) x 250 Formalin-fixed mammary carcinoma tissue, microwave pretreated and stained by
immunohistochemistry stained for MIB-1 by ABC-HRP diaminobenzine. Demonstrates a carcinoma showing
a high level o f proliferation.
■v *
dr
n
&
w
% *
*
*
I
t
*
*
I
.
%
124
The proliferative indices of the study group ranged between 1 to 80 with the overall mean
proliferative index of the group 12.3. The mean values for this study group increased with
histological grade with a mean values for tubular carcinomas of 6.6% (range 2.7 to 10), for
well-differentiated infiltrating ductal carcinomas 6.9 (range 1 to 15.4), for well to
moderately differentiated infiltrating ductal carcinomas 10.7 (range 1.4 to 30) and poorly
differentiated infiltrating ductal carcinomas 33.13 (range 1 to 80).
The mean proliferative index for ER positive carcinomas was 8.9% compared with a mean
proliferative index of 40.2% for the ER negative tumours. Two of 7 (28.57%) carcinomas
with a H score greater than
200 had a proliferation index greater than that of mean
proliferative index and 6 o f 32 (18.75%) carcinomas with H score between 50 and 200 had a
proliferation index greater than the mean index (Table 2.11). ER positive/PgR positive
tumours had a mean proliferative index o f 7.02% compared with a mean proliferative index
o f 11.45% for the ER positive/PgR negative phenotype. The majority of carcinomas with
high proliferation rates were moderately to poorly differentiated (12 of 13 carcinomas - 6 of
each: well to moderately differentiated and poorly differentiated, and 1 well differentiated).
T able 2.11 Comparison o f oestrogen receptor H-score and proliferation index (MIB-1 cut
off point = mean value) in relation to type and grade of mammographically detected
Carcinoma
H -score 50-199
H-score 200 and
greater
< 12,.64 > 12.64
^ 12.64
> 12.64
Grade I
6
1
j
0
Grade II
9
j
2
2
Grade III
2
2
0
0
IDC/ILC
1
0
0
0
Lob/Tub
1
0
0
0
Tubular
7
0
0
0
MIB-1
IDC
125
2.14.2 O estrogen R eceptor V ariant / M u tan t Analysis
2.14.2.1 Single Stranded Conformational Polymorphism Analysis
mRNA was isolated from 44 asymptomatic, mammographically detected, primary breast
carcinomas using TRIzol extraction with mRNA enrichment using oligo d(T);5 Dynabeads,
reverse transcribed, and used as template for isotopic PCR. PCR primers were designed to
specifically amplify regions o f the DNA binding and hormone binding domains of the
human oestrogen receptor. Radioactive PCR products were fractionated on non-denaturing
MDE polyacrylamide gels and exposed to autoradiography (Figure 2.21(a-g) for ER Set 2,
Figure 2.22 for ER Set 4, Figure 2.23 for ER Set 5).
SSCP bandshifts were identified in 8 o f the 44 mammographically detected tumours. Seven
were found in the DNA binding domain when RT product were used as template (Figure
2.21. numbers 4, 8, 10, 22, 40, 51, 52) but bandshifts were only seen for 1 of these 7 cases
when nested PCR was used (Figure 2.21(g), number 52). An SSCP bandshift was identified
in 1 of the 44 mammographically detected tumours in the proximal half of the hormone
binding domain (Figure 2.22, number 43) with no bandshifts observed in the proximal half
of the hormone binding domain (Figure 2.23). For all analyses, a known p53 point mutation
(Mismatch Detect Kit) was co-analyzed as a positive control (Figure 2.24).
The nature of the SSCP bandshifts were studied further by the cloning of the shifted bands
and characterization by sequence analysis. Five of the 7 bandshifts detected using ER Set 2
were characterized as having normal sequence (data not shown). Due to constraints of time,
only 1 o f the remaining 2 clones from the DNA binding domain (ER Set 2) was successfully
characterized with the identification o f an A—»G substitution (CAA—>CGA at nucleotide
898 in exon 2). This would theoretically result in an amino acid substitution, exchanging a
hydrophilic glutamine for a basic arganine (Figure 2.25(a)). Sequence characterization o f the
nature of SSCP bandshifts are summarized in Table 2.11.
Sequence characterization o f the bandshift observed in the region screened using ER Set 4
identified 2 point mutations in the same region of the gene. The first was identified as a
T—>C substitution (TTG—>TCG at nucleotide 1282 in exon 4) which would theoretically
result in an amino acid substitution, exchanging a nonpolar leucine for an uncharged polar
serine (Figure 2.25(b)). The second mutation, 53 bases upstream from the first, identified an
A—>T substitution (ATA—>TTA at nucleotide 1336 in exon 4) which would theoretically
126
F igu re 2.2 1 (a -e) Single stranded conform ational polym orphism analysis o f ER S et 2, j5S labeled, PCR
products using Dyneabead oligo d(T)25 m RNA isolated from asym ptom atic breast carcinomas as template.
Tumour products were heat denatured, allow ed to co o l and co-fractionated on 0.5 x non denaturing MDE gels
alongside an ER cD N A control (not shown). Bandshifts w ere observed for asym ptom atic carcinomas 4, 8, 22,
40, 51, and 52 (indicated by an asterix).
F igu re 2 .2 1 (f-g ) Single stranded conform ational polym orphism analysis o f ER S et 2, 35S labeled, nested PCR
products. First round PCR was carried out using primers 6 7 1 /1 3 1 6 with Dyneabead o lig o dT(25) mRNA
isolated from asymptomatic breast carcinom as as tem plate. Second round 35S labeled PCR was carried out
using nested ER Set 2 primers. Products were fractionated on 0.5 x non denaturing MDE gels. A bandshift was
only observed for asymptomatic carcinoma 52 (as indicated by an asterix).
4.
5.
6.
8.
9.
10. 11.
Fig 2.21(a)
12. 13.
14.
15. 17. 18. 19.20.
Fig 2. 21(b)
22. 23. 24. 25. 26. 27. 28.
Fig 2.21(c)
36. 37. 38. 39. 40. 41. 42. 43.
Fig 2.21(d)
51. 52. 53. 54. 55.
.
goi
9 r-* i0 :
(flH fijf-'';
Fig 2.21(e)
<
Q
10.11. g
41.42. 43. w 44. 51.52.53.
12. 13.14. 15.17.
T
Fig 2.21(f)
Fig 2 .2 1(g)
127
Figure 2.22 Single stranded conformational polymorphism analysis o f ER Set 4 35S labeled
o f Dynabead oligo d(T)25 isolated m RNA from frozen, asymptomatic breast carcinomas.
products were heat denatured, allow ed to cool and co-fractionated on 0.5 X non
polyacrylamide gels alongside an ER cD N A control. A bandshift was only observed
carcinoma 43.
<N
(N
m
rf
(S M
RT-PCR products
Tumour RT-PCR
denaturing MDE
for asymptomatic
vo
(N (N
Figure 2.23 Single stranded conform ational polymorphism analysis o f ER Set 5 35S labeled RT-PCR products
o f Dynabead oligo d(T)25 isolated m R N A from frozen, asymptomatic breast carcinomas. Tumour RT-PCR
products were heat denatured, allow ed to cool and co-fractionated on 0.7 X non denaturing MDE
polyacrylam ide gels. Additional am plification products below the expected products were also observed.
Figure 2.24 Single stranded polymorphism analysis o f p53 positive and negative mutation controls. 35S
labeled PCR products o f w t and mt plasmid controls were heat denatured, allowed to cool and fractionated on
non denaturing M DE polyacrylam ide gels alongside ER test samples using ER Set 2, ER S et 4 and ER S et 5.
128
Figure 2.25(a) Sequence analysis o f
asymptomatic 8 ER Set 2 SSCP bandshift
identifying an A->G substitution
Figure 2.25(b) Sequence analysis o f
asymptomatic 43 ER Set 4 SSCP bandshift
identifying an T->C substitution
Figure 2.25(c) Sequence analysis o f
asymptomatic 43 ER Set 4 SSCP bandshift
identifying an A ->T substitution
result in an amino acid substitution, exchanging a nonpolar isoleucine for an nonpolar
leucine (Figure 2.25(c)).
Isotopic amplification of some cases also revealed additional amplification products which
demonstrated gross altered migration to wt products when fractionated on non denaturing
SSCP gels, o f which a representative example is given in Figure 2.23. The presence of
additional products was therefore assessed by nested PCR amplification using Dynabead
oligo d(T)25 isolated mRNA reverse transcription product as template, and primer set
739/1384 with ER Set 2 nested, and primer set 1160/2092 with ER Set 4 and ER Set 5
nested. Products were resolved on 2% agarose gels, stained with ethidium bromide and
visualized under UV illumination (Figure 2.26(a-c)).
Four o f the 44 asymptomatic breast carcinomas exhibited additional, lower molecular
weight products o f approximately 270 bp when analyzed at the ER DNA binding domain
using first round 739/1384 RT-PCR product as template for nested ER Set 2 primers (Fig
2.26(a) asymptomatic carcinomas 37, 43, 51, and 54). All cases exhibited the product in
addition to wt except asymptomatic 43 which only exhibited the lower molecular weight
product.
Thirteen o f the 44 asymptomatic breast carcinomas exhibited additional, lower molecular
weight products o f approximately 270 to 310 bp, when analyzed at the ER hormone binding
domain using first round 1160/2092 RT-PCR product as template for nested ER Set 4 (Fig
2.26(b)) and ER Set 5 (Fig 2.26(c)) primers sets.
Three asymptomatic carcinomas demonstrated additional products using ER Set 4, 2 of
lower molecular weight, approximately 310 bp in size (Fig 2.26(b) asymptomatic
carcinomas 14 and 56), and 1 o f higher molecular weight, approximately 550-600 bp in size
(Fig 2.26(b) asymptomatic carcinoma 21) representing a possible exon duplication. Each
cases exhibiting the additional amplification product coexpressed wt product and all
expressed the additional product in the distal half of the hormone binding domain.
Eleven asymptomatic carcinomas demonstrated additional products using ER Set 5, all of
lower molecular weight, approximately 310 bp in size (Fig 2.26(c) asymptomatic
carcinomas 14, 21, 22, 25, 26, 31, 36, 37, 44, 51, 56). All cases exhibited additional product
130
- o - mr o r*i
r<)
m
r*lr»"> <—
i m Tj-
m
603 - m
~t
-rt
Tt-
ir> ir>
</~>
mm
mm m
3 1 0 _ B
1/-1
Figure 2.26(a) Agarose gel analysis o f nested RT-PCR analysis from mRNA isolated by Dynabead oligo
d(T)25 from asymptomatic breast carcinomas. PCR was carried out on reverse transcribed m RNA isolated
from frozen, asymptomatic breast carcinomas using primers 739/1384 and nested primers ER Set 2. Products
were fractionated on 2% agarose gels and stained with ethidium bromide. Upper arrow refers to w t product
with lower arrow referring to additional amplification product. Cases demonstrating additional products
indicated by an astrix.
<
X rf
c4 </"> 'O >—r o v i v O r - ' — tJ- — vO
-e- — 0 4 <N <N <N m
W
Figure 2.26(b) Agarose gel analysis o f nested RT-PCR analysis from mRNA isolated by Dynabead oligo
d(T)25 from asymptomatic breast carcinomas. PCR was carried out on reverse transcribed m RN A isolated
from frozen, asymptomatic breast carcinomas using primers 1160/2092 and nested primers ER S et 4. Products
were fractionated on 2% agarose gels and stained with ethidium bromide. The centrally located arrow refers to
wt product with lower and upper arrows referring to additional amplification products. Cases demonstrating
additional products indicated by an astrix.
*
•
*
•
*
*
•
*
*
*
*
*
X Tf « cn io s d - - f i v i s d r ' ' —' ^
<NCS C
N <N ro n n n n if
*
*
\o
>n v-)
Figure 2.26(c) Agarose gel analysis o f nested RT-PCR analysis from mRNA isolated by Dynabead oligo
d(T)25 from asymptomatic breast carcinomas. PCR was carried out on reverse transcribed m RNA isolated
from frozen, asymptomatic breast carcinomas using primers 1160/2092 and nested primers ER S et 5. Products
were fractionated on 2% agarose gels and stained with ethidium bromide. The upper arrow refers to w t product
with the lower arrow referring to additional amplification product. Cases demonstrating additional products
indicated by an astrix.
131
in addition to wt except asymptomatics 25 and 26 which only exhibited the lower molecular
weight product. Five asymptomatic carcinomas exhibited additional product at more than
one loci, 3 with both ER Set 4 and ER Set 5 (asymptomatics 14, 21, and 56) and 2 with ER
Set 2 and ER Set 5 (asymptomatics 37 and 51).
Aberrant amplification products showing gross molecular weight alterations were isolated as
described in section (2.12.2.10), subcloned into pCR-Script Amp SK+ cloning vector,
transfected into Epicurian Coli XL-1 Blue MRF’ Kan supercompetant cell, cultured
overnight,
plasmids
isolated
and
characterized
by
dideoxynucleotide
sequencing.
Representative examples of variant species characterized are shown in Figure 2.27(a-d).
Sequence characterization o f aberrant amplification products are summarized in Table 2.12.
Aberrant species isolated from asymptomatic carcinomas amplified with ER Set 2 were
found to contain wt sequence for exons 2 and 4 with exon 3 perfectly deleted (Figure 2.27(a)
asymptomatic carcinoma 51). This deletion of 117 bp of exon 3 would theoretically result in
an ER protein molecule 39 amino acids smaller than a wt molecule, i.e., a 556 amino acid
molecule.
Sequence characterization of aberrant species, of lower molecular weight with respect to wt,
isolated from asymptomatic carcinomas amplified with ER Set 4, identified a variant species
which contained wt sequence for exon 4 up to nucleotide 1230 followed by wt exon 6
sequence which was preceded by a 27 nucleotide insertion (Fig 2.27(b) asymptomatic
carcinoma 56). This alteration would theoretically introduce a frameshift in the reading
frame, introducing a TAA stop codon at nucleotide 1722 resulting in a prematurely
truncated protein molecule containing 340 amino acid residues.
Sequence characterization o f aberrant species, of lower molecular weight with respect to wt,
isolated from asymptomatic carcinomas amplified with ER Set 5, identified 2 types of
variant species. The first type contained wt sequence for exon 6 and 8 with exon 7 perfectly
deleted (Fig 2.27(c) asymptomatic carcinoma 31). This alteration would theoretically
introduce a shift in the reading frame, introducing a TGA stop codon at nucleotide 1943
giving rise to a prematurely truncated protein molecule containing 466 amino acid residues.
132
Exon 4
Exon 2
5’
A
A
A
A
G
C
A
T
A
G
G
G A T C
T.r;
G
A
A
C
T
T
A
T
F igure 2.27(a) Sequencing analysis o f aberrant asymptomatic 51 ER Set 2 nested PCR product cloned into
pCR-Script Amp SK+ cloning vector identifying an in frame exon 3 deletion.
F igure 2.27(b) Sequencing analysis o f asymptomatic 56 ER Set 4 nested PCR product cloned into pCR-Script
Amp SK+ cloning vector identifying a partial deletion o f exon 4, the insertion o f an additional 27 nucleotides
and com plete deletion o f exon 5.
133
5’
G
G
A
Exon 8
GA T C
C
A
T
_
| m*
^
**•
«K
/“ ^ *
^
G
cI ’ m
T
T
A
A
T
T
C
3’
Exon 6
**
wv
F igure 2.27(c) Sequencing analysis o f asymptomatic 31 ER Set 5 nested PCR product cloned into pCR-Script
A m p SK+ cloning vector identifying an in frame deletion o f exon 7.
5’
T
£
G A
T C
deleted T^>
T
T
T
3’
F igure 2.27(d ) Sequencing analysis o f asymptomatic 26 ER Set 5 nested PCR product cloned into pCR-Script
Amp SK+ cloning vector identifying an com plete deletion o f exon 7. In addition, a T->C bp substitution was
observed 5 nucleotides from the exon 7 splice site, with the first nucleotide o f exon 8 deleted.
134
The second type o f alteration observed contained wt sequence for exon 6 with a T->C bp
substitution 5 bases from the normal splice site and all of wt exon 8 minus the first T (Fig
2.27(d) asymptomatic carcinoma 26). This alteration would theoretically introduce a shift in
the reading frame, introducing a TGA stop codon at nucleotide 1970 giving rise to a
prematurely truncated protein molecule containing 471 amino acid residues.
Table 2.12 Summary sequence data for oestrogen receptor SSCP bandshifts and PCR
variants
No. Cases
Sequence Data
SSC P Bandshifts
DNA Binding Domain
Hormone Binding Domain
1
No sequence data
1
A->G at 898 in exon 2
1
T->C at 1282 in exon 4 with an
A->T at 1336 in exon 4 *
PCR Variants
DNA Binding Domain
2
Deletion of exon 3
Distal Hormone Binding
Domain
8
i) Deletion of exon 7
ii) Deletion of exon 7 with T->C 5
bases from exon 6/7 splice site and
lacking first T exon 8
Proximal and Distal Hormone
Binding Domains
i) Deletion of exon 5
3
ii) Partial deletion of exon 4 with
complete deletion of exon 5 and
insertion of 27 nucleotides
iii) possible exon duplication
DNA and Distal Hormone
Binding
2
Key * A variant was also detected in DNA binding domain of this case
135
2.14.2.2 Mismatch Detect II Analysis - Non Isotopic RNAse Cleavage Assay (NIRCA)
An additional mutational screening method known as NIRCA (Non Isotopic RNAse
Cleavage Assay) was attempted. The method is based on that described by Myers et al.,
(1985) and Winter et al, (1985) which relies on the ability of RNase A to cleave single
unpaired bases (mismatches) in an RNA probe hybridized to an experimental target
containing a mutation.
cDNAs synthesized from RNA isolated from 44 asymptomatic, with mRNA enrichment
using oligo d(T)25 Dynabeads, were reverse transcribed, and used as template for PCR using
ER Set 6 (ER DBD) and ER Set 7 (ER HBD). Each primer set contained opposable T7 and
SP6 phage promoters which were incorporated into the amplification product. Resultant
PCR products were used as templates for in vitro transcription of sense and antisense strands
o f experimental (asymptomatic carcinomas) and wild type PCR products. ER cDNA was
used as source o f wt for ER Set 6 with endogenous wt sequence expressed within carcinomas
as a source for ER Set 7). After heat denaturation, hybridization of sense and antisense
experimental RNA transcripts with complementary wt transcripts produced RNA duplexes
targets for RNase digestion, the products o f which were analyzed on 2% agarose gels
(Figure 2.28(a-c)).
Figure 2.28(a) shows representative NIRCA analysis of p53 controls supplied with the kit in
which hybridization of the following transcripts combinations are digested with different
combinations of RNAse solutions (#1, #2, #3): i) wt T7 and wt SP6 (no mismatch control) in
lane 1s; ii) mt T7 and wt SP6 (mismatch control) in lane 2s; iii) wt T7 mt SP6 (mismatch
control) in lane 3s; mt T7 and mt SP6 (mismatch control) in lane 4. Digestion products were
observed in both mismatch and no mismatch controls reactions for all RNAse combinations.
In contrast, no digestion products were observed for asymptomatic carcinoma NIRCA
analysis at the DNA binding domain (representative analysis using RNAse digestion system
#2 given in Figure 2.28(b) lanes 1-11) whereas numerous digestion products / smears were
observed for analysis at the hormone binding domain (representative analysis using RNAse
digestion system #1 given in Figure 2.28(c) lanes 1-10).
136
F igu re 2.28(a) Representative NIRCA analysis o f p53 controls with digestion using RNAse systems #1, #2,
and #3 on the follow ing T7/SP6 transcript duplexes: lane Is - wt T7 and w t SP6 (no mismatch control); lane 2s
- m t T7 and wt SP6 (mismatch control); lane 3s wt T7 mt SP6 (mismatch control); lane 4s - mt T7 and mt SP6
(m ism atch control). Products o f RN Ase digestion can be seen in both mismatch and no mismatch controls for
all R N A se system s. Molecular weight marker used was a lOObp ladder.
F igu re 2.28(b ) Representative NIRCA analysis o f asymptomatic carcinomas with digestion using RNAse
system #2 on ER S et 6 generated wt T7 and test SP6 transcript duplexes. N o observable digestion products
w ere observed.
£
S
1.
2.
3.
4.
5.
6.
7.
8.
9. 10.
F igure 2.28(c) Representative NIRCA analysis o f asymptomatic carcinomas with digestion using RNAse
system #1 on heterogeneous wt and mt ER Set 7 generated T7 and SP6 transcript duplexes. In contrast to ER
S et 6 analysis multiple products o f digestion and smears were observed. Molecular weight marker used was a
1OObp ladder
137
2.14.3 Correlation Of Variant Oestrogen Receptor Species With Pathological Features
The pathological data of mammographically detected carcinomas demonstrating altered
oestrogen receptor transcripts are shown in Table 2.13.
Altered transcripts were detected in fifteen o f the 36 infiltrating ductal carcinomas, and 2 of
the 8 tubular and tubular/lobular carcinomas. The presence of an alteration was considered
with respect to grade for infiltrating ductal carcinomas (summarized in Table 2.13). Four of
the grade I carcinomas (well differentiated, 4 of 10 (40%), had alterations: 1 with a mutation
in the DBD, 2 with exon 7 deletions in the distal HBD and 1 with an exon 3 deletion in
DBD and exon 7 deletion in distal HBD. Six of the grade II carcinomas (moderately
differentiated, 6 o f 19 (31.6%) had alterations: 1 with a mutation in the DBD, 1 with an
exon 3 deletion in DBD, 2 with exon 7 deletion in distal HBD, 1 with an exon 3 deletion in
DBD and exon 7 deletion in distal HBD, and 1 with exon 5 and 7 deletions in proximal and
distal HBD. Five of the 7 grade III carcinomas (poorly differentiated, 71.4%) had
alterations: 4 with exon 7 deletions in distal HBD, and 1 with exon 5 and 7 deletion in both
proximal and distal HBD respectively. When the lobular/tubular and tubular carcinomas
were included with the grade I infiltrating ductal carcinomas the incidence o f alterations was
6/18 (33%). O f the two tubular carcinomas, 1 had an exon 3 deletion in DBD plus mutations
in the proximal HBD, the other had an exon 7 deletion in distal HBD with a possible exon
duplication in proximal the HBD.
Considering presence o f mutants/variants with steroid receptor expression (summarized in
Table 2.13), 21 one o f the 44 carcinomas expressed both oestrogen and progesterone
receptor, and 6 o f these had altered ER transcripts (28.6%), 3 involving the distal HBD, 2
involving the DBD and distal HBD and 1 involving both proximal and distal HBD. Eighteen
o f the 44 carcinomas were ER+/PgR- and 7 of these had altered transcripts (38.9%). Four
showed alterations to the DBD (1 mutation, 2 exon 3 deletions 1 with mutation in proximal
HBD). Two had alterations in distal HBD (exon 7 deletions) and 1 deletion in both proximal
and distal HBD (exon 5 and 7 deletions respectively). Four of the 5 ER-/PgR- carcinomas
had altered transcript (80%). Three had exon 7 deletions in distal HBD and 1 with exon 5
and exon 7 deletions in distal and proximal HBD respectively.
O f the 17 carcinomas demonstrating altered ER transcripts, all but 4 coexpressed wt
transcript (carcinoma 43 (tubular carcinoma) - exon 3 deletion, and carcinomas 25 (ID II),
138
T a b le 2 .1 3 Correlation o f presence of altered oestrogen receptor transcripts in relation to
pathological data of mammographically detected carcinomas
Tumour
No.
8
Type
Grade
Size
ER
PgR
MIB-1
DBD
IDC
II
15
187.5
neg
12.5
*
14
IDC
II
15
190.5
neg
21
Tub
I
15
151
22
IDC
II
15
25
IDC
II
26
IDC
31
Proximal
HBD
Distal
HBD
27.3
•
•
123
4.8
•
•
123
neg
2.3
•
13
130
62
6.6
•
III
11
176
neg
17.6
•
IDC
I
15
210
180
5.5
•
IDC
III
15
neg
neg
60
•
35
IDC
III
15
neg
neg
80
•
36
IDC
III
15
neg
neg
1
•
37
IDC
I
13
225
190
8
•
43
Tub
I
12
171
neg
10
•
44
IDC
I
15
132
169
12.2
51
IDC
II
10
232
62
1.4
•
52
IDC
I
13
136
10
1
*
54
IDC
II
14
130
15
5
•
56
IDC
III
11
neg
neg
30
•
*
•
•
•
•
Key IDC (infiltrating ductal carcinoma), Tub (tubular carcinoma), Lob/Tub (lobular and
tubular carcinoma), neg (negative), ER (oestrogen receptor), PgR (progesterone receptor),
DBD (DNA binding domain), HBD (hormone binding domain).# indicates splice variant
detected, * indicates mutation detected.
139
Table 2.14 Correlation of the presence of altered oestrogen receptor species with grade and steroid receptor expression
Total
Mutation
DBD (exon 3)
1 DBD, 1 HBD
1 DBD
1 plus mutation in
proximal HBD
1
Grade I
18
No. with
alteration
6
Grade II
19
6
Grade III
7
5
ER+/PgR+
21
6
1 HBD
ER+/PgR-
18
7
2 DBD
ER-/PgR-
5
4
Distal HBD
(exon 5)
2
DBD and Distal
HBD (exons 3 and 7)
1
Proximal and Distal
HBD (exons 5 and 7)
1
2
1
1
1
4
3
1 plus mutation in
proximal HBD
3
2
1
2
1
3
1
Key DBD (DNA binding domain), HBD (hormone binding domain), ER (oestrogen receptor), PgR (progesterone receptor), + indicates positive, indicates negative
26 (IDC III) and 31 (IDC I) - exon 7 deletion). O f the remaining 13 carcinomas there was
only 1 which coexpressed wt transcript, at a level of altered transcript greater than that of wt
(carcinoma 37 (IDC I) - exon 7 deletion).
Six o f the 17 carcinomas demonstrated expression of multiple transcripts. Two carcinomas,
carcinomas 14, and 56 (IDC grade II and III respectively), coexpressed exon 5 and exon 7
deletion transcripts; 2 carcinomas, carcinomas 37 and 51 (IDC I and II respectively),
coexpressed exon 3 and exon 7 deletion transcripts. O f the 2 remaining carcinomas (both
tubulars), 1 coexpressed an exon 3 deletion transcript and a transcript containing a point
mutation in the DBD (carcinoma 43), and 1 coexpressed an exon 7 deletion in addition to a
possible exon duplication in the proximal half of the HBD (carcinoma 21).
O f the 6 carcinomas demonstrating multiple alterations, i.e., more than one splice variant
and/or a point mutation, 4 had either a proliferation rate above the mean (cases 14 and 56) or
a relatively high ER H score o f above 150 (Cases 14, 21 and 51).
2.14.4 Oestrogen Receptor CpG Island Methylation Analysis
To determine patterns o f ER CpG island methylation in the mammographically detected,
asymptomatic breast carcinomas group, DNA from these tumours was studied using a PCR
based method. As the ER CpG island contains a methylation sensitive Hhal restriction site
in the vicinity o f a polymorphic TAn microsatellite repeat sequence, it is possible to assess
the methylation status of heterozygous alleles by PCR (method depicted in Figure 2.29(a)).
Prior to PCR amplification, MCF-7 or T47-D and MDA-MB-231 cell line DNA was treated
with and without Hhal. Amplification with primers ESR1 and ESR 2 (primers 1 and 2
respectively in Figure 2.29(a)) ensured equivalent template concentration in the PCR
reaction with amplification using ESR1 and ER Meth 2 (primer 3 in Figure 2.29(a)) to assess
methylation status. In this analysis, unmethylated DNA results in amplification of products
for both ESR 1+2 and ESR1/ER Meth 2 (differing in size by approximately 250-300 bp),
while methylated DNA results in product for ESR 1+2 only.
PCR analysis o f MDA MB 231, and T47-D breast cancer cell line DNA resulted in
amplification products for both Hhal restricted and unrestricted DNAs from both cell lines,
for ESR 1+2 and ESR1/ER Meth 2 primer sets (Figure 2.29(b) and Figure 2.29(c)
respectively) when fractionated on agarose gels. 35S PCR analysis of MDA MB 231 and
141
MCF-7 cell line and asymptomatic carcinoma DNA resulted in amplification, again for
restricted and unrestricted DNAs but only for primer set ESR 1+2 (Figure 2.29(d) and
Figure 2.29(e)) with no observable product for ESR1/ER Meth 2 when fractionated on 6%
PAGE. Due to the constraints of time it was not possible to extend this work further.
3
H hal
y
TA„
2
U nm ethylated
M ethylated
Hha I Digestion
Digestion
No Digestion
PCR
Polyacrylam ide Gel
Electrophoresis
Figure 2.29(a) Oestrogen receptor CpG island methylation PCR method. Malignant tissue D N A was treated
with H hal methylation sensitive restriction endonuclease prior to multiplex PCR amplification with primers
ERS 1 + 2 and ER Meth 2. After PCR, amplification products were fractionated on 6% PAGE. A s depicted,
methylated CpG island are restricted resulting in ESR 1+2 amplification only. Unmethylated CpG islands are
not restricted resulting in ESR 1+ 2 and ESR1/ER Meth 2 amplification.
Tf
r><
+
+
—•
m
<N
Q
t^>
rt
h
—
en
<n
e'­
h
m
603-
m m
194- 1
#-*4f19
Figure 2.29(b ) Agarose gel analysis o f oestrogen receptor CpG island methylation PCR products using ESR
1+ 2 PCR primers. M DA MB 231 and T47-D breast cancer cell line DNA was treated with (+) and without (-)
H hal methylation sensitive restriction endonuclease prior to PCR amplification. Products were observed for
both cell lines, with and without Hhal restriction.
142
Figure 2.29(c) Agarose gel analysis o f oestrogen receptor CpG island methylation PCR products using ESR
1/ER Meth 2 PCR primers. M DA MB 231 and MCF-7 breast cancer cell line DNA was treated with (+) and
without (-) H hal methylation sensitive restriction endonuclease prior to PCR amplification. Products were
observed for both cell lines, with and without H hal restriction.
+
m
<N
CQ
i
CQ
s
<
Q
m
<N
N
<
Q
CQ
+
q
f iTf
2
1
q
•*3H
<
Q
Figure 2.29(d) Polyacrylamide gel analysis o f oestrogen receptor CpG island methylation PCR products using
ESR 1+2 PCR primers. MDA MB 231 and T 47-D breast cancer cell line DNA was treated with and without
35
H h al methylation sensitive restriction endonuclease prior to S PCR amplification. Products were observed
for both cell lines, with (+) and without (-) H h a l restriction.
Figure 2.29(e) Polyacrylamide gel analysis o f oestrogen receptor CpG island methylation PCR products using
ESR 1+2 PCR primers. Asym ptomatic carcinoma D N A s (1-5) was treated with and without H hal methylation
sensitive restriction endonuclease prior to 35S PCR amplification. Products were observed for both restricted
(+) and unrestricted (-) templates for hll cases analyzed.
143
2.15 Discussion
2.15.1 Technical Problems
A number o f technical difficulties were encountered during this thesis, with key
methodologies, in particular with reverse transcriptase-polymerase chain reaction (RTPCR), single stranded conformational polymorphism analysis (SSCP), Non-Isotopic RNase
Cleavage Assay (NIRCA) and ER promoter methylation analysis.
The main problems encountered with RT-PCR fell into 2 main categories, i) negative
controls and ii) sensitivity. For RT-PCR a negative control consisting of all the constituents
for reverse transcription, without the addition of reverse transcriptase, was included.
Therefore no cDNA would have been available for primer annealing, and consequently for
amplification during subsequent PCR. Nonetheless, in the early stages o f the project, this
control was sometimes found to amplify for both ER and actin primers as efficiently as other
samples which had reverse transcriptase enzyme included, casting doubt on the validity of
sample amplification. If contaminating genomic DNA was present in the original RNA
isolation then this could be utilized as template for amplification. Whilst an attractive
explanation for the actin primers, which both annealed within a single exon, this could not
account for amplification with ER Set 2 which had been designed to anneal to individual
exons (exons 2 and 3). These primers flanked a greater than 19 kb intron (Ponglikitmongkol
et al, 1988) and consequently would yield a product far greater in molecular weight than the
correctly sized product that was observed. A more likely explanation was that aerosol
contamination might be responsible, with carry over of enzyme from one reaction to another
via a pipette. To minimize this, additional precautions were undertaken, including,
pretreatment of all pipettes and consumables, i.e., tips and eppendorfs, by repeated UV
irradiation for approximately 30 minutes and with all manipulations being carried out in a
class II cabinet using filtered pipette tips. Using this approach it was possible to eliminate
the inappropriate amplification of this reverse transcription negative control and so
confirmed the validity o f amplification seen with the study group.
The second problem encountered, was lack of detection of RT-PCR products by ethidium
bromide staining from very limited amounts of tissue samples. As a consequence an isotopic
PCR approach was utilized (35S). However, there are drawbacks with this approach in terms
of cost, and safety issues with the isotope. In addition, products are analyzed by denaturing
PAGE and autoradiography increasing the time and cost of an otherwise straightforward
144
agarose gel analysis. The problem was mainly due low RNA yields from very small
amounts o f carcinoma tissue for RT-PCR (3 x lOpM sections) since tissue was required for
other studies within the laboratory. A majority of studies screening for ER variants/mutants
utilized larger amounts o f RNA (l-10pg) as template for reverse transcription (Daffada et
al, 1994, 1995; Castles et a l, 1995; Gotteland et al, 1995; Villa et a l, 1995; Zhang et al.,
1996; Desai et a l, 1997), alternative methodologies, e.g., RNase protection (Hirata et al,
1995), isoelectric focusing (Marsigliante et al, 1996), western blotting (Pink et al, 1995) or
genomic DNA (Roodi et al, 1995) although Leygue et al, (1996) analyzed 600ng total
RNA for their analysis using 40 cycles of amplification. In order to overcome this problem 2
additional steps were undertaken. Firstly mRNA enrichment by poly A selection using oligo
d(T)25 paramagnetic beads which other studies have demonstrated enable the amplification
o f RNA transcripts from small amounts of tissue (Bicknell et al, 1996). Secondly a nested
PCR approach was adopted which enabled analysis by simple ethidium bromide staining.
The initial problems encountered optimizing the SSCP technique, resulted from the lack of
available appropriate positive (mutation containing) and negative (wt) controls. Many
variables affect the ability to resolve mutations by this method, including temperature and
gel composition, e.g., with or without urea or formamide, which affect the stability of the
artificial secondary structures formed (Glavac and Dean, 1993). By excluding appropriate
controls, it was not possible to determine whether a mutation was or was not present or
whether conditions were not optimal for mutation detection. p53 controls supplied with the
NIRCA kit were used which provided confidence as to validity of results. In addition, the
use o f MDE polyacrylamide gel matrix simplified the analysis as the matrix contained a
polymer which stabilizes single stranded secondary structures such as hair pin loops thereby
leaving the gel concentration (% acrylamide) the only variable to be empirically determined.
The development o f a non-isotopic SSCP method would be desirable in terms o f safety and
cost as already described above. Unfortunately the main drawback for this type of approach
is the lack o f sensitivity. Silver staining gels can overcome this problem but is a relatively
laborious method compared to other detection methods such as ethidium bromide staining.
To some degree, sensitive SSCP using MDE gel was achieved using the mRNA selection,
nested RT-PCR approach already discussed. Although the results obtained would not be
suitable for publication purposes, it could be useful as an initial screen as the results can be
easily observed by the naked eye. Confirmatory analysis could then be achieved to a
145
publication standard using an isotopic incorporation thereby reducing the amount of isotope
used.
Another non isotopic method with which I had difficulties with was the NIRCA analysis.
Unlike the non isotopic SSCP method, results obtained using this method should be easily
observed and o f publication quality due to the production of high levels of in vitro
transcripts. Unfortunately, mutation analysis using this method was unsuccessful in my
hands. This was predominantly due to two main problems. The first was that the technique
is primarily designed for the detection of single nucleotide mutations and not splice variants.
During the hybridization step, numerous duplexes may form which when digested with
RNases result in many digestion products visualized on agarose gels which are difficult to
interpret.
The second problem encountered was due to the different sensitivity of unpaired nucleotides
to RNase digestion. In order to determine the presence of a mutation, many combinations of
RNAse solutions must be used with the optimum combinations determined empirically for
each sample. Unfortunately, due to lack of time, a full evaluation of the conditions to detect
mutations was not performed, and therefore no information using this method was obtained.
DNA methylation at the 5’ position o f cytosine residues is an important mechanism which
may regulate gene expression (Cedar and Razin, 1990) with an inverse relationship between
methylation and expression being demonstrated for many genes, indicating an important
role for gene methylation in repression of transcription (Doefler, 1983). A role for DNA
methylation in cancer has been suggested with amongst others Goelz et al., (1985)
demonstrating that DNA from both benign polyps and carcinomas are hypomethylated
compared with normal colon, indicating that altered DNA methylation could be a key event
in the development and progression o f cancer. Most methylation studies utilize enzymatic
digestion of genomic DNA followed by Southern analysis as this also allows the
methylation status o f each allele to be assessed. The position of a polymorphic microsatellite
within the promoter o f the ER could permit a rapid, inexpensive evaluation o f the
methylation status o f the ER gene on individual alleles in a large number of cases.
Oestrogen receptor promoter methylation analysis was unsuccessful in this study, due
primarily to lack o f available time to fully evaluate the method. The basic idea remains
146
fundamentally sound and potentially useful in as much that it utilizes well established
methodologies (DNA extraction, enzymatic digestion, and PCR) that are sensitive, quick,
relatively inexpensive and permit the assessment as to the methylation status of individual
alleles. From the outset the experimental plan was fundamentally flawed due to the
assumption that the methylation status of the ER gene promoter in the cell lines used as
positive and negative controls, were as described in the literature, e.g., MDA-MB-231
methylated (Lapidus et al.} 1996). With hindsight, evaluation of ER status of these cell lines
by northern, western or immunocytochemical analysis would have been advantageous. In
addition, digestion of genomic DNA from these cell lines, at methylation sensitive
restriction endonuclease sites within the ER 5’ UTR (Figure 2.30), followed by Southern
analysis with radiolabelled ER probe (e.g., Pvull/EcoRl fragment from pOR3) would have
allowed the methylation status of these cell lines to be assessed. Alternatively, treatment of a
methylated ER promoter cell line by the demethylating agent 5-aza-2’-cytosine could
provide controls if the appropriate cell lines were not available. In this way an evaluation as
to the usefulness of this approach could have been made, e.g., is restriction at one Hhal
flanked by two PCR primer annealing sites sufficient to prevent PCR amplification?
0.3 kb pOR3
1
0o4
CO
■a
I
#
n
I
exon 1
a
o4
o
3
a*
I
#
1111
I
Figure 2.30 Restriction map o f the human ER gene promoter (Taken from Lapidus et al., 1996). Methylationinsensitive flanking cut is £coR I, m ethylation-sensitive enzym es are N otl, SacII, Hhal, H pall.
With respect to analysis on clinical tissue, I was unable to obtain the larger of the two PCR
fragments, to be generated by ER Meth 2 and ESR1. Genomic DNA was extracted using the
phenol phase after Trizol reagent extraction of RNA from the mammographically detected
breast carcinomas. Unfortunately, DNA extractions were not carried out immediately after
the RNA extractions and consequently the phenol phases containing the DNAs were stored
147
at 4°C for approximately 1 year. Long term storage of DNA in phenol would most likely
result in the nicking of DNA thereby making amplification of large fragments difficult, if
not impossible.
2.15.2 Interpretation O f Data
An abundance of variant oestrogen receptor mRNAs including those harbouring mutations,
deletions, insertions and aberrant splicing have been characterized in breast tumours and
breast cancers cells lines (reviewed in section 2.7) with multiple spliced variants frequently
observed in individual breast tumours. Many of these studies have focused on well
established, breast tumours including those which have distant metastasis, with little
attention to the earlier stages o f the disease, (e.g., <2cm, node negative). Little is known
about the presence and nature o f variants/mutant ERs in the earlier stages. If there
are differences from the later stage carcinomas this could provide information about
divergence of pathways and increase our knowledge regarding the natural history of the
disease. It was therefore the aim o f this study to screen a large group of well to moderately
differentiated breast carcinomas, which were node negative, 15mm or less in size, and
detected by the first round o f screening mammography for the expression of variant/mutant
ERs and to correlate their presence with the pathology of these tumours.
Multiple ER variants in breast carcinomas were detected in this study in agreement with the
frequent expression o f multiple variants reported in other published studies. These included
both point mutation variants and variants with gross molecular weight abnormalities. These
were characterized by sequence analysis and found to correspond to aberrantly or
alternatively spliced transcripts precisely and imprecisely lacking particular exons (notably
3, 5 and 7 - translation products summarized in Figure 2.31). Due to the constraints o f time,
not all aberrant species were characterized. Given that these cases comigrated with
characterized cases on SSCP gels, it is reasonable to assume that they were probably the
same type o f variant. Nonetheless, for publication purposes these also need to be confirmed.
The observed ER exon variants were present in 15 of the 44 (34.1%) carcinomas examined
(9.1% (4 of 44) exon 3, 4.5% (2 of 44) exon 5 and 29.5% (13 of 44) exon 7 deletions), with
6 carcinomas showing multiple variants. Most also contained the corresponding wt
sequence. This analysis focused on screening for the 4 key variants (exons 2, 3, 5, and 7) as
these have been shown to have profound effects on the DNA and hormone binding domains.
148
UGA
AUG
TT
X
Ex 2 Ex 3
rsc
O
C
r-
o
sO
•'f
Ex 5 Ex 6
Ex 7
/
/
}I
/ /
Ex S
(N
<
N
r*~.
sC
/ /
A
B
C
r*.
302
OKI
1
wt Protein
(aa)
Ex 4
263
wt mRNA
(nt)
Ex 1
r-~
oc
oc
r<
nCS
ir.
E
D
is
^*,
595 aa
IF
Protein generated
by deletion of.
Exon 3
(in frame)
556 aa
214s* /
Exon 5
(out of frame)
Exon 5
(out of frame)
595
371 aa
365 ^
(ziT ) ( z T )
(6 a a )
(27 aa)
340 aa
289
—
Exon 7
(out of frame)
(42 aa)
466 aa
456
Exon 7
(out of frame)
(10 aa)
417 aa
350 (6 aa)
F igu re 2.31 Schematic representation o f p u tative E R -sp lice variants. Top to bonom: n t E R -specific m R N A
m apping the position o f the various exon s, fo llo w e d by a schem atic representation o f die functional dom ains
o f the wt protein and the predicted structure o f the various deleted proteins. Functional regions o f the w t ER
(595 aa) include the amino-terminal dom ain A /B . the D N A binding domain C with the two zinc fingers (Z n),
the hinge region D. the hormone binding dom ain E and the carboxy-terminus F (Green and C ham bon, 1991).
The size o f each putative deleted protein is g iv e n on the right.
149
Consequently due to the position of the PCR primers evaluation of exon 4 and 6 variants
was not carried out.
Correlation of the incidence of variants/mutations with pathological data showed that 41.7%
of infiltrating ductal carcinomas and 25% of tubular and tubular/lobular carcinomas
demonstrated these altered transcripts. With respect to grade, a relatively good association
was observed between the number of variants/mutants and increasing grade with 33.3% of
grade I (6/18), 31.6% o f grade II (6/19), and 71.4% of grade III (5/7) carcinomas
demonstrating altered ER transcripts, suggesting that higher grade, more aggressive
carcinomas are more likely to express altered ER transcripts.
Approximately twenty nine percent of the carcinomas which were positive for ER and PgR
(6/21), 38.9% o f ER+/PgR- negative carcinomas (7/18) and 80% of ER-/PgR- carcinomas
(4/5) had altered ER transcripts indicating that loss of steroid receptor could be associated
with the frequent occurrence o f these variants.
Approximately 35% o f carcinomas in the study group (6/17) expressed more than one
altered or mutated ER transcripts and o f these 66.7% (4/6) had proliferation rates above that
of the mean proliferative index indicating that the presence of aberrant ER are possibly
associated with a higher rate o f tumour cell division.
The least common variant detected was the exon 5 deletion variant (4.5% - 2/44) comprising
one in frame and one out o f frame deletion. The perfect deletion of exon 5 resulted in a shift
of the reading frame and the introduction of a stop codon at the seventh position of exon 6.
This premature truncation would give rise to a protein of 40 kD lacking hormone binding
ability and TAF-2 activity, but retaining TAF-1, nuclear localization and intact DNA
binding. Originally exon 5 deletion variants were thought to contribute to the ER-/PgR+
phenotype. No tumours of this steroid receptor phenotype were characterized in this study,
but the exon 5 deletion variant was detected in an ER+/PgR- carcinoma, which is in
agreement with previous studies where
its expression was not confined to ER-/PgR+
tumours but also found in ER positive tumours (Zhang et al., 1993).
In addition, an out of frame exon 5 deletion variant with partial loss of exon 4 and insertion
of 27 nucleotides was identified. This alteration would theoretically result in the
150
introduction of a TAA stop codon at the 43rd position of exon 6, giving rise to a truncated
protein o f 343 amino acids which would hence lack similar elements to the perfect deletion
variant already described. Although the functional activity of this variant is unknown, it is
attractive to postulate that it may have the constitutively active properties associated with
the perfectly deleted exon 5 counterpart.
The insertion o f 27 novel nucleotides into the coding sequence is an interesting finding since
insertion o f bases, especially o f such large numbers, other than exon duplication, seems to
be a rare event with respect to ER variants/mutants (Sluyser, 1995; Murphy et al., 1996).
One explanation could be that mutations to intron and or exon sequences could reveal or
generate cryptic splice sites. These could result in the deletion of exon sequences at these
newly generated splice sites within the affected exon or within intron sequences which result
in the inclusion o f intron sequences in mature ER mRNA. Such a situation has been
demonstrated by Murphy et al. (1996) who identified an ER-like mRNA which contained an
insertion o f 69 novel nucleotides between exon 5 and 6 sequences of normal ER mRNA in 3
breast tumours. Additional investigation of these tumours led to the cloning and
characterization of the corresponding region in genomic DNA (intron 5) which was isolated
from a breast tumour expressing the 69 bp inserted ER mRNA. This revealed an A->G point
mutation immediately 3’ to the 69 bp sequence (Wang et al., 1997). This point mutation was
subsequently found to result in the generation of a consensus splice donor site. Comparison
with the same sequence from wt gene identified a consensus splice acceptor site
immediately 5’ to the 69 bp sequence, but not a donor splice site immediately 3’ to the 69
bp sequence. These data indicate that the 69 bp sequence was being recognized as an exon
by the splicing machinery, and resulted in the processing of a mature ER mRNA containing
the 69 bp insert.
In addition to the two carcinomas in which deletions within the proximal half of the
hormone binding domain (exon 5), another carcinoma demonstrated a product of
approximately 150 bp greater than that of wt product. Due to the constraints of time,
sequence characterization o f this variant was not possible, but it is feasible to postulate that
this variant may represent an exon duplication. Given the position of the PCR primers
within exon 4 (forward) and exon 6 (reverse) and the approximate difference of 150 bp from
wt product, it is attractive to propose that this variant contains a duplication o f exon 5 (139
151
bp). Whilst duplications have been observed for exon 3 with exon 4, exon 6 alone (Murphy
et al., 1996), and exon 6 with exon 7 (Pink et al., 1997), duplication of exon 5 has not been
previously reported and as a consequence it is difficult to postulate that if present, as
confirmed by sequence analysis, what the functional activity of this variant would be.
The exon 3 deletion variants were present in 9.1% (4/44) of the carcinomas, all of which
were positive for ER, with half o f these carcinomas negative for PgR expression. This
variant represents an in-frame deletion, internally lacking the sequence for a part of the
DNA binding domain and the second zinc finger, but would otherwise encode an intact
receptor protein. Whereas this receptor would be inactive by its own as a homodimer, it may
inhibit the transcriptional activation of wt ER by the formation o f non-functional
heterodimers, unable to bind DNA. O f the four carcinomas expressing the exon 3 deletion
variant, one expressed this variant in the absence of wt. Interestingly the carcinoma was
ER+/PgR- indicating that the variant might have a significant influence on the biology of
this tumour, with detectable aberrant ER protein responsible for the inability to induce the
expression of PgR.
Exon 7 deletion variants were most frequently observed (29.5% - 13/44). These were found
in 42.9% o f ER positive tumours (9/21) and 30% PgR negative tumours (7/23). Deletion of
exon 7 results in a frameshift and premature termination at the tenth position in exon 8,
giving rise to a truncated protein missing the C-terminal part of the hormone binding
domain including the TAF-2 region. Based on yeast expression system experiments, the
exon 7 deletion variant was suggested to be a dominant-negative inhibitor of normal ER
function, like the exon 3 deletion variant forming inactive heterodimers with wt receptor
(Wang and Miksicek (1991); Fuqua et al, (1992)).
A second type o f exon 7 deletion variant was also observed in which there was a complete
deletion o f exon 7 in addition to a T->C nucleotide substitution 5 bp proximal to the exon
6/intron 6 border and with deletion o f the first nucleotide of exon 8. Again these alterations
would introduce a shift in the reading frame, introducing a stop codon with a prematurely
truncated protein as a result. The presence of a nucleotide substitution only 5 bp from the
end of exon 6 and the first nucleotide of exon 8 missing could be suggestive o f a situation
similar to that described above for the insertion of nucleotides, with these point mutations
generating cryptic splice signals which could result in the aberrant splicing of the mRNA.
152
Confirmatory sequence analysis at the genomic DNA level was not carried out, but it would
be interesting to investigate its presence, confirming it as a true somatic alteration. In
addition, analysis similar to that described by Wang et al, (1997), i.e., cloning the region
surrounding the 29 bp insertion and isolation of the corresponding region using genomic
DNA from the carcinomas which expressed this variant, would also be advantageous in
order to test the generation o f slicing signal hypothesis.
Interestingly, this variant and an in frame exon 7 deletion variant already described, were
expressed in carcinomas in which they were the only species detected, i.e., no wt ER product
was detected. Both these carcinomas were ER positive with one negative for PgR expression
(imperfectly spliced) and the other with a low PgR H score of 62 near the cut off point of 50
for positivity (perfectly spliced). Again, given the functional activity o f this variant it is
attractive to postulate that it might have a significant influence on the biology of this
tumour, with detectable, inactive ER protein which is either unable to induce the expression
of PgR or affects the activity o f any wt ER activity to do so.
When the exon 7 deletion variants were correlated with steroid receptor profiles it could be
seen that although these variants was observed in all steroid receptor phenotypes, they were
observed in the majority o f ER-/PgR- carcinomas (6/21 ER+/PgR+ (28.6%), 3/18 ER+/PgR(16.7%), 4/5 ER-/PgR- (80%)).
O f the 44 carcinomas screened, none expressed the exon 2 deletion variant.
This is in
contrast to other studies in which this variant was observed in a significant proportion of
breast tumours e.g., 12% (Leygue, et al., 1996), and 41% (Gotteland et a l, 1995 and Zhang
et a l, (1996). Despite this, the finding o f no exon 2 deletion variant is in agreement with
these other studies which examined considerably larger numbers of breast carcinomas,
representing different stages o f the disease, in as much that of all splice variants detected,
the exon 2 deletion variant was the one which was least frequently observed.
Interpretation regarding the relative levels of these splice variants with that of wt transcripts
is difficult due to the preferential amplification of smaller molecules by the PCR process.
Despite this, it is still possible to see that the spliced variants were expressed at a lower level
than wt in all cases except one (cases 37 with ER Set 5). In hindsight it would probably have
been advantageous to examine the level of expression using the isotopically labeled
153
products analyzed by SSCP which would have been a more sensitive method for addressing
this feature. In addition it may have revealed whether there was a low level of expression of
wt receptor in those cases where there was no wt as determined on agarose gels with
ethidium bromide staining (cases 25 and 26 with ER Set 5).
Although this thesis, and other studies mentioned, demonstrate a wide range of variant ERs
in breast cancers they do not provide any clues as to whether wt and variant ERs are
coexpressed within the same cell or in different cells within the same tumour. Despite this, it
is not unreasonable to suspect that at least some of these variants may interfere with the
normal ER function if coexpressed. Another consideration is that despite the detection of a
considerable number and variety o f variants/ mutants at the mRNA level, there is a lack of
supporting data as to their presence at the protein level, i.e., are they translated? This has
prompted some to propose that their occurrence might be a laboratory artefact which results
from the high sensitivity o f RT-PCR (Dowsett et al., 1997). At the same time, these authors
also note that this is a difficult argument to sustain due to the fact that some variants
demonstrate a tissue-specific expression (Daffada and Dowsett, 1995). Only through the
generation of specific antibodies to the major splice variants, will the significance, if any,
regarding the occurrence o f variant ER expression in breast cancer, become evident.
To date only Desai et al., (1997) have raised an antibody to an ER splice variant, a
monoclonal antibody to the exon 5 deletion variant, which did not recognise the wt ER
protein. Although there was no evidence of this variant in breast carcinoma lysates as
assessed by electroblotting and gel shift assays, evaluation of frozen sections by
immunocytochemistry resulted in approximately 30% of tumours demonstrating the nuclear
staining which is characteristically seen for wt ER protein using the ID5 monoclonal
antibody (Dako). For these tumours, the presence of the exon 5 variant was related to the
presence of ER and PgR which is in agreement with the findings of the present study and
that of Daffada et al., (1995). The study of Desai et al., (1997) also showed that the
presence of the exon 5 deletion variant was associated with longer disease-free survival, but
was not an independent prognostic factor. These findings are controversial in that it has been
suggested that this variant could account for resistance to endocrine therapy, but they are in
agreement with the findings o f Daffada et al., (1995) who demonstrated, using RT-PCR,
similar levels o f the exon 5 deletion variant in tamoxifen-resistant cancers and primary,
untreated tumours.
154
These studies need to be extended to hyperplastic breast lesions and normal breast tissue to
determine the significance of these variants in normal mammary tissue biology and to put
into perspective their expression in breast cancers.
Even though ER splicing variants have been shown to be widely expressed in human breast
cancers (e.g.. Zhang et al, 1996), the number of naturally occurring missense mutations
identified in primary breast cancers are extremely low. It has been estimated that missense
mutations are present in only about 1% (2 of 188) of primary breast tumours (Roodi et al.,
1995). In agreement with this, a low frequency of missense mutations was detected in this
study with three mutations identified from two individual breast carcinomas (7% 3/44).
Although higher than the study described by Roodi et al, (1995), this may represent a
different tumour population and the lower numbers of carcinomas examined.
The first identified an A-»G immediately before the DNA binding domain boundary, which
resulted in the substitution o f the amino acid directly before this domain from a hydrophilic
glutamine (CAA) for a basic arganine (CGA). The amino acid substitution exchanges a
hydrophilic amino acid usually found on the exterior of proteins to a very basic amino acid
which would also normally be found on the exterior surface of a protein. This mutation has
not been identified previously and consequently the activity of the resulting protein, if
expressed, is unknown. It is difficult to postulate the activity of such a mutant receptor, but
the position o f this amino acid (amino acid 179) immediately before the highly conserved
DNA binding domain rather than within it, and the exchange of an amino acid with similar
properties suggests that the alteration may not have a profound influence on receptor
activity, especially since no significant conservation or function of the region proximal to
the DNA binding domain has been found.
The remaining mutations were identified in the same tumour and within the same region o f
the gene (exon 4 the ER hormone binding domain). In both cases, the mutations would
theoretically result in amino acid substitutions. The first, a T-»C substitution, would result in
the exchange o f a nonpolar leucine (TTG) for an uncharged polar serine (TCC) at amino
acid 308. This amino acid substitution would result in the exchange of an amino acid which
would normally be found clustered within the interior of a protein for an amino acid which
is relatively hydrophilic and usually found on the exterior of protein molecules although this
may not be necessarily so for the ER.
155
The second, an A->T substitution, would result in the substitution of a nonpolar isoleucine
(ATA) to a nonpolar leucine (TTA) at amino acid 326. This amino acid substitution would
result in the exchange o f an amino acid that would normally be found clustered within the
interior o f a protein with another of similar properties.
Again, mutants characterized by these alterations have not been previously seen, and
therefore the function o f the resulting proteins, if expressed is unknown. Interestingly,
Kumar et al., (1986) demonstrated, using a number of mutants, that the integrity of the
region located within amino acids 305 and 552 was indispensable for high affinity binding
of oestradiol. Mutants lacking portions of this domain resulted in impaired oestradiol
binding. Consequently, it is possible to speculate that these mutations together, or even if
they occurred individually, could code for mutant protein which had an altered ability to
bind oestrogens and consequently their ability to influence ER mediated gene transcription.
The degree to which each alteration could potentially disrupt oestradiol binding may be
indicated by the amino acid substitutions, which potentially have an effect on protein
secondary structure. The first, with the exchange of a relatively hydrophobic amino acid to a
hydrophilic, could potentially have greater influence than the substitution of a like for like
amino acid.
In all cases, due to lack o f time, the presence, of these mutations at the genomic DNA level
was not carried out. This analysis is critical in determining the validity of the mutations as
being true, somatic alterations and not due to other factors. These factors include the natural
error rate o f Taq DNA polymerase which was used in the PCR process. This is especially
important as three rounds o f PCR amplification were utilized, although one round used Pfu,
a thermal stable polymerase with high proofreading ability.
Four carcinomas demonstrated altered ER transcripts in the absence of wt ER expression
which is in stark contrast to other studies of breast cancers and cell lines in which wt ER is
always expressed, even at a lower level of expression than the aberrant ER species or no ER
species are of any description are observed (Daffada et al., 1994; Castles et al., 1995; Roodi
et al., 1995; Gotteland et al., 1995; Murphy et al., 1996; Leygue et al., 1996; Zhang et al.,
1996). Lack o f wt expression may be a consequence of a silencing mechanism such as
aberrant methylation o f a wt allele or other mechanisms leading to loss of wt allele at the
chromosomal level. In contrast, many of these studies have carried out semiquantitative
156
analysis of wt to variant levels in breast cancers and cell lines in which in many cases,
especially with more aggressive disease and cell lines, the variant ER species are expressed
at higher levels than wt ER species. Although no direct semiquantitative comparison was
carried out in this study, assessment by eye could only identify one carcinoma in which
there was an appreciable increase in an exon variant (exon 7 deletion) with respect to wt ER
transcript product.
To put this study into perspective it is useful to summarize the findings of the other major
studies on breast cancers and cell lines which represent different groups of breast tumours at
different stages of the disease. Gotteland et al., (1995) screened a large number of untreated
and non-metastatic breast carcinomas and found frequencies of 41%, 7.2% and 21% for
exons 2, 5 and 7 deletions respectively whereas Leygue et al., (1996) screened a large
number of a wide range o f breast carcinomas which were a mixture of the different steroid
receptor profiles, ranging in 1-6.3 cm in size and both node positive and negative, and
found frequencies o f 12%, 40%, 26% and 63% for exon 2, 3, 5, and 7 respectively. In
contrast Zhang et al., (1996) examined a large number of randomly chosen breast cancers,
from patients having an early recurrence (within 2 years of primary surgery) during adjuvant
treatment (tamoxifen resistant) and tumours from patients with long recurrence-free survival
(more than 5 years) during and after 2-5 years of adjuvant treatment (tamoxifen sensitive).
These breast tumours again had a mixture of steroid receptor profiles, were <20mm to
>50mm in size and both lymph node positive and negative. They observed frequencies of
41%, 74%, 66%, and 88% for deletions o f exons 2, 3, 5, and 7 respectively. Finally, Castles
et al., (1995) examined a large number o f ER positive and negative cell lines and expressed
the presence o f wt. and both exon 5 and exon 7 deletion variants in terms of the total percent
composition of ER mRNA in cells. They found that all three were expressed in ER positive
cell lines and in some ER negative cell lines but the striking feature was that in 87.5% of
cell lines expressing ER mRNA, the exon 5 deletion variant was the predominant transcript,
with in general wt transcripts the second most abundant and the exon 7 deletion variant the
least abundant. Contrasting this study with the findings of the 4 others suggests that as
tumours develop to more aggressive disease they express a greater range of variants with the
most striking trend being that the exon 5 deletion variant is more frequently expressed as
malignancy increases until it is the predominant ER transcript expressed, i.e., that the
acquisition o f the exon 5 deletion variant correlates with more aggressive disease. Given
that this variant has been shown to be constitutively active and resistant to antioestrogen
157
therapy this could account for the acquisition of hormone independence and resistant to
adjuvant therapy.
To date the significance for aberrant expression of ER spliced variants in human breast
cancer remains to be explained. Their identification in normal tissue has led to the
postulation that they may be expressed independently from cellular transformation
(Gotteland et al., 1995; Pfeffer et a l,
1996) by a process of differential splicing.
Consequently if translated, they may be important in modulating the overall oestrogen
responsiveness o f tissues in addition to the expression of oestrogen responsive genes. The
expression o f altered or uncontrolled levels of these variants, due to aberrations to the
splicing machinery, could be a contributing factor to the progression of breast cancer.
Pfeffer et al., (1995) addressed this hypothesis by examining mRNAs of genes closely
related to the ER (GR, aR A R , yRAR) in MCF-7 cells expressing ER splice variants but they
were found to be intact, indicating a process specific to the ER. In contrast, the findings of
Zhu et a l, (1997) support the hypothesis for aberrations to the splicing machinery in breast
cancer. They identified an exon 3 deletion splice variant androgen receptor (AR) in human
breast carcinomas and cell lines which was predicted to have reduced ability to activate
transcription. This variant was found to be expressed in excess of wt AR in 15% of breast
cancers, at low levels in 3 breast cancer cell lines but was not detected in normal breast
tissue.
An alternative or additional explanation for altered splicing of the oestrogen receptor in
human breast cancer could be because alterations to nucleotide sequences within both
introns and exons which direct the splicing machinery. Altered nucleotide sequences could
remove or create consensus splice elements which would be recognized by the splicing
machinery and lead to the processing o f mature modified ER transcripts, e.g., the A->G
point mutation which resulted in the creation of 3’ splice donor site within intron 5, not
present in wt ER gene, leading to the creation of a 69 nucleotide insertion between exons 5
an 6 (Wang et al., 1997). Considering this example, it is not impossible to imagine that
mutations within consensus splice elements (donor or acceptor) within introns could result
in a permanently altered allele which only expresses a particular spliced form of the gene.
Put into the context o f other alterations to the remaining allele, promoter methylation
abnormalities, and other factors which modulate the oestrogenic response (e.g., the
158
phosphorylation state of expressed receptor protein) this could make a profound contribution
to the pathogenic development o f breast cancer.
159
3.
L O S S O F H E T E R O Z Y G O S I T Y A T C H R O M O S O M E 6 q 2 5 .1 -2 7
3.1
Loss O f Heterozygosity
One of the most frequent genetic alterations found in human malignancies is due to deleted
fragments of chromosomes, which include genes critical for the maintenance of cellular
interactions and growth control. These include chromosomes 5q, 17p and 18q which
represent loss at the APC gene, the p53 gene and the DCC gene respectively. This feature is
commonly known as allelic loss or loss o f heterozygosity (LOH).
The significance o f somatically occurring LOH in tumour cells stems from studies on
retinoblastoma. These have shown that two mutational hits, inactivating both alleles of the
tumour suppressor gene RBI are crucial for the development of the tumour (Knudson,
1989). The first hit, either a point mutation or intragenic deletion, is thought to be recessive
at the cellular level, as the remaining intact allele would be sufficient to prevent tumour
growth. For this reason this first inactivating hit may be passed through the germline and
thus predisposes the recipient to retinoblastoma (Figure 3.1).
The second hit appears to occur as a consequence of either interstitial deletions,
chromosome loss, or as most recently observed, most frequently due to aberrant mitotic
recombination events (Cavenee et al., 1983; Seemayer et al., 1989; Levine, 1990; Gupta et
al., 1997), and thus always occurs somatically giving rise to cells which are hemi- or
homozygous for the mutation in R B I. As a consequence the function of R B 1 is lost and
tumour growth follows. It has been estimated that the frequency of LOH in tumour cells is
at least two orders o f magnitude higher than that of point mutations (Weinberg, 1991;
Harwood et al., 1993) and is thus favoured over other inactivating mechanisms.
The implication of the retinoblastoma model was that the finding of LOH in any other
malignancy could signify the presence of a nearby tumour suppressor gene locus. This
prompted many investigators to compare the constitutional genotype of cancer patients with
that of their tumours to find clues as to the location of such genes. The outcome o f this
research was that the genetic mechanisms proposed for RBI were subsequently found to
exist also in many other tumour suppressor genes, for example, the APC gene in colon
cancer, the WT1 gene in Wilms' tumour and p53 gene in breast cancer (Lasko and Cavenee,
1991; and Devilee and Comelisse, 1994).
161
CONSTITUTIONAL
PREDISPOSED CELL
Heritable
TUMOUR
Loss
O
9
9
Loss/duplicate
? ^ Germline Mutation
►
9
+
I
Recombination
Sporadic
Somatic Mutation,
Localized
9
*
■ *
F igu re 3.1 A general m odel for m echanism eliciting loss o f heterozygosity. Upper left: In heritable disease, a
recessive defect (labeled *) at the tumour locus is inherited, giving a genotype */+ in all cells. The
predisposition is unmasked by elim ination o f the w ild-type alleles, usually by one or the chrom osom al
m echanism s described in the text, follow ed by other alterations during malignant progression. Lower left: In
sporadic disease, som atic mutation at the tumour locus results in formation o f a predisposed precursor cell.
Predisposition is unmasked in m echanistically sim ilar w ays as in heritable disease. (Taken from Lasko et al.,
1991)
162
Before 1985 the assessment of allelic loss in chromosomes was limited, as tumours had to
be o f a sufficient size so samples could be frozen for extraction of high quality DNA
without hindering histological evaluation, thereby excluding small, early breast carcinomas.
Also the position and frequency of heterozygosity was limited as restriction fragment length
polymorphisms were used. Then in 1985. the development of two new molecular tools
revolutionized LOH analysis. Firstly, the polymerase chain reaction (PCR), a technique for
amplifying small quantities of DNA (Saiki et al., 1985), and secondly the identification of a
new type o f DNA marker known as microsatellites (reviewed by Weber and May, 1989).
Microsatellites are short, simple repeated sequences, (l-6 n) e.g., (CA)n, that are highly
polymorphic, evenly distributed throughout the genome, and can easily be detected with
PCR (Weber and May, 1989). By exploiting these microsatellites, it is possible to define
areas o f common deletion by comparing the tumour and normal DNA from the same
individual. In the tumour DNA o f a patient heterozygous for a certain polymorphism, it is
possible to observe that one o f the alleles is lost relative to the corresponding normal DNA
o f that patient. One problem in using this type of analysis is that polymorphic DNA markers
o f this kind reveal any imbalance in parental alleles, i.e., loss of an allele in addition to gain
in allele copy number. Although in practice it is difficult to discriminate between loss and
gain relative to normal DNA, a number of controlled studies have demonstrated that
approximately 80% o f observed allelic imbalances are in fact losses (Devilee and
Comelisse, 1994) and consequently a useful tool for LOH analysis.
3.2
Loss O f Heterozygosity In Breast Cancer
Our knowledge of LOH in breast cancer is predominantly due to three main studies which
carried out genome wide analysis on over 400 breast tumours and identified consistent LOH
on chromosomes 3p, 16q, and 17p in over 45% of informative cases (Larsson et al., 1990;
Sato, et al., 1990: Devilee et al., 1991). Other researchers have focused on particular areas
of specific chromosomes which identified key LOH regions which are summarized in Table
3.1. The cumulative evidence highlights the complexity of chromosome involvement in
breast cancer given that half the cases studied showed LOH at more than 2 different
chromosomes. Studies looking at LOH at loci on different chromosomes have identified 3p
and 17p, 13qand 17p, and l i p and 17p (Devilee et al.. 1991; Borg et al.. 1992; Takita et al.,
1992).
163
Table 3.1 Summary of loss of heterozygosity data in breast cancer
Chromosomal Arm
Cases LOH
Range (%)
Example
% LOH
lp
(%)
26.5
3-47
lp31
28%
Nagai et al., 1995
iq
33.0
0-65
1q21-31
65%
Bieche et al., 1995
6q
36.4
9-52
6q25-27
39%
Orphanos et al., 1995;
7q
23.7
0-41
7q23
41 %
Bieche et al., 1992
8p
38.7
27-50
8p 12-21
50%
Kerangueven et al., 1995
9q
24.0
9-58
9q34
58%
Brenner and Aldaz, 1995
llq
53
42-66
1 lq23
42%
Koreth et al., 1992
lip
27.1
8-41
1lp l 5.5
35%
Winquist et al., 1995
13q
27.7
0-42
13ql2-13
42%
Schott etal., 1994
16q
52.3
40-62
16q24.4qter
45%
Cleton-Jansen et al., 1994
1?P
57.0
37-75
17pl 3.3
48%
Ito et al., 1995
18q
20.0
3-36
18q21
31 %
Thompson et al., 1993
164
Reference
Deng et al, (1996) detected LOH in morphologically normal lobules adjacent to breast
tumours. They observed LOH most frequently at 3p24 (48%), 1I p l5.5 (29%), 13 q l3 (64%)
and at 17 p l3.1 (80%). In a high proportion of cases, the alteration observed in the tumour
was also observed in the adjacent normal lobules. These findings suggest that the molecular
heterogeneity that characterizes invasive breast carcinomas may occur at the earliest stages
of development.
In addition, Devilee suggested that breast cancer develops, like colon cancer, through a
defined progression of morphologically distinguishable stages beginning with benign
hyperplasia, which progresses to atypical hyperplasia, to an in situ carcinoma and finally to
invasive carcinoma (Figure 1.4). Many small invasive carcinomas do not have atypical
components (Deng et a l, 1996), suggesting that they may have developed directly from
normal epithelium. Given that cancer develops by clonal selection one might expect to find
evidence that at least some o f the genetic alterations found in invasive carcinomas are also
present in normal epithelium, a hypothesis which is supported by the findings by Deng et
al, (1996).
3.3
Loss o f Heterozygosity A t Chromosome 6q In Breast Cancer
Cytogenetic analyses of primary breast tumours have identified frequent alterations to a
number of chromosomes, notably deletions, suggesting the potential localization of tumour
suppresser genes (Devilee and Comelisse, 1994). These studies demonstrated that deletion
of chromosome 6q was one of the most frequent chromosomal changes (Dutrillaux et al.,
1990; Mars and Saunders, 1990). A subsequent study, using Southern analysis of restriction
fragment length polymorphisms to compare constitutional and tumour DNAs, identified
chromosome 6q as one of the most frequent sites for allelic loss (LOH) after 17p in breast
cancer (Deville et al., 1991).
Other evidence for the presence of putative tumour suppresser genes on chromosome 6q
comes from chromosome mediated transfer experiments of normal chromosomes into
melanoma cell lines (Trent et al., 1990), uterine endometrial cell lines (Yamada et al., 1990)
and the breast cancer cell lines MDA-MD231 and MCF-7 (Negrini et al., 1994) all resulting
in the suppression of tumourigenesis.
165
Ivvase et al., (1995) hypothesized that ER negative breast tumours might be induced by the
mutation of one allele and loss or replacement of a chromosomal segment containing the
other allele. They tested their hypothesis by screening the DNA from a group of breast
tumours for LOH at the ER gene locus but found no significant relationship between LOH
and ER status.
The advent of PCR analysis o f microsatellite polymorphisms has confirmed the cytogenetic
evidence for chromosomal deletion at 6q, and has enabled construction of a more detailed
deletion map. Allelic loss at 6q24-27 has been observed in different tumour types including
breast carcinoma (Orphanous et al., 1995), ovarian carcinoma (Rodabaugh et al., 1995;
Saito et al., 1992b), hepatic carcinoma (De Souza et al., 1995a, b), small cell lung
carcinoma (Merlo et al., 1994), renal cell carcinoma (Morita et al., 1991), malignant
melanoma (Walker et al., 1994; Millikin et al., 1991), and non-Hodgkin’s lymphoma
(Menasce et al., 1994). This shared region of allelic loss may harbor putative tumour
suppressor genes that are pleiotropic for these tumour types and reflect a common
mechanism of tumourigenesis.
Recent detailed analyses o f microsatellite markers on chromosome 6q in breast cancers have
defined two key regions on chromosome 6q showing LOH levels higher than that o f
background at 6ql3 (26-35%) and 6q25-27 (27-50%), indicating the presence of at least two
tumour suppressor genes (Devilee et al., 1991; Orphanous, et al., 1995; Fuji et al., 1996c).
Noviello, et al., (1996) carried out a detailed deletion map on 6q and suggested the existence
of three, distinct regions of allelic loss at 6ql3, 6q24-25, 6q27 with a possible fourth region
at 6q21. Since these studies were concerned with symptomatic, well established breast
carcinomas it is not clear whether allelic loss on chromosome 6q is an early event in the
development of breast cancers. Small, mammographically-detected breast cancers form a
useful group for study o f the involvement of tumour suppresser genes in tumour
development and earlier stages of progression.
The ER gene has been mapped to 6q25.1 (Menasce et al., 1993), within a region researchers
have found to be commonly affected by LOH in breast cancer. In the early stages o f the
disease, the proliferation of tumour cells depends on oestrogen, after which the tumour cells
acquire new proliferative pathways as a result of multiple genetic alterations. This then
enables the tumour cells to bypass oestrogen-dependent proliferation (Liu et a l, 1993).
166
Although there are many reports concerning variant ER genes (Sluyser et al., 1995), the
exact mechanism(s) for the existence o f ER negative breast tumours is still not understood.
3.4
Candidate Tumour Suppressor Genes On Chromosome 6q
3.4.1
T h e iM a n n o se 6 -P h o s p h a te /I n s u lin -L ik e G r o w th F a c to r 2 R e c e p to r
Experimental evidence suggests that paracrine interactions between stromal and epithelial
cells are important influences on the growth and malignant behavior of breast cancers
(Singer et al..
1995). The use o f both in situ hybridization (Paik,
1992) and
immunohistochemistry (Ellis et al.. 1994) has demonstrated that insulin-like growth factor II
(IGF-II) is expressed by fibroblasts in both benign and malignant breast lesions. IGF-II is a
potent mitogen for a number o f breast cancer epithelial cell lines in vitro and it is thought to
exert its mitogenic effect primarily through the Insulin-like growth factor I receptor. In
addition, IGF-II also binds to its cognate receptor the mannose 6-phosphate/IGF-II receptor
(M6P/IGF2R).
Primary functions of the M6P/IGF2R include trafficking of newly synthesized lysosomal
enzymes from the Golgi to the lysosomes, and the endocytosis of extracellular lysosomal
enzymes (Dahms et al., 1989). Flowever, apart from binding IGFII, growth factors, e.g.,
proliferin, and the inactive complex o f TGFpl also bind the M6P/IGF2R (Purchio et al.,
1988; Kovacina et al., 1989). Although binding leads to internalization and subsequent
degradation, the extracellular activation of TGFpl by plasmin is greatly facilitated by the
binding of the TGFpl latent complex to the receptor (Dennis and Rifkin, 1991; Kojina et
al, 1993). Consequently the M6P/IGF2R is required for both the activation of the growth
inhibitor, TGFpi and the degradation o f the mitogen IGFII (Komfeld, 1992), and therefore
plays an important part in negative cell growth control.
M6P/IGF2R mRNA has been detected in both breast cancer cell lines and breast cancer
tissue (De Leon et al., 1988; Cullen et al., 1990). In situ hybridization analyses on breast
cancer biopsies suggest a higher level o f expression in breast carcinomas than that in benign
epithelium or stroma (Zhoa et al., 1993). Comparisons of M6P/IGF2R RNA levels between
tumour and non-tumour breast tissue using northern analysis demonstrated expression in all
tissues tested with no significant differences in the level of expression between tumour and
non-tumour tissue (Hebert et al.. 1994). Analysis of the M6P/IGF2R gene copy number in
this same tumour group, showed no amplification of the gene whatever the clinical
167
presentation o f the tumour and irrespective of a concomitant amplification of c-erb&2 or int2 genes in several tumours.
The M6P/IGF2R gene has been mapped to chromosome 6q26-27 (Laureys et al., 1988) and
recently, De Souza et al., (1995a. b) demonstrated frequent LOH at the M6P/IGF2R locus in
human hepatocellular tumours and identified point mutations in the remaining allele in 25%
of these cases, strongly suggesting that the M6P/IGF2R gene functions as a tumour
suppressor gene in human liver carcinogenesis. More recently Hankins et al., (1996)
demonstrated frequent LOH in both invasive and non-invasive breast carcinomas and found
missense mutations in 2 out of 5 (40%) of the later group giving rise to amino acid
substitutions. These data indicate that LOH at this locus is an early event in the etiology of
breast cancer and that the M6P/IGF2R functions as a tumour suppressor in the breast.
3.4.2
T h e T A T A B o x -B in d in g P r o te in (T B P )
In eukaryotes, transcription is carried out by three different RNA polymerases, RNA
polymerase A, II, and III, each of which is dedicated to the transcription o f different sets of
genes. The genes in each class contain characteristic promoters, which often consist of two
types o f elements: the basal promoter and the modulatory promoter elements. The basal
elements are sufficient to determine RNA polymerase specificity and direct low levels of
transcription, whereas the modulatory elements enhance or reduce the basal level of
transcription. None o f the RNA polymerases can recognize its target promoters directly.
Instead, basal promoter elements are first recognized by specific transcription factors, which
then recruit the correct RNA polymerases. The TBP is one such factor and is a subunit of
TFIID and has been shown to bind to the TATA box present in a large number of RNA
polymerase II promoters (Swadogo and Sentenac, 1990; Buratowski et al.,
1991).
Transcriptional regulation above basal RNA polymerase II transcription requires a complex
formation containing TFIID consisting of TBP and a number of associated TBP-associated
factors.
One mechanism by which p53 protein appears to function as a tumour suppressor is by
inducing the expression of gene products that are responsible for inhibiting or arresting cell
growth and proliferation (El-Deiry et al., 1993). Recent studies have provided circumstantial
evidence that p53 communicates with the transcriptional machinery by means of a direct
interaction with TBP (Seto et al., 1992; Truant et al., 1993; Liu et al., 1993 and Chen et al.,
168
1993). Consequently the ability of p53 to regulate transcription of its target genes is in part
dependent on its binding to the TBP and subsequent DNA binding (Raycroft et al., 1990;
Unger et al., 1992; Kern et al., 1992). Consequently it is possible to hypothesize that a
reduction in TBP expression has the potential to alter the transcriptional activation of genes
critical for the inhibition of growth and proliferation of cells. Additionally, the
transcriptional activation o f other genes by TATA box-RNA polymerase II mechanisms,
e.g. the oestrogen receptor, could also be affected.
The TBP gene has been mapped to 6q27 by fluorescence in situ hybridization (Saito et al.,
1994) a region demonstrating a high frequency of loss in many neoplasms including breast.
The N-terminal region o f human TBP contains a characteristic primary structure: a stretch of
glutamine residues, that are encoded by the trinucleotide (CAG)n repeat (Hoffmann et al.,
1990).
169
3.5
Microsatellite Instability
In addition to defining regions of common deletion, microsatellite markers can also be used
to detect a distinct genetic phenomenon known as microsatellite instability (MI). Originally,
microsatellite markers were used in genetic mapping and linkage analysis, identifying
specific mutations in inherited disorders such as Spinal and Bulbar Muscular Atrophy (La
Spada et al., 1991) and Myotonic Dystrophy (Brook et al., 1992). These conditions were
characterized at the molecular level by the expansion of trinucleotide repeats within
susceptibility genes.
In 1991, Loeb was the first to suggest the concept that mutations in stability genes may be
involved in carcinogenesis. These initial mutations can be inherited (as HNPCC) or arise
from acquired DNA damage. These mutations in stability genes generate a series of
secondary mutations throughout the genome, some of which involve, e.g., oncogenes that
alter the regulatory mechanisms o f the cell. This concept is summarized in Figure 3.2.
Early M utations In
S tab ility G enes
Chemicals
X-rays
Endogenous
Processes
r hMSH2
hM LH l
hPM Sl
hPMS2
hMSH6
HMSH3
M utations In C ancer
A ssociatedG en es
MUTATOR - y f
PHENOTYPE
(MSI)
ak.
TGFPRII
M6P/IGF2R
Bax
APC
Cancer
Development
and
Progression
+ d)
F ig u re 3.2 Schem atic representation o f the sources for m ultiple mutations in cancer as initially proposed by
Loeb, 1991, 1994. The hypothesis proposes that mutations in stability genes m ight lead to an increase in
mutation rate that could in turn lead to mutations at candidate cancer associated loci, e.g., p53 and R B I. The
detection o f MI in a high proportion o f colon cancers has provided excellent confirmation o f this hypothesis.
Inactivation o f the mism atch repair genes (HMSH2, hMSH3, h M L H l, h P M S l, hPM S2, and HMSH6) leads to
widespread genetic instability detected as MI which in turn may target specific cancer associated loci for
m utations (e.g., TGFPRII, M 6P/IGF2R, Bax and APC). Recent evidence has demonstrated a “cascade effect”
w hereby inactivation o f both alleles o f mism atch repair genes (e.g., hM SH 2) leads to inactivation o f a second
m ism atch repair gene (e.g., hM SH 6), Rampino e t al., 1997 (1).
In 1993, a number o f workers in colorectal cancer research identified a more widespread MI
detected as expansion or contraction of mono-, di- and tri- repeat elements in hereditary
non-polyposis colorectal cancer (HNPCC) as well as sporadic colorectal cancer cells
(Aaltonen et a l, 1993; Thibodeau et al., 1993; Ionov et al., 1993). HNPCC patients develop
170
colorectal cancer at an average age o f 42, more than two decades earlier than the general
population. Although colorectal carcinoma is the major cancer in HNPCC families, about
35-40% o f the people in these families have other types of tumours, of which endometrial
and ovarian tumours are the most common. In addition, approximately 35% of people in
HNPCC families develop more than one tumour. In addition, three articles reported that an
unusual form o f somatic mutation occurred in 12-15% of all colorectal tumours and was
found in virtually all the colorectal tumours of HNPCC (Aaltonen et al., 1993; Thibodeau et
al., 1993; Ionov et al., 1993). These mutations consisted of changes in the length of DNA
tandem repeat sequences (microsatellites) that are normally interspersed throughout the
genome. All three studies recognized that errors of insertion or deletion of dinucleotide
repeats or other short repeated sequences occurred during the replication of DNA by
neoplastic cells in HNPCC tumours. Therefore it was recognized that replication errors
(RERs) could possibly represent a new mechanism contributing to tumour development.
In bacteria this mutational fingerprint is characteristic of a failure of the mismatch repair
(MMR) enzymatic complex. The best known MMR pathway is the E. coli MutHLS system
that promotes a 'long-patch* methyl-directed repair of single base-pair mismatches and short
mismatched loops in newly synthesized DNA, thus increasing the fidelity of DNA
replication (reviewed in Modrich, 1991). The bacterial MutS protein binds to DNA
mismatches, and other proteins of the MutHLS system, which distinguish the newly
synthesized unmethylated strand, excise the incorrectly synthesized nucleotides, and
resynthesize double-stranded DNA. Failure of the bacterial MutHLS repair system results in
genetic instability.
3.5.1
Mismatch Repair Genes
The human homologue o f the MutS, hMSH2 was subsequently cloned and disease-causing
mutations were characterized in HNPCC kindreds (Fishel et al., 1993, Leach et al., 1993).
Several additional human mismatch repair genes {hMLHl and hPMSl through hPMS8) have
been cloned by virtue o f homology to bacterial and yeast MutL homologues (Nicolaides et
al., 1995). Together, mutations in /*MSH2 (44%) and /?MLH1 (23%) account for the major
contribution o f all HNPCCs (Bronner et al, 1994) and therefore it is easy to hypothesize
that these related genes have a causative role in sporadic, non-familial cancers also. But
recent studies in sporadic colorectal cancers demonstrate that germline mutations in
mismatch repair genes are uncommon which may reflect the type o f mutation analysis
171
utilized (Borrenson et al, 1996). In addition, it has been demonstrated that microsatellite
alterations in cultured tumour cells can occur in the absence of germline mutations,
suggesting that mutations in genes other than those discussed here are responsible for
permitting replication errors in sporadic colon cancer (Liu et al., 1995), e.g., DNA repair or
chromosomal segregation.
3.5.2
Mechanism O f Mismatch Repair
In human cells, mismatches are recognized by a heterodimer, hMutSa, comprising hMSH2
and hiMSH6 (Fig. 3.3). Biochemical and genetic data indicate that inactivation of each
partner has a different effect on genome stability (Karran, 1995). Mutations in hMSH6 are
associated with a selective inability to repair single base mismatches and single nucleotide
loops. In contrast, cells with HMSH2 mutations are unable to repair single base mismatches,
single base loops or dinucleotide loops. It has now been proposed that hMSH2 participates,
either alone or in combination with an unidentified partner (hMSH3 is a strong candidate) in
a second mismatch repair function that stabilizes dinucleotide or larger repeats.
After mismatch recognition binding o f MutL occurs. The MutL component is also a
heterodimer (hMutLa) o f hM LHl and hPMS2. The fact that these MutL homologues form
functional heterodimers, combined with evidence that a mutation in hPMSl is also
associated with HNPCC, suggests that hPMSl may form a heterodimer with a different
MutL homologue. This putative partner may be among the family of six MutL-like genes
(hPMS3-hPMS8) on chromosome 7.
The latter steps in mismatch repair are less clear. The interaction of MutS and MutL
heterodimers is likely to lead to the recruitment of polymerases, nucleases and other proteins
required for the repair processes. Recently, the hMSH2 gene has been postulated to be a
novel p53-regulated target gene (Scherer et al, 1996). This may indicate a direct
involvement of p53 in repair mechanisms via DNA binding of a mismatch repair gene.
3.5.3
Microsatellite Instability In Sporadic Cancers
The RER+ or microsatellite instability (MI) phenotype has been found in a significant
proportion of a wide variety o f
sporadic tumours (Eshleman and Markowitz, 1995),
possibly suggesting that many sporadic tumours may have acquired somatic mutations
172
similar to those inherited in HNPCC. Despite this, little is known regarding MI as an
indicator o f mismatch repair defects in sporadic disease, as studies of sporadic colon and
endometrial tumours have suggested that somatic mutations in hMSH2, hM LH l , hPM Sl and
hPMS2 are rare (Liu et a l, 1995).
M ism a tc h
A
CA
A
r\
MSH6
ZV
Recognition
■
MSH3
R ecruitm ent
M
L
Assembly
Figure 3.3 Early stages o f mismatch repair D isplaced single bases (e.g. -A-) are recognized by the mismatch
binding heterodimer MSH2:MSH6 which bind to the mispaired region. Dinucleotides loops (e.g., -CA -) are
bound by an alternative com plex com prising M SH2:M SH3. The MLH1:PMS2 heterodimer is then recruited to
the bound com plexes and this facilitates further assembly o f other proteins leading to excision and resynthesis
o f a stretch o f D NA containing the m ism atched region (adapted from Karran, 1995 and 1996).
3.5.4
Microsatellite Instability In Breast Cancer
To date there have been 20 studies o f microsatellite instability in breast tumours. Invasive
carcinomas represent the most studied group (summarized in Table 3.2), however, a few
studies have studied pre-malignant in addition to preinvasive breast lesions. The status and
relevance o f microsatellite instability in breast cancer is currently being debated. For
173
example, Peltomaki et al., (1993) found no evidence of MI in breast carcinomas, whereas
Patel et a l, (1995) detected instability in all of the tumours studied.
3.5.4.1 Invasive Carcinomas
The lack o f a standard definition for the presence of microsatellite instability in a tumour has
led to widely divergent estimates of its incidence in sporadic cancers. Therefore the results
have been divided in to tumours showing instability at a single locus and those with
alterations at more than one locus (Table 3.2).
3.5.4.2 Non-Invasive Carcinomas
The most comprehensive study to date of DCIS has been described by Aldaz et al., (1995).
This study investigated MI in 23 cases of pure DCIS. Three (13%) tumours with MI were
identified. Toyama et al., (1996b), studied two foci of DCIS from two invasive tumours
which had previously been demonstrated to show microsatellite instability (Toyama et al,
1996a). Both areas demonstrated the same pattern of instability to that seen in the invasive
components.
3.5.4.3 Pre-Malignant Disease
The study described by DeMarchis et al., (1997) included seven cases of fibroadenoma.
Eight microsatellites were analyzed, but there was no evidence of instability in these cases.
Rosenberg et al., (1997) examined 12 separate atypical hyperplastic lesions, from 6 breast
cancer patients, for microsatellite instability. Alterations were detected in 5 of the 12 ADH
lesions, from three patients. Kasami et al., (1997) demonstrated MI in 2 of 8 breast biopsies
with hyperplastic lesions.
3.5.5
Mismatch Repair Genes In Breast Cancer
To date mutational analysis or expression studies of candidate MMR genes have not been
reported in breast tumours displaying MI. However, a recent study has identified a breast
tumour (as an integral tumour within the HNPCC syndrome) which displayed MI and had a
frameshift mutation in hMLHl (Risinger et al., 1996). This was the first evidence that breast
cancer, albeit within the context of HNPCC, may result from the inheritance of a mutant
mismatch repair gene.
174
Table. 3.2. Summary o f microsatellite instability in invasive breast carcinomas from other
groups
No. Tumours
studied
MI
No. Tumours
with MI
Reference
Single locus
Multiple loci
Peltomaki et al., 1993
84
0 (0%)
-
-
26
1 (4%)
1
0
Han et al., 1993
104
11 (10%)
10
1
Wooster et al., 1994b
20
4 (20%)
1
3
Yee et al., 1994
14
6 (3%)
6
0
Glebov et al., 1994
13
13 (100%)
1
12
Patel et al., 1995
30
2 (7%)
?
?
Jonsson et al., 1995
28
6(21% )
5
1
Contegiaco et al., 1995
52
12 (23%)
?
7
Aldaz et al., 1995
69
20 (29%)
19
1
Kamick et al., 1995 -
37
25 (67%)
14
11
Paulson et al, 1996
100
8 (8%)
6
2
Toyama et al., 1996a, b
81
27 (33%)
7
20
De Marchis et al, 1997 "
42
11 (26%)
9
2
Sourvinos et al., 1997
46
13 (28%)
8
5
Rush et al., 1997
Key MI (microsatellite instability)
175
-
3.5.6
Mismatch Repair Mechanisms And Cancer
A model is now emerging whereby the mutator phenotype unfolds in gradual steps
(Perucho, 1996). Rare homozygous mutations in a mismatch repair gene (e.g., hMSH2 or
hM LH l) may lead to mutations in secondary mismatch repair genes (e.g., hMSH3 and
hMSH6) because they contain the targets for the mutagenic action of the first mutator
mutations. This model has been confirmed by studies carried out by Malkhosyan et al.,
(1996). Various tumour types, which display MI, also have ffameshift mutations of
polyadenine (Ag) and polycytidine (C8) tracks within the coding regions of hMSH3 and
hMSH6 respectively. This enhanced genomic instability may accelerate, even more, the
accumulation of mutations in cancer genes during tumour progression as originally
proposed by Loeb, 1991. If these types of alteration cause mutations to accumulate in
tumour suppresser genes and/or oncogenes, then the mutant cells will probably progress to
become tumour cells. Target genes o f this type are now beginning to emerge.
3.5.6.1 Transforming Growth Factor p Type II Receptor
TGFP is one of the most potent inhibitors of normal cell growth. However, many
malignancies of epithelial origin are resistant to TGFp, suggesting that developing
resistance to growth inhibitory cytokines plays an important role in tumourigenesis
(reviewed in Polyak, 1996). Tumours can become resistant to TGFp in many different ways.
However, mutation in one o f the TGFp signaling receptors (type II) has recently been
associated with microsatellite instability in colon cancer (Markowitz et al, 1995). The
TGFp type II receptor (TGFpRII) contains a region of 10 consecutive adenines which is a
target for up to 90% o f colon tumours with the RER+ phenotype. Insertions or deletions of
adenines within this repeat produce ffameshift mutations, results in truncated type II
receptors, and hence resistance to TGFp. However, this does not seem to be a universal
target for all RER+ tumour types. Endometrial tumours with microsatellite instability do not
show mutations at the TGFpRII poly A tract (Myeroff et al., 1995).
3.5.6.2 Mannose 6-Phosphate/Insulin-Like Growth Factor 2 Receptor
As already discussed in section (3.4.1) the mannose-6-phosphate receptor or insulin-like
growth factor type II receptor (M6P/IGF2R) plays roles in a number of cellular functions,
the most critical of which is the activation of the potent growth inhibitor TGFp (Dennis and
Rifkin, 1991). This receptor also inhibits cellular proliferation mediated by IGFII, which is a
potent growth stimulant, by internalizing and degrading this protein (Komfeld, 1992).
176
Recent studies have suggested that M6P/IGF2R functions as a tumour suppresser gene, in a
number of tissues, including breast (Hankins et al., 1996). The M6P/IGF2R gene harbors a
tract o f 8 guanines in its coding region. Recent studies have characterized mutations of this
sequence in RER+ tumours of the gastrointestinal tract and endometrium (Souza et al.,
1996, Ouyang et al., 1997).
3.5.6.3 Bax
The development and growth of tumours is controlled by a combination of cellular
replication and cell death (Williams, 1991). The most common mechanism of cell death is
programmed cell death which is a precisely regulated process that culminates in a set of
characteristic morphological and structural alterations referred to as apoptosis (Wyllie et al,
1980). Many apoptotic stimuli induce cell death through a pathway that is regulated by
members of the Bcl-2 family which function to either promote or inhibit apoptosis. Bcl-2 is
an inhibitor o f apoptosis and can heterodimerize with Bax, a proto-apoptotic member of this
family of proteins (Otavi et al., 1993).
The ability o f Bax to promote programmed cell death suggests that it may function as a
tumour suppressor gene, with a reduction in functional Bax giving cells a growth advantage
and contributing to their expansion. The Bax gene contains a tract of 8 consecutive guanines
in exon 3 (Oltvai et al., 1993). One nucleotide insertions or deletions within the poly G tract
were identified in 51% o f colon tumours which were also R E R \ but not in any tumours
characterized as RER*(Rampino et al., 1997). The ffameshift mutations resulted in truncated
proteins which result in loss o f BAX function. Cells lacking BAX protein are likely to have
a diminished capacity to trigger apoptosis after receiving a death signal, thereby facilitating
tumourigenesis.
3.5.6.4 E2F-4
E2F-4 is a member o f a family o f transcription factors which are involved in the regulation
of the cell cycle and apoptosis (La Thangue, 1994). E2F-4 is involved in the transition from
G0 to G, phase of the cell cycle, as well as in early G,, facilitating the transactivation of
genes necessary for cellular proliferation (Sardat et al., 1995). This gene contains a long
spacer region encoded by 13 consecutive serine (CAG) residues, which may function as a
transactivation domain (Sardat, 1995). This repeat is polymorphic in the normal human
population and therefore a functional effect of mutation in this region can not be assumed.
177
However, modification of the transactivation domain may alter the expression of E2F-4
responsive genes. Yoshitaka et al. (1996) demonstrated that 2/20 (10%) of patients with
sporadic colorectal carcinoma had tumour specific alterations at the 13 consecutive
trinucleotide repeats. O f the twenty tumours studied only the two with E2F-4 alterations
were classified as RER \
An extended study by Souza et al. (1997), who examined 86 neoplastic lesions, including
46 sporadic colorectal carcinomas (31 of which were REIT and 15 RER'), gastric
adenocarcinomas, endometrial cancers and prostate carcinomas, all of which were classified
as RER". Of the RER" gastrointestinal tumours, 37% demonstrated tumour specific
alterations of the CAG repeats. None of the RER* gastrointestinal tumours or the RER+
endometrial or prostate tumours showed alterations, suggesting that this gene is a frequent
target of microsatellite instability in gastrointestinal tumourigenesis.
3.5.6.5 Other Targets
At present, the four genes described above represent the only targets which have definitively
been correlated with the presence o f microsatellite instability in a large number of tumours.
However, there are a wealth o f candidate genes which possess similar repeat motifs in their
coding regions. For example, the BRCA2 gene contains a number of poly A tracts in its
coding region. Insertion of an adenine at of one of these regions has been described in an
ovarian tumour which also showed microsatellite instability (Takahashi et al, 1996). A
recent review o f BRCA1 mutations also found that mutations at sites of homonucleotide and
short repeats was common (Rodenhiser et al, 1996).
A im s O f Chapter
The objectives of this chapter were to determine if loss of heterozygosity on the region of
chromosome 6, to which the oestrogen receptor gene has been mapped (6q25.1), was present
in cases of preinvasive breast cancers (DCIS) and mammographically detected, impalpable,
early invasive breast carcinomas and tubular carcinomas. Microdissected tumour foci were
to be obtained and analyzed using microsatellites to determine allele loss. In addition the
region surrounding the oestrogen receptor gene (6q25.1-27) was to be examined in detail
including loci to which putative tumour suppressor genes have been mapped including the
M6P/IGF2R and TBP genes. LOH was then to be correlated with steroid receptor
expression, proliferative activity and pathological features of the tumours.
178
Additionally, the use of microsatellites allows the simultaneous evaluation of microsatellite
instability which if present was to be correlated with pathological features of the tumour
group. This was to contribute to a parallel study carried out by a fellow Ph.D. student, Tom
Walsh.
179
3.6
Materials A nd Methods
Source and grade of all chemicals, DNAs, radioisotopes, enzymes, antibodies and serum,
materials for electrophoresis, autoradiography, and commonly used solutions used in this
analysis were as described in section (2.12.1)
3.6.1.9 Patients
59 invasive, breast carcinomas which were impalpable and detected by mammography were
studied. All were from the first round o f mammographic screening and were detected by the
Leicestershire Breast Screening Service, and were 15mm or less in maximum diameter. All
had either axillary node sampling or axillary dissection. None of the tumours were from
women with a strong family history o f breast cancer or any known inherited predisposition
to the development of tumours. 56 cases were node negative.
16 cases of pure ductal carcinoma in situ (DCIS) were studied. These comprised 3 low, 3
intermediate, and 10 high nuclear grade cases. 10 of these were mammographically detected
and 6 clinically presenting. An additional 6 cases of pure DCIS were also studied at the
M6P/IGF2R locus, and were comprised of 6 low nuclear grade cases, three of which were
detected by mammography with the other three clinically presenting.
3.6.1.10 Tissues
All tissues were fixed in 4% formaldehyde in saline for 18-36 hours, dehydrated and
processed through paraffin wax. Following review of haematoxylin and eosin stained
sections, representative blocks were chosen for further study.
3.6.1.11 Histology
Carcinomas were reported according to the Royal College of Pathologists working party
guidelines (1990), and graded using the modified Bloom and Richardson system (Elston
and Ellis, 1991). All histology was undertaken by Dr R.A. Walker. Clinicopathological
features of the invasive tumours studied for LOH and MI at 6q are summarized in Table 3.3.
180
Table 3.3 Clinicopathological features of 59 paraffin embedded "early” sporadic, invasive
carcinomas
Type
Grade
No. of Cases
Tub
I
Lob/Tub
No. of Cases
6
Tumour Size
(mm)
< 10
I
1
10
14
IDC/ILC
II
1
11
4
ILC
II
1
12
5
IDC
I
17(1)
13
3
IDC
II
29 (2)
14
2
IDC
III
4
15
21
Total
59
10
59
Key Tub (tubular carcinoma), Lob/Tub (lobular and tubular carcinoma), IDC/ILC
(infiltrating ductal with infiltrating lobular carcinoma), ILC (infiltrating lobular carcinoma),
IDC Infiltrating ductal carcinoma, number in brackets, node positive cases.
181
3.6.2 Methods
3.6.2.1 Immunocytochemistry
Carcinomas were immunostained for ER, PgR and MIB-1 as described in Chapter 2 section
(2.13.2.3).
3.6.2.2 DNA Extraction
3.6.2.2.1 Whole Tumour Sections
Both normal and tumour DNA was extracted from formalin fixed, paraffin-embedded tissue
with non-involved lymph nodes (invasive group) or normal breast (non-invasive group)
serving as the source o f normal DNA. For each tumour-normal pair, DNA was extracted
from 3 x lOjaM paraffin-embedded sections. In brief, sections were dewaxed and dehydrated
by the sequential addition, mixing for 30 minutes and removal of 2 x 1ml xylene, 2 x 1ml
99% ethanol and 2 x 1ml 95% ethanol. Air dried pellets were resuspended in 250pl
digestion buffer (1 mg/ml Proteinase K in 50mM Tris-HCl, pH 8.0, 1% SDS) and incubated
for 3 days at 37°C. DNA was then extracted twice with phenol/chloroform and precipitated
from the resultant aqueous phase with 1/10th volume 3M sodium acetate (pH 4.8) and 2-3
times resultant volume o f cold 100% ethanol overnight at -70°C. DNA was then pelleted,
washed with 70% ethanol, air dried and resuspended in 200pl PCR grade water and allowed
to resuspend overnight at 4°C. Once resuspended, optical densities were measured at 260nm
and 280nm to assess quantity and quality of DNA.
3.6.2.2.2 Tumour Microdissection And DNA Extraction
The method used for DNA extraction from microdissected material was adapted from that of
Koreth et al., (1995). 10pm paraffin sections were deparaffinised in xylene ( 2 x 5 mins) and
dehydrated in 99% ethanol ( 2 x 2 mins) and 95% ethanol ( 1 x 2 mins), and rehydrated in
water before staining. Tissues were stained in 0.5% eosin solution for 20 seconds, washed in
water and allowed to air dry. A serial reference slide for each tumour was further stained
with haematoxylin, dehydrated and coverslipped. Tumour foci of interest were identified by
Dr R.A Walker and included tubular, solid, invasive lobular components and non-invasive
lesions o f DCIS within infiltrating ductal carcinomas and tumour foci from the invasive
lobular carcinomas. Individual lesions o f DCIS were the foci of interest from non-invasive
lesions. These areas were visualized using the H & E reference section and microdissected
from the corresponding eosin section using a x 40 magnification microdissection
182
microscope (American Optical Corporation) by the use of sterile. 20pm drawn out glass
capillary tubing (Figure 3.4). Glass tips, containing microdissected cells, were broken
directly into 0.5ml sterile eppendorf tubes containing 25pl of digestion buffer (lOOmM TrisHC1. pH 8.8; ImM EDTA, pH 8; 200pg/ml Proteinase K), and incubated at 55°C for 3
hours, then at 94°C for 10 minutes to deactivate Proteinase K. 5pl was then removed and
used as template for 25pl volume PCR.
3.6.2.3 PCR Analysis A t 6q25.1-27
59 invasive carcinomas and 16 non invasive breast carcinomas were studied for LOH in
comparison to corresponding normal DNA at three polymorphic markers from chromosome
6q25.1-27: the oestrogen receptor (ESR) at 6q25.1 (Del Senno et al., 1992), D6S186 (6q26)
and D6S193 (6q27) (Saito et al., 1992; Orphanos et al., 1995). 28 invasive breast
carcinomas were studied for LOH at the TATA box-binding protein (TBP) gene locus at
6q27 (Polymeropoulos et al., 1991; Saito et al., 1994) with 41 invasive breast carcinomas
and 22 non-invasive carcinomas studied for LOH at the M6P/IGF2R (Hoi et al., 1992).
Primer sequences and amplification conditions used are summarized in Table 3.4 and
appendix I respectively. Dinucleotide repeats comprised: (CA)n - D6S186 and D6S193, and
(TA)n - ESR, one trinucleotide repeat microsatellite (CAG)n - TBP, and one dinucleotide
repeat (GT)n with a tetranucleotide insertion deletion polymorphism
(ACAA) -
M6P/IGF2R.
PCR reactions were carried out in 35S PCR buffer with 0 .3 |l i 1 [a-35S] deoxyadenosine-5’triphosphate (600Ci/mmol, lOmCi/ml), lOpmol of forward and reverse primers, lOOng or
5pi microdissected DNA and 1 unit Taq DNA Polymerase in a total volume of 25pl. Hot
start PCR was carried out using the following cycles: 5 min. denaturation at 94°C, followed
by 35-40 cycles of 1 min. denaturation at 94°C, 1 min annealing, and 1 min extension at
72°C with a final extension o f 7 min at 72°C on a DNA Thermal Cycler (Perkin Elmer
Cetus, UK).
3.6.2.4 Polyacrylamide Gel Electrophoresis Analysis O f Labeled PCR Products
Labeled PCR products were electrophoreised through 6% denaturing polyacrylamide gels at
approximately 70W for 2-3 hours depending on fragment size. Gels were dried and exposed
183
Table 3.4 Microsatellite repeat sequence PCR primer sequences
Primer Set
Primer Sequence
6q Locus
Annealing
Temp (°C)
Size
(bp)
Reference
(5 ’- » 3 ’)
ESR1
F 5’-GAC GCA TGA TAT ACT TCA CC-3’
R 5’-GCA GAA TCA AAT ATC CAG ATG-3’
6q25.1
56
178-194
Del Senno et al, 1992
D6186
F 5’-TTA CCC ACT ACC TAC CCA GAG-3’
R 5’-GTC CCT TGG AAA ATT CTC CCT-3’
6q26-27
56
235
Orphanous et al, 1995
D6SJ93
F 5’-AGA GCA GGC TCT GCA TGG TTA-3’
R 5’-CTG ACA AAA GAA CAT ATT GTT TCC C-3’
6q26-27
56
190
Orphanous et al, 1995
F 5’-TTG CCG GCT GGT GAA TTC AA-3’
R 5’-GTA TCA TGA GAA CCT GAA GAG-3’
6q26-27
60
158-168
Hoi etal., 1992
F 5’-GAC CCC ACA GCC TAT TCA GA-3’
R 5’-TTG ACT GCT GAA CGG CTG CA-3’
6q26-27
65
185-206
Polymeropoulos et al, 1991
M6P/1GF2R
TBP
Key F denotes forward primer (sense); R denotes reverse primer (antisense)
184
to radiographic film for 1-7 days. LOH was considered to be present when the constitutive
DNA was heterozygous (informative) for the locus under investigation, and where there was
loss either partial (at least 50%) or total of an allele in the corresponding tumour DNA as
estimated by visual inspection. Shift, or gain, of alleles in the tumour sample served to
indicate microsatellite instability. Densitometric analysis was not considered necessary due
to almost complete loss of alleles in tumour DNA prepared by microdissection.
185
3.7.
Results
3.7.1 Loss O f Heterozygosity Analysis At 6q25.1-27
59 early invasive breast tumours and 22 non-invasive lesions of DCIS were screened for
LOH with 5 polymorphic microsatellite markers, mapping to chromosome 6q25.1-q27.
LOH was considered to be present when the constitutive tissue DNA was heterozygous
(informative) for the locus under investigation, and where there was complete or > 50% loss
of one allele in the corresponding tumour DNA as estimated by visual inspection (Table
3.5). The complex heterogeneity of the disease and the presence of non-tumour cells can
mask LOH, therefore all analyses were confirmed using DNA prepared by microdissection
from different histological tumour foci within the same tumour section. The use of
microdissected material produces almost complete allelic loss, such that densitometric
analysis of the data was not considered necessary. Representative examples of the
microdissection technique are shown in figure 3.4(a-b) for DCIS and figure 3.5(a-b) for an
invasive carcinoma.
Table 3.5 Heterozygosity frequencies for the invasive and non invasive tumour groups at 5
microsatellite loci
Locus
ESR
No. Cases Tested
DCIS
Invasive
15
57
No. Informative Cases (%)
Invasive
DCIS
32 (56)
9(60)
D6S186
54
16
23 (42.6)
8(50)
D6S193
57
16
35(61.4)
12(75)
TBP
28
-
13(46)
-
M6P/IGF2R
40
22
25 (62.5%)
18 (82%)
Examples of breast carcinomas demonstrating heterozygous and homozygous alleles at each
marker studied are shown in Figure 3.6(a-f) and Figure 3.7(a-f) respectively. Allelic
heterozygosity was detected as two different sized alleles representing differences in the
number of repeats in the microsatellite from each allele whereas allelic homozygosity was
detected as single bands representing an identical number of repeats in the microsatellite
from each allele.
An example demonstrating the benefits of a LOH using microdissected material is
illustrated in Figure 3.8(a-b) which compares LOH analysis of DNA prepared using a whole
186
F igure 3.4(a-b ) Representative microdissection o f individual ducts from a case o f ductal carcinoma in situ.
Individual lesions o f DCIS were visualized using a H&E reference section (Figure 3.4(a). Necrotic material
w as first removed (Figure 3.4(b) and tumour cells m icrodissected from a serial eosin stained section using a x
40 m agnification microscope and a sterile, 20pl drawn out glass capillary tube. Glass tips containing
microdissected cells were broken directly into digestion buffer.
F igure 3.4(a)
187
F igure 3.5(a-b ) Representative microdissection o f individual tumour foci from an invasive breast carcinoma.
Individual tumour foci were visualized using a H&E reference section (Figure 3.5(a) and microdissected from
a serial, eosin stained section (Figure 3.5(b) using a x 40 magnification microscope and a sterile, 20pl drawn
out glass capillary tube. Glass tips containing microdissected cells were broken directly into digestion buffer.
Figure 3.5(a)
188
ESR
T
D6S193
T
D6S1S6
N
N
Case 3
Case 43
Case 135
F ig u re 3.6(a)
F ig u re 3.6(b)
F ig u re 3.6(c)
M 6P/IG F2R
T
TBP
N
Case 13
Case 31
F ig u re 3.6(d)
F ig u re 3.6(e)
Figure 3.6(a-e) Representative carcinomas which were heterozygous for all microsatellite markers utilized
(D 6S186 Figure 3.6(a), D 6S193 Figure 3.6(b), ESR Figure 3.6(c), M6P/IGF2R Figure 3.6(d) and TBP Figure
3.6(e). Genom ic D N A samples from paired normal lymph node (N) and tumour (w hole sections T or
microdissected M ) samples were compared by PCR amplification, electrophoresis on 6% denaturing
polyacrylamide gels and autoradiography. A llelic heterozygosity was detected as two different sized alleles
(indicated by arrowheads) representing differences in the number o f repeats in the microsatellite from each
allele.
189
D6S193
D6S186
T
N
Case 43
Case 13
Case 111
F ig u re 3.7(a)
F ig u re 3.7(b)
F igure 3.7(c)
TBP
M 6P/IG F2R
T
N
Case 43
Case 21
F ig u re 3.7(d)
F ig u re 3.7(e)
Figure 3.7(a-e) Representative carcinomas which were homozygous for all microsatellite markers utilized
(D 6S186 Figure 3.7(a), D 6S193 Figure 3.7(b), ESR Figure 3.7(c), M6P/IGF2R Figure 3.7(d) and TBP Figure
3.7(e). Genomic D N A sam ples from paired normal lymph node (N ) and tumour (w hole sections T or
microdissected M) sam ples were compared by PCR amplification, electrophoresis on 6% denaturing
polyacrylamide gels and autoradiography. A llelic hom ozygosity was detected as single products (indicated by
arrowheads) representing an identical number o f repeats in the microsatellite from each allele.
T N
Case 3
Case 3
F ig u re 3.8(a)
F ig u re 3.8(b)
Figure 3.8(a-b) LOH analysis at the ESR locus illustrating the advantages o f the microdissection approach
compared with w hole tumour sections. Genomic DNA samples from paired lymph node (N ), w hole tumour
sections and m icrodissected tumour foci were compared by PCR amplification, electrophoresis on 6%
denaturing polyacrylam ide gels and autoradiography. N o LOH was observed using genom ic D N A prepared
from w hole tumour sections (Figure 3.8(a), but was revealed using genomic D N A prepared from
m icrodissected foci (Figure 3.8(b). A lleles are indicated by arrowhead with the sample demonstrating lost
indicated by asterix.
190
F igu re 3 .9 (a -h ) Representative m icrodissection LOH in m am m ographically detected, breast carcinom as.
G enom ic D N A sam ples from paired normal lymph node (N ) and m icrodissected tumour (M ) sam ples were
com pared by PCR am plification, electrophoresis on 6% denaturing polyacrylam ide g els and autoradiography.
LOH w as considered present when constitutive D N A was h eterozygou s (inform ative) for the locus in addition
to loss, either partial (at least 50%) or total o f an allele in the corresponding tumour D N A as estim ated by
visual inspection. Position o f alleles are indicated by arrowheads w ith lanes dem onstrating LOH indicated by
an astrix. Figure 3.9(a and b) LOH at D 6S186 in infiltrating ductal carcinom as 102 and 23 respectively; Figure
3.9(c and d) LOH at D 6S193 in infiltrating ductal carcinom as 7 and 37 respectively; Figure 3.9(e and f)i LOH
at ESR in infiltrating ductal carcinom as 21 and 102 respectively; LOH at TB P locus in tubular carcinom as 13
and 23 respectively.
a
—
......
■■■■■■ ■ '
Case 102
Case 23
Figure 3.9(a)
Figure 3.9(b)
•»
~
Case 7
Case 37
F igure 3.9(c)
Figure 3.9(d)
^
<N
S
£2
2
S
N
' « ' *<
J i t # ' ' '
f
Case 21
Case 102
F igure 3.9(e)
Figure 3.9(f)
* *
-
cs n
2
2
S
* *
—> (N
N
;H b | |
M
Case 13
Case 17
Figure 3.9(g)
Figure 3.9(h)
191
3
F ig u re 3 .1 0 (a -g ) Representative m icrodissection LOH in preinvasive lesions o f DCIS. G enom ic D N A
sam ples from paired normal lymph node (N ) and m icrodissected ducts (M ) sam ples w ere com pared by PCR
am plification, electrophoresis on 6% denaturing polyacrylam ide gels and autoradiography. LOH was
considered present w hen constitutive D N A w as heterozygous (inform ative) for the locus in addition to loss,
either partial (at least 50% ) or total o f an allele in the corresponding tumour D N A as estim ated by visual
inspection. Position o f alleles are indicated by arrowheads with ducts dem onstrating LOH indicated by an
astrix. Figure 3.10(a) LOH at D 6S 186 in high grade DCIS case D5; Figure 3.10(b and c) LOH at D 6S193 in
DCIS cases D5 (high grade) and D 4 (low grade) respectively; Figure 3.10(d and e); LOH at ESR in DCIS
cases D3 (low grade) and D5 (high grade) respectively; LOH at M 6P/IGF2R locus in DCIS cases D2 (high
grade) and D4 (low grade) respectively.
C a se D 5
F ig u r e 3 .1 0 (a )
* *
>—1
<N
m
2
S
N
C a se D 5
C ase D 4
F ig u r e 3 .1 0 (b )
F ig u r e 3 .1 0 (c )
— <N m
2
S
S
N
C a se D 3
C a se D 5
F ig u r e 3 .1 0 (d )
F ig u r e 3 .1 0 (e )
*
* *
<N m
N
C ase D 2
C ase D 4
F ig u r e 3 .1 0 (f )
F ig u r e 3 .1 0 (g )
10pm section o f tumour (Figure 3.8(a) with DNA isolated from a microdissected foci from a
serial section o f the same carcinoma (Figure 3.8(b) at the ESR locus. No allelic loss was
observed for DNA isolated from a 10pm section, whereas heterogeneity was observed for
the same tumour using microdissected material, with almost complete allelic loss detected at
an individual tumour foci.
Representative examples o f breast tumours demonstrating LOH at all microsatellite loci
examined are shown in Figure 3.9(a-h) for invasive breast carcinomas and Figure 3.10(a-g)
for preinvasive lesions o f DCIS. Figures 3.9(a and h) and 3.10(b) shows 2 invasive
carcinomas and one case o f DCIS that exhibit complete LOH for separate microdissected
foci at the markers studied. The tumours analyzed in Figure 3.9(d) and Figure 3.10(c) show
an invasive carcinoma and a case o f DCIS which some evidence of heterogeneity with
variation between different microdissected foci. For example, Figure 3.9(d) is an infiltrating
ductal carcinoma grade III that exhibits LOH at D6S193. Analysis of six distinct
microdissected foci shows 4 areas o f solid component of the tumour with LOH (M l, M2,
M4, and M6), and 2 areas o f solid component with heterogeneity (M3 and M5).
This
discrepancy could be attributable to the presence of contaminating non-neoplastic stromal
cells in some areas o f the carcinoma, even when it was dissected away from normal tissue.
Although microdissection analysis revealed occasional heterogeneity of distinct structural
components e.g. in situ, solid or tubular lesions within a tumour section, with some foci
showing clear LOH and others showing no evidence of LOH, no clear correlation was seen
between specific structural components and LOH at any particular locus.
Tables 3.6 and 3.7 summarize the observed patterns of LOH at 6q25.1-q27 for the invasive
and non-invasive study groups respectively.
LOH was seen for all types and grades of
disease that were studied. 24 o f 59 invasive carcinomas (48%) showed evidence of LOH.
Of these, 17 exhibited LOH only at a single locus (Table 3.6). The situation was similar for
the cases of DCIS with 8 o f 16 cases (50%) showing evidence of LOH and 5 of these only
exhibiting LOH at A single locus (Table 3.7). The frequency of LOH at individual markers
ranged from 23 % to 40.6% for the early invasive cases and from 33.3 % to 50 % for the
DCIS group (Table 3.8). The highest frequency of LOH was observed at the ESR locus for
the invasive carcinomas and at the D6S186 locus for the cases of DCIS. LOH was observed
in both high and low grade DCIS. 0 o f 25 (0%) cases demonstrated detectable LOH in 40
invasive breast carcinomas at the M6P/IGF2R locus with LOH detected in 4/18 (22%) of
193
the DCIS group. The DCIS group were not studied for LOH at the TBP marker due to the
paucity of available material for study.
The early invasive carcinomas were studied for oestrogen and progesterone receptor status,
and MIB-1 indices by immunohistochemistry. 53 (90%) were oestrogen receptor positive,
27 (46%) were progesterone receptor positive and MIB-1 (Ki67) proliferative indices ranged
from 0 to 59.9 with a mean index o f 11.25 (Raw data shown in Appendix III). 13 of the
early invasive carcinomas showed LOH at the ESR locus. Of these, 12 were ER positive
(92%) and 4 were PgR positive (31%) by immunohistochemistry (Table 3.6).
194
T ab le 3.6 Pattern o f loss o f heterozygosity and MI observed using 4 m icrosatellite m arkers from the 26 early invasive breast carcinom as in which the
alterations were observed
C ase
No.
3
5
7
13
15
17
19
21
23
29
31
37
41
49
55
57
59
70
76
78
80
98
102
106
108
122
Type
ILC
1DC/ILC
IDC
II
11
I
Tub
IDC
Tub
IDC
1
I
I
11
II
I
I
I
III
II
II
IDC
IDC
Lob/Tub
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
Tub
IDC
IDC
IDC
IDC
IDC
U
I
I
I
I
I
I
II
u
11
III
II
H S cores
Loss o f H eterozygosity at M arkers
G rade
M IB -l
ESR (q 25.1)
D 6S186 (q26)
D 6 S I9 3 (q27)
T B P (q27)
M 6P (q 26-27)
ER
PgR
(% )
•
NI
O
O
MI
O
o
O
•
O
NI
NI
•
o
o
o
20.8
15.6
•
175
215
182
106
•
O
Ml
NI
142
181
94
o
o
O
NI
•
•
•
MI
NI
NI
NI
MI
•
•
•
•
NI
NI
•
•
•
•
NI
MI
O
•
NI
NI
O
O
•
F
NI
•
NI
NI
•
NI
•
NI
NI
NI
NI
NI
o
0
•
•
o
•
NI
NI
NA
NI
o
o
o
o
o
o
o
o
o
O
NI
•
•
NI
MI
O
•
NA
NA
O
•
•
•
•
•
•
•
O
O
NI
NT
NT
NI
NI
NI
NT
NI
NT
NT
NT
NI
NI
Ml
o
o
NT
NT
NT
NT
NT
NT
NT
NT
NT
195
162
151
175
196
202
22
192
215
183
173
196
231
219
202.5
162
187.5
142
195
0
160
49
49
97
3
168
92
31
0
115
3
170
46
234
192
19.2
0
0
85
83
0
0
33.5
0
0
NTA
1
6.3
1.5
2.5
15.4
12.3
1.9
118
1.5
6.8
23.7
3.3
15.4
22
5.1
8.2
8.9
15.4
12.5
15.8
7
25.8
3
Key • , loss of heterozygosity; O, heterozygosity; MI, microsatellite instability; NI, not informative; NA, no amplification; NTA, no tissue available; NT,
not tested, Tub (tubular carcinoma), Lob/Tub (lobular and tubular carcinoma), ILC (infiltrating lobular carcinoma), IDC (infiltrating ductal carcinoma).
195
T a b le 3.7 Pattern o f loss o f heterozygosity and MI observed using 5 m icrosatellite m arkers in the 12 non-invasive lesions o f DCIS in which the
alterations were observed
Case No.
Chromosome 6q Markers
Grade
ESR (q25.1)
D6S186 (q26)
D6S193 (q27)
M6P (626-27)
TBP(q27)
NT
D2 (M)
High
NA
•
MI
D3 (M)
Low
•
NI
D4 (M)
Low
NI
D5 (M)
High
D7 (M)
High
MI
MI
D5 (M)
High
•
O
•
•
•
O
•
•
•
•
D8 (M)
Low
NI
NI
•
DIO (M)
Inter
NI
NI
MI
D12 (C)
High
NI
•
D13 (C)
High
•
NI
•
O
•
O
•
o
o
o
o
o
o
•
D14 (C)
High
NI
NI
•
•
NT
D17(C)
High
NT
NT
MI
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
Key • , loss of heterozygosity; O, heterozygosity; NI, not informative; NA, no amplification; NT, not tested; (M), mammographically-detected; (C),
clinically presenting.
196
Table 3.8 Summary of chromosome 6q LOH data in early invasive tumours and non-invasive cases of DC1S
Locus
Chromosomal
Location
No. Cases Tested
No. Informative Cases (%)
No. Cases with LOH (%)
DCIS
15
Invasive
32 (56)
DCIS
9(60)
Invasive
13(40.6)
DCIS
3 (33.3)
ESR
6q25.1
Invasive
57
D6S186
6q26
54
16
23 (42.6)
8(50)
7 (30.4)
4(50)
D6S193
6q27
57
16
35(61.4)
12 (75)
11(31.4)
5(41.7)
TBP
6q27
28
-
13(46)
3(23)
-
M6P/IGF2R
6q26-27
40
22
25 (62.5)
0(0)
4(22)
-
18(82)
Key TBP(TATA box binding protein), M6P/IGF2R (mannose 6-phosphate/insulin-like growth factor 2 receptor), DCIS (ductal carcinoma in situ), LOH
(loss of heterozygosity)
197
3.7.2 Microsatellite Instability Analysis At 6q25.1-27
As already described, as a consequence o f using microsatellite repeats to define regions of
common deletion, it is also possible to simultaneously detect tumours exhibiting
microsatellite instability. Three dinucleotide repeat microsatellites (ESR, D6S193 and
D6S186), one dinucleotide repeat microsatellite with a tetranucleotide deletion/insertion
polymorphism (M6P/IGF2R) and one trinucleotide repeat microsatellite (TBP) were used
for analysis o f the early invasive group. The non-invasive tumour group were not analyzed
at the TBP due to paucity o f non-invasive tumour tissue. Microsatellite instability was
deemed to have occurred when one or both alleles, in the tumour, shifted in size relative to
the alleles from the corresponding normal DNA.
The results for the invasive and non-invasive tumour groups are summarized in Tables 3.6
and 3.7 respectively. In the invasive group, two tumours (cases 5 and 55) demonstrated
instability at 3 o f the 4 loci, with one other (Case 31) demonstrating instability at the ESR
locus alone. In the non-invasive group, one tumour demonstrated MI at 2 of the 4 loci, Case
D7 (D6S186 and D6S193), with three others demonstrating instability at the D6S193 locus
alone (Cases D2, DIO and D17). Invasive tumours exhibiting MI were all infiltrating ductal
carcinomas, well to moderately differentiated and oestrogen receptor positive, with the noninvasive tumours exhibiting MI being predominantly of high grade (3 out of 4) with one of
intermediate grade.
Representative examples o f microsatellite instability in the invasive carcinomas group are
illustrated in Figure 3.1 l(a-d) showing MI in two invasive carcinomas (cases 5 and 55) at
the D6S193 locus (Figure 3.11 (a and b); MI in three invasive carcinomas (cases 5, 31 and
55) at the ESR locus (Figure 3.11(a-c); and MI in one invasive carcinoma (case 55) at the
M6P/IGF2R locus Figure 3.11(d).). Two types of instability can be seen. The first type of
instability is where both alleles have contracted, e.g., case No. 55 at the ESR locus (Figure
3.11(c); and the second type o f instability where individual alleles have expanded or
contracted, e.g., cases 5 and 31 at the ESR locus (Figure 3.11(c), case No. 5 at the D6S193
locus (Figure 3.11(a) and case No. 55 at the M6P/IGF2R locus (Figure 3.11(d).
Representative examples o f microsatellite instability in the non-invasive tumour group are
illustrated in Figure 3.12(a-d) showing MI in 3 non-invasive tumours; D7 at both the
D6S186 and D6S193 loci (Figure 3.12(a and b) respectively) and DIO and D34 at the
D6S193 locus alone.(Figure 3.12(d)).
198
T
N
T N
►
p S
►
T
Case 5
Case 55
Figure 3.11(a)
Figure 3.11(b)
T
N
T
N
N
►
|J
&
Case 5
►
►
Case 31
Case 55
Figure 3.11(c)
Case 55
Figure 3.11.(d)
Figure 3.11(a-d) Representative m icrosatellite instability in mammographically detected breast
carcinomas. DNA samples from paired normal lymph node (N) and tumour (T) samples were
compared by PCR amplification, electrophoresis on 6% denaturing polyacrylamide gels and
autoradiography. M icrosatellite instability w as deemed to have occurred when one, or both alleles in
tumour D N A shifted in size relative to the alleles from constitutive DNA from the same individual.
Expansion (a) and contraction (b) o f alleles in tumour D N A (ILC/IDC and IDC respectively) at
D6S193; Instability at the ESR locus (c) with contraction (case 55), and contraction with expansion
(cases 5 and 31) o f alleles in tumour DNA (IDC, IDL/IDC tumour types respectively); and expansion
(d) o f alleles in tumour D N A (IDC) at the M6P/IGF2R locus. The position o f alleles are indicated by
arrowheads (wt alleles to the right (N ) tumour alleles to the left (M)).
199
Case D7
Figure 3.12(a)
S
CN
CO
2
2
N
Case D7
Case DIO
Figure 3.12(d)
Figure 3.12(c)
* *
— <N
S 2 N
Case D34
Figure 3.12(d)
F igure 3 .1 2 (a -d ) Representative microsatellite instability in mammographically detected breast carcinomas.
D N A sam ples from paired normal lymph node (N ) and tumour (T) samples were compared by PCR
am plification, electrophoresis on 6% denaturing polyacrylamide gels and autoradiography. Microsatellite
instability w as deem ed to have occurred when one, or both alleles in tumour DNA shifted in size relative to
the alleles from constitutive D N A from the same individual. Expansion (a) and contraction (b) o f alleles in
tumour D N A (ILC/IDC and IDC respectively) at D6S193; Instability at the ESR locus (c) with contraction
(case 55), and contraction with expansion (cases 5 and 31) o f alleles in tumour D NA (IDC, IDL/IDC tumour
types respectively); and expansion (d) o f alleles in tumour DNA (IDC) at the M6P/IGF2R locus. The position
o f alleles are indicated by arrowheads ( wt alleles to the right (N) tumour alleles to the left (M )).
200
3.8
Discussion
3.8.1
Loss of Heterozygosity at Chromosome 6q25.1-27
Genetic alterations are believed to play an important role in the evolution of breast cancer
(Sato et al., 1991). Frequent loss of heterozygosity in breast cancer, implying the presence
of tumour suppresser genes, has been detected on many different chromosomes.
Deletions o f parts o f the long arm of 6q have previously been reported in several
malignancies suggesting the presence of one or more tumour suppresser genes, evidence
which is supported by chromosome-mediated transfer experiments (Trent et al., 1990;
Yamada et al., 1990; Negrini et al., 1994). The combined LOH data in breast cancer at all
chromosomes has implied that chromosome 6q is the second most frequent site after 17p for
LOH in breast cancer. As many of these studies have concentrated on the well established,
invasive tumours, I have carried out deletion analysis on 59 mammographically screen
detected, early invasive carcinomas, and 22 non invasive lesions of DCIS, using highly
informative dinucleotide and trinucleotide microsatellite repeat sequences to assess whether
allelic loss at this chromosome is an early event in the development of breast cancer.
Using microdissection o f distinct structural components from within a tumour tissue section
and PCR amplification o f microsatellite repeats, I have demonstrated loss of heterozygosity
(LOH) at chromosome 6q25.1-27 in foci of both DCIS and “early” invasive carcinomas.
Comparing the proportion o f in situ lesions with the proportion o f invasive lesions
exhibiting LOH at each locus revealed similar frequencies. Moreover, there was a general
spread of LOH detected for all types and grades of disease studied. These data for the
“early” carcinomas suggest that the majority of allele losses previously reported at these loci
in symptomatic invasive cancers (Orphanos et al., 1995; Iwase et al, 1995) can be found in
pre-invasive carcinomas. This suggestion is supported by evidence from a small number of
the infiltrating ductal carcinomas where it was possible to analyze an invasive and in situ
component from the same tumour section (not shown). In each case LOH was detected in
both lesions. In combination, these data suggest that loss of alleles on chromosome 6q is an
early event in the progression o f malignancy in the breast.
The informativity of each marker was 73% (ESR); 72% (D6S193); 50% (D6S186); and
69.3% (M6P/IGF2R) and were lower than those previously reported for ESR, and D6S186
(82%, and 72% respectively), considerably higher for M6P/IGF2R (58%) and identical for
201
D6S193 (72%). These differences may represent different tumour populations between
studies and also differences in the number o f cases analyzed.
Although the highest frequency o f LOH was observed at the ESR locus (40.6%) for the
invasive carcinomas and at the D6S186 locus (50%) for the cases of DCIS, these differences
may merely reflect the different groups o f cases studied and the difference in sample size
between the two groups. In addition, the slightly higher frequency of LOH noted for the
cases of DCIS may reflect the fact that most were of high nuclear grade, and therefore a
more aggressive disease type. It is interesting to note that LOH was found for all three
markers studied in both high and low grade DCIS, suggesting the early involvement of loci
on 6q in the development o f these lesions. In a study of chromosome 1, differences were
found between chromosomal regions for the different sub-types of DCIS with no alteration
at two regions in low grade DCIS (Munn et a l, 1995).
LOH analysis at the ESR, D6S193 and D6S186 locus in this study indicate that allelic loss
at these loci occur at an early stage in breast cancer development. The frequency of LOH at
the D6S193 and D6S186 loci are slightly higher, but generally agree quite well with those
previously reported (Orphanos et al., 1995). The frequency of LOH at the ESR locus in this
study (55%) was considerably higher than that described by Iwase et al., (1995) who
observed 19.1% LOH at this locus.
The significant difference in observed LOH frequencies may well reflect the difficulty in
interpreting LOH data. Three main problems confuse the comparison of the genotypes in
normal and tumour DNA from the same patient i) interpretation of ‘allelic imbalance’ as
loss or gain; ii) tumour heterogeneity; and iii) the presence of non-malignant cells. The first
o f these i), is that microsatellite markers will reveal any imbalance in parental alleles, i.e.,
loss and gain in allele copy number. In addition, this problem is compounded by the fact that
with the methodology used, it is very difficult to ensure that the amounts of normal and
tumour DNA are the same. Also there could be unpredictable fluctuations in the efficiency
of the PCR reaction itself. Consequently, it is difficult to discriminate between the loss and
gain although in a number o f controlled studies, it appears that about 80% o f observed
allelic imbalances are in fact allele losses (Devilee and Comelisse, 1994). Although this is a
problem when interpreting LOH data, it is relatively unimportant when trying to explain the
significant differences in LOH at the ESR locus in this study and that by Iwase and
202
colleagues. The use o f microdissection allows complete loss to be demonstrated and thus
overcomes this confusion, but it is only as good as the hand of the individual carrying out
the technique.
A factor which may contribute to this discrepancy to a greater extent, is the nature of tumour
heterogeneity. An illustration o f this problem can be seen by the work of Stratton et al.,
(1995) where LOH was seen in certain foci, e.g., in in situ ductal carcinoma (a non-invasive
component) but less obvious or undetectable in the invasive component. In some cases, the
allele lost in the DNA from the in situ lesion was different from that apparently lost in the
invasive cancer, raising the possibility that genetically distinct subclones exist within a
single specimen. Therefore it can be seen that the heterogeneity of tumours results in
diluting the actual allelic loss in tumours.
Consequently this may have the effect of
lowering LOH frequency.
Thirdly and probably the most influential problem in interpreting LOH data is the presence
of non-malignant cells within tumour which, i) reduces the proportion of tumour cells, and
ii) obscures any LOH present in the tumour cells. Devilee and Comelisse (1994) empirically
established that tumours with >50% tumour cells gave reliable results in LOH screening
whereas Stratton et al., (1995) determined that the presence of >20% non-malignant cells
made LOH difficult to detect. These studies emphasize the importance of enriching for
tumour DNA.
The problems discussed illustrate the difficulty in interpreting LOH data, and thus may
explain the differences in LOH frequency described in this study compared to that of Iwase
and colleagues. Additionally, the analysis o f different tumour populations, and the
subjectivity o f visually determining LOH may have also contributed to this lack of
agreement o f results.
The prevalent detection o f LOH at a single locus rather than multiple loci in both the “early”
invasive carcinomas and cases o f DCIS argues against random losses due to general
chromosomal instability and gross chromosomal alterations. Some invasive carcinomas and
cases o f DCIS showed LOH at more than one locus. Since the markers studied map from
6q25-27, a distance o f several megabases, it is not possible to say whether a contiguous
region harbouring the relevant loci has been lost, or whether there are distinct areas of LOH
203
within this region o f chromosome 6q. In order to determine this, analysis using a much
higher density o f markers would need to be carried out.
Knowledge o f the ER status o f a carcinoma is of value in aiding prediction of hormone
responsiveness and can provide some prognostic information. Tumours lacking ER and
PgR generally grow faster than those containing both ER and PgR (McGuire and Clark,
1989).
46.1% o f informative cases for the invasive carcinomas exhibited LOH at the
oestrogen receptor locus (ESR), and 33% of the informative cases of DCIS showed LOH at
ESR. These frequencies are higher than the 19% LOH at ESR observed by Iwase et al.
(1995) and may reflect different groups of cases, or might be due to the more informative
analysis o f material prepared by microdissection in this study. The high incidence of LOH
at the ESR locus in the invasive carcinomas was not reflected by loss of ER at the protein
levels as detected by immunohistochemistry. Indeed, the majority of the group of early
invasive lesions were ER positive. This would be expected since the group studied were
predominantly well or moderately differentiated. Evolving tumour cells may later acquire
new proliferative pathways as a consequence of multiple genetic alterations, enabling the
tumour cells to by-pass oestrogen-dependent proliferation (Liu et al., 1988).
Other studies have found no relationship between LOH on chromosome 6q and ER status
(Devilee et al., 1991; Iwase et al., 1995; Noviello et al., 1996) suggesting that allele loss
may not play an important role in the lack of ER function in breast cancer tissues. However,
our results have identified 9 o f 13 tumours exhibiting LOH at ESR that were ER positive
and PgR negative. These results are similar to the findings of Noviello et al., (1996) who
demonstrated that LOH in the region 6q23-24, which although not clearly colocalized with
the ESR locus was associated with a PgR negative phenotype. This might indicate
mutational inactivation o f the remaining ESR allele leading to production o f an inactive but
detectable ER protein, and hence loss o f PgR. These cases are candidates for screening for
either mutations or spliced variants o f the oestrogen receptor gene. Alternatively it may be
that loss o f one allele o f the ESR has a direct consequence on the level of ER, which would
then act as a cascade on the PgR expression in these tumours (Noviello et al., 1996).
Only three invasive carcinomas showed evidence of LOH at the TBP locus (23%), at 6q27.
This frequency o f LOH is only marginally raised above expected levels for random
background loss. These data suggest that inactivation of this region of the chromosome is of
lesser importance than o f that harboring the ESR, D6S186 and D6S193 loci in these early
lesions.
I examined DNA obtained from 40 invasive breast carcinomas and microdissected tumour
foci from 25 informative cases and found no evidence of LOH (0%) at the M6P/IGF2R
locus mapped to 6q26-27. However, three other loci that map to 6q25.1-q27 show evidence
of frequent LOH in the same population of carcinomas. This would tend to suggest that
discrete regions o f this region o f the chromosome are lost rather than whole chromosomal
arms, but a more detailed analysis would be required to fully address this feature.
Normally in LOH studies o f cancer, a background level of LOH of 5-15% is usually
observed (Sato et a l, 1990; Chen et al., 1992) which are randomly acquired and irrelevant
to tumour development. This background incidence o f random allelic losses are believed to
be higher in late stage lesions because they have a longer time for randomly acquired lesions
to be co-selected with other mutations which confer a selective advantage associated with
malignant progression. The frequency o f LOH at the M6P/IGF2R locus in the early,
invasive carcinoma group falls below the values described by Sato et al., (1990) and Chen et
al, (1992), and reflects their findings that random chromosomal losses are rare in the early
stages of the disease.
An additional factor which cannot be ignored and which may interfere with detection of
LOH, is that polymerase amplification o f dinucleotide repeats is known to produce slippage
bands below the true allele (Louis et al., 1992). Given that in a high proportion of
informative cases the individual alleles differ by only 2 base pairs at the M6P/IGF2R locus,
these slippage bands would tend to mask the loss of the lower allele and as a consequence
mask the true frequency o f allelic loss. As a result it has been reported that frequency o f
LOH associated with the bottom allele is lower than that for the top and thus the true
frequency is higher then that seen (De Souza et al., 1995a).
I observed a high frequency o f heterozygosity at the M6P/IGF2R locus among both the
invasive (62.5%)
and
non-invasive
(82%) tumour groups. The original
reported
heterozygosity was only 58% (Hoi et al., 1992) and others have observed similar
frequencies, e.g., 65% (Hankins et al., 1996). Slippage bands could account for this
205
discrepancy in heterozygosity frequency making data interpretation difficult leading to the
scoring of some cases as informative which may actually be non-informative. A large
number o f amplification cycles (40) would exacerbate this slippage problem, consequently
my analysis was repeated using 30 amplification cycles. This improved the assessment of
heterozygosity and the detection of LOH in the groups analyzed with the frequency in the
invasive group a little above the frequency seen by others. The high frequency of
heterozygosity in the DCIS group can be accounted for by the analysis of a small number of
tumours. As a consequence the frequency of any event occurring is sensitive to small
variations. Therefore, analysis o f a much larger group could possibly result in a
heterozygosity frequency nearer to that previously described by other workers.
Hankins et al., (1996) demonstrated 33% and 26% LOH in this region in invasive and noninvasive breast tumours respectively. No clinicopathological data was given regarding the
invasive carcinomas analyzed, consequently no correlations with LOH could be made. LOH
in the non-invasive tumours was almost exclusively associated with comedo type carcinoma
in situ (21.5%), which is equivalent to high grade DCIS using the histological classification
used by ourselves, which would tend to suggest that loss at the M6P/IGF2R is associated
with a more aggressive phenotype. High grade DCIS is believed to be a precursor of
invasive carcinomas and is associated with an aggressive phenotype (e.g., steroid receptor
negative, high proliferation rates, high nuclear pleomorphism. greater risk o f local
recurrence and associated with invasion). Given that the tumour population examined in my
study comprised of tumours with a non aggressive phenotype, it may not be so surprising
that I detected LOH at the M6P/IGF2R locus at a low frequency suggesting that loss may
well be a pathway specific alteration, associated with a more aggressive tumour behaviour.
A good example o f a molecular change associated with high grade DCIS is the
overexpression o f the oncogene c-erb-B2 (Bartkova et al., 1990).
My LOH analysis was extended to the non-invasive tumour group, of which 10 of the 22
were o f high grade DCIS, at the M6P/IGF2R locus. 4 out of 18 (22%) informative cases
demonstrated LOH, three high grade DCIS and one low grade DCIS. I observed a lower
frequency of LOH than that observed by Hankins et al., (1996) who observed 5 of 19 (26%)
cases o f DCIS which again may reflect a small sample population which can be distorted by
small variations in positive results.
206
Other mechanisms of tumour suppresser gene inactivation cannot be discounted, including,
alternative
splicing,
promoter
and
intragenic
mutations.
Alternatively,
aberrant
hypermethylation of 5' CpG islands within proximal promoter regions has been implicated
as a mechanism by which tumour suppresser genes can be inactivated (reviewed in Jones et
al, 1996). Examples of this mechanism have been demonstrated for E-Cadherin (Graff et
al., 1995), for the VHL and p i 6 tumour suppresser genes (Herman et al., 1994; Merlo et al.,
1995). Consequently it would be of interest to investigate the CpG island methylation status
within regulatory regions o f the M6P/IGF2R gene.
In summary, I have detected frequent LOH at five polymorphic loci from chromosome
6q25.1-27, in both cases o f high and low grade DCIS, and all types and grades of early
invasive carcinomas. In combination these data confirm distal chromosome 6q as a major
site for genetic change in the early stages of development of some sporadic breast cancers.
3.8.2
M icrosatellite Instability
The accumulation o f widespread alterations (insertions and deletions of one or a few
nucleotides) in simple repeated sequences of a tumour cell genome is the diagnostic feature
o f cancers o f the microsatellite instability phenotype (MI+) (Ionov et al., 1993; Perucho et
a l, 1994). MI+ tumours represent a well characterized example of the cancer as a mutator
phenotype hypothesis (Loeb et al., 1991, 1994). Mutations in these unstable repeat
sequences (reviewed in Eshelman and Markowitz, 1995) accumulate because of the failure
by the DNA mismatch repair machinery (MSH2, MLH1, PMS2, PMS1, and MSH6) to
correct the errors o f replication due to slippage by strand misalignment (Streisinger et al.,
1966). Mutations targeted in MI+ tumours differ from those in tumours of the classical
“suppressor pathway” (Fearon and Vogelstein, 1990) because they contain gene targets for
slippage-induced frameshift mutations. For example, the TGFfiRII, IGF2R, and Bax genes
contain polynucleotide regions within coding regions which are subject to one or two base
insertions or deletions in a high proportion of sporadic gastrointestinal and urogenital MI+
tumours (Myeroff et al., 1995; Ouyang et al, 1997; Rampino et al, 1997).
Although well characterized in the colon, MI+ tumours of breast are less well understood.
One o f the problems associated with the analysis of microsatellite instability in breast cancer
is that clear criteria to define the phenomenon have not been established. In this thesis, a
tumour which demonstrated alterations at two or more independent genomic locations was
207
described as having the microsatellite instability phenotype and designated as MI+, as first
defined by Aaltonen et al., (1993).
The DNA template procured from microdissected tumour foci produced consistent
amplification o f the microsatellites used in this study. In general, microdissection of a single
involved duct or group o f carcinoma cells produced sufficient template for analysis with
multiple markers. However, this method is only as good as the hand/eye of the investigator.
One problem associated with MI analysis of breast tumours is that, to date, each
investigation has used a different panel of microsatellite markers, therefore making
comparisons between different groups of tumours and different tumour types very difficult.
Other problems for the analysis o f invasive carcinomas relate to the investigation of
unselected groups, for example early and advanced groups combined. In order to establish
the role o f MI in early stages of invasive tumours, small (less than 15mm) node negative
tumours were investigated. The analysis in this group was initially performed on DNA
prepared from whole tumour sections, similar to the majority of other studies performed on
invasive carcinomas (summarized in Table 3.2). The analysis with chromosome 6q markers
identified two tumours (cases 5 and 55) that displayed instability at all of the 6q loci tested.
A more detailed investigation was carried out by a fellow Ph.D. student, Tom Walsh, on this
tumour group using an extended panel of microsatellites and demonstrated that these two
carcinomas displayed instability at 9 of 10 microsatellites mapping to 5 chromosomes
(Shaw et a l, 1996).
Breast carcinomas with similar widespread alterations have also been reported by other
investigators. For example, Contegiacomo et al., (1995) characterized a tumour with
alterations to 6 o f 8 markers, and Toyama et al., (1996a), demonstrated alterations to 8 of 12
loci in one tumour in their study. These cases have instability similar to that reported in the
HNPCC kindreds, however, the pattern of instability is different. In contrast to HNPCC
patients, these carcinomas were unstable at a microsatellite in the 3’ UTR of the gene
responsible
for
myotonic
dystrophy,
not
unstable
in
HNPCC
tumours.
Also,
mononucleotides are generally unstable in HNPCC whereas in these tumours they were on
the whole stable. In tumours from HNPCC patients there is generally a ladder of unstable
alleles which results in ‘smearing’ of the PCR products. The tumours described here
generally had two new alleles not present in the corresponding normal sample. The
208
possibility o f a sample mix up was discounted by repeating the analysis with alternative
tumour blocks from the same patient, since heterozygous stable alleles were observed for at
least one other locus.
The group o f invasive carcinomas studied were predominantly infiltrating ductal, with a
small number o f tubular carcinomas or lobular-tubular carcinomas. MI+ was only identified
in the infiltrating ductal carcinomas. Other studies have included infiltrating lobular
carcinomas and have correlated the presence of MI with this phenotype (Aldaz et al., 1995;
Contegiacomo et al., 1995).
Microdissection o f individual foci from each of the carcinomas confirmed the initial MI
results obtained from DNA prepared from an entire tumour tissue section, and did not lead
to the identification o f any additional cases with alterations. This suggests that
microdissection may not be necessary for the analysis of infiltrating carcinomas where the
majority o f cells in the tissue section are malignant, in contrast to the non-invasive
carcinomas where there are high proportions of non-malignant cells surrounding malignant
foci (described below).
The benefits o f microdissection analysis were particularly evident when studying MI in the
group of DCIS. Microsatellite analysis in these lesions using DNA template prepared from
whole tissue sections revealed no alterations, as would be expected since tumour cells may
often represent a low percentage o f the total cells from cases of this type.
Microdissection enabled enrichment for tumour cells and the analysis of multiple ducts from
the same tumour section. The type o f MI was similar to that seen in the early invasive
carcinomas. Four cases demonstrated MI+, three of which showed alteration of the D6S193
marker and a trinucleotide repeat, and one case which was altered at both D6S193 and
D6S183 markers (Walsh et al., submitted). Although three of the four cases were of high
nuclear grade it is difficult to interpret this association since the tumour population was
predominantly o f high grade. However, analysis by Tom Walsh of equal numbers of high,
intermediate and low grade cases (total 23) has confirmed this association.
Aldaz et al., (1995) studied MI in 23 cases of pure DCIS, and found a MI+ frequency of
13% (3 o f 23). Unfortunately, the markers showing instability in these cases were not
described. The nuclear grade o f the cases was also not reported. However, the data recently
209
presented by Dillon et al, (1997) seems to contradict these findings. They examined 33
cases of DCIS with a panel o f microsatellites, and found only one case to be altered.
However, it is not clear from this paper whether microdissected tumour foci or DNA from
whole sections were used. Considering the evidence presented in this thesis it seems likely
that the vast quantities of non-malignant cells present in cases of pure DCIS would lead to
‘masking’ o f tumour specific alterations.
Microsatellite instability was detected at a higher frequency in the high and intermediate
grade cases, compared to the low grade cases. It is not possible to demonstrate correlation
between MI+ and grade from the other DCIS studies of Aldaz et al, (1995) and Dillon et al,
(1997). However, MI+ in invasive breast carcinomas has been correlated with indicators
commonly associated with poor disease prognosis, such as involvement of lymph node (De
Marchis et al, 1997), and reduced overall survival (Paulson et al, 1996).
The data from the DCIS cases together with evidence from other investigations suggests that
in cases o f DCIS and invasive breast carcinomas instability is associated with high nuclear
grade and an aggressive phenotype. This is also supported by our findings of low frequency
of MI+ in well to moderately differentiated early invasive carcinomas (Shaw et al, 1996).
Despite a large number o f MI studies in breast cancer, there have been no reports correlating
this phenotype with alterations in the mismatch repair proteins or any of the cancerassociated loci containing targets for slippage-induced frameshift mutations. The only data
reported thus far is a frameshift mutation in MLH1 in a MI+ breast tumour from a patient
within a HNPCC kindred (Risinger et a l, 1996). This was the first evidence that breast
cancer, albeit as an integral tumour within the HNPCC syndrome, may result from the
inheritance o f a mutant MMR gene.
Following the identification and characterization of MI+ in various breast malignancies, it
would be important to determine whether the observed instability is a consequence of
defects to the mismatch repair proteins, MSH2, MLH1, PMS1 or PMS2. This could be
achieved by examining the tumour specimens for the presence or absence of mismatch
repair protein expression by immunohistochemistry (Thibodeau et al,
1996). This
investigation examined the expression pattern of MSH2 and MLH1 in a large series of
patients with colorectal cancer (both HNPCC and sporadic colorectal cases). The results
demonstrated that an absence of protein expression for both proteins was associated with the
210
presence o f MI+. The technique therefore has the potential to be a rapid and simple screen
for identifying defects in mismatch repair genes.
It would also be o f interest to determine whether the MI+ breast tumours contain mutations
at the repeated motifs within the coding regions of TGFftRII, IGF2R or BAX, as has recently
been described in gastrointestinal MI+ tumours (Myeroff et al., 1995; Ouyang et al., 1997;
Rampino et al., 1997). PCR-based assays exist for each of these genes and can readily be
applied to paraffin embedded tumour samples. Another particularly good candidate for
future mutation analyses in MI+ breast tumours is the oestrogen receptor. Exon 1 of the ER
gene contains a combined trinucleotide and mononucleotide repeat motif, which has been
shown to be unaltered in MI+ gastrointestinal tumours (Simms et al., 1997).
In the absence o f any “targeted” mutations, the significance of the phenomenon of
miscrosatellite instability in breast cancer remains unclear. It may merely reflect evolving
genomic instability in tumours. However since all but one of the MI+tumours did not show
any evidence o f LOH, this might indicate that distinct “mutator” and “suppressor” pathways
indeed exist in breast cancer development and progression, as first demonstrated in
colorectal cancer, and therefore warrant further investigation.
211
4.
CON CL USIONS AND FUTURE WORK
4.1
Conclusions
Alterations to the ER gene, are frequently observed in breast cancers and these abnormal
forms could be implicated in the evolution of breast cancer. This thesis addressed the
frequency o f variant/mutant ERs in mammographically detected, node negative, small,
moderately or well differentiated, invasive carcinomas by RT-PCR and SSCP analyses. A
number o f exon splice variants were detected with deletions of exon 3 ,5 , and 7 in addition
to three missense mutations. Notably there was a high incidence of variants in grade III,
steroid receptor negative tumours and is in contrast to other studies of larger tumours with
metastatic disease, with a low incidence of the exon 5 deletion variant. These data suggest
that acquisition o f altered splicing profiles, with the acquisition of the ER exon 5 deletion
variant, is involved in certain more aggressive pathways of development/progression of
breast cancer.
Using microdissection and amplification of microsatellite repeats, I demonstrated significant
loss of heterozygosity at chromosome 6q25.1-27 in both preinvasive ductal carcinoma in
situ and early invasive carcinomas. There was a general spread of LOH detected for all types
and grades o f disease studied indicating that loss of alleles on chromosome 6q is an early
event in the development and progression of malignancy of the breast and must be involved
in multiple pathways. Two other specific alterations are noteworthy. Firstly the appearance
of altered length alleles (or microsatellite instability) in tumour DNA which correlates with
more aggressive features, e.g., high grade ductal carcinoma in situ, but must be an early
event since all ducts examined showed the same alteration. Secondly, alterations to a
candidate tumour suppressor gene, the mannose 6-phosphate/insulin-like growth factor 2
receptor (M6P/IGF2R), was only noted in high grade ductal carcinoma in situ, again
suggesting that inactivation may occur only in certain, more aggressive subgroups, (poorly
differentiated cases) o f breast cancers. Since I have detected frequent LOH within this
chromosome interval in all types and grade of disease, this must suggest the inactivation of
other, as of yet uncharacterized tumour suppressor genes on chromosome 6q, as well as the
M6P/IGF2R. Consequently these genes remain to be identified.
4.2
Future Work
Primarily, the first objective would be to complete the work carried out in this thesis, to
characterize the nature of the remaining variants and SSCP bandshift observed which was
not possible due to lack o f time. These include not only those RT-PCR products which
demonstrated apparently identical migration on SSCP gels to variants already characterized,
but also those which probably represent exon duplications. In addition, confirmation as to
the presence o f the point mutations at the genomic DNA level is also desirable to validate
that these are true somatic alterations.
Since mutants which have previously been unidentified were observed in this thesis, e.g.,
point mutations and out o f frame splice variants containing novel nucleotide sequences, it
would be interesting to determine their functional activity. This could be achieved by
cloning these variants into expression vectors and co-transfecting them with reporter
plasmids containing EREs into both ER positive, e.g., MCF-7 cells, and ER negative, e.g.,
HeLa or COS-1, cells since altered ER species with alterations to similar regions have been
shown to exhibit altered activities. These include point mutations which result in loss of
oestradiol binding and nuclear localization (Graham et al, 1990; Leslie et al., 1992),
hyperactive receptor (Fuqua et al., 1996) and both resistance to (Kami et al, 1994) and
stimulation by (Zhang et al., 1996) Tamoxifen. This type of analysis could be extended to
cotransfection with wt expression constructs to determine if the out of frame exon 7 deletion
variant acts as a dominant negative receptor, since a similar alteration has been observed in
other studies (Wang and Miksciek, 1990; Fuqua et al, 1992).
The frequent detection o f a range o f ER splice variants in breast cancer, seemingly
culminating in the predominance (in excess of wt receptor) of the exon 5 deletion variant,
seems to suggest alteration(s) to the mechanism of alternative splicing. Furthermore, the
recent detection of an alternatively spliced androgen receptor expressed in excess of wt in a
small proportion o f breast tumours (Zhu et al., 1997), may suggest aberrations to the
splicing machinery itself. It this was true, then it is not unreasonable to assume that this
would result in the aberrant splicing of many genes. Therefore, screening tumours for
splicing of multiple genes, e.g., the androgen receptor and other members of the steroid
receptor super family, could identify tumours which contain alterations to the splicing
machinery and are therefore good candidates for further investigation.
An alternative explanation is that mutations within introns could reveal or destroy consensus
splice signals which could direct an apparently normal splicing machinery to process spliced
species resulting in their abnormal level of expression. Therefore it would be interesting to
compare the relevant intron sequences (approximately 200-300 nucleotides) at the exon
borders ot the ER gene in tumours which express these spliced variants with the
corresponding sequences of wt gene. Consequently if this hypothesis was valid, it could be
possible to account, at least in part, for the aberrant expression of these variants and identify
potential mutation hotspots.
This study demonstrated that a high proportion of carcinomas in which allelic loss at the
ESR locus was detected were positive for ER with a majority of these cases negative for
progesterone receptor expression (9/13). It is attractive to postulate that these findings
represent loss o f one ER allele with mutation or aberrant splicing of the remaining allele,
resulting in the expression o f negative or dominant negative receptor which could account
for the steroid receptor phenotype of these cases. Unfortunately such a correlation could not
be made in this study as there was incomplete overlap between the populations screened for
allelic loss and that screened for altered/mutated ER. Therefore it would be interesting to
carry out both types o f analyses on the same tumours. This could be achieved by the
sequential isolation o f RNA and DNA from frozen breast carcinomas, made possible using
commercially available reagents, e.g., TRIzol, allowing allelic loss to be determined at the
ESR locus with ER variant/mutant analysis as described in this thesis. In addition to loss of
the oestrogen receptor gene at the chromosomal level, an alternative mechanism which
results in loss o f ER expression is methylation of promoter CpG islands. Consequently it
would also be interesting to develop the PCR based methylation assay and determine the
methylation state o f ER CpG islands in these same tumours, and correlate the findings to
ESR allele loss and the expression o f spliced ERs.
The presence o f alternatively spliced ERs in normal cells, breast cancer tissue and cell lines
has led some to propose that the regulation of ER transcripts by alternative splicing may be
involved in the expression o f the ER gene independently of cellular transformation
(Gotteland et al., 1995; Pfeffer et a l, 1996). Therefore, if differential splicing was regulated,
then splicing could control the level o f functional ER in the cell, and given the biological
activity o f these receptors, the overall tissue responsiveness to oestrogens, and the
expression o f oestrogen responsive genes. One of the major problems with this hypothesis is
the lack o f evidence demonstrating the coexpression of multiple variants, including wt,
within individual cells. One approach which could address this problem is the development
of an in situ RT-PCR method to detect individual species. Primers could be designed to
214
amplify specific splice variants which if labelled with different flouresence labels could
allow the simultaneous amplification and differential detection of multiple receptor species,
including wt receptor. These images could be superimposed in order to assess coexpression
of variants and to correlate their expression with histology, e.g., does the expression of a
particular variant correlate with a histological subtype.
This analysis could also be extended to assess the expression of these variants in normal
breast tissue. Their expression could be correlated with a number of different variables
including age, phase o f the menstrual cycle and menopausal status. In this way it could be
possible to identify the possible roles o f these variants in normal breast physiology.
The M6P/IGF2R gene has been shown to be a tumour suppressor gene in both
hepatocellular and breast cancers. This thesis has demonstrated that LOH at the M6P/IGF2R
locus is associated with high nuclear grade DCIS. As a tumour suppressor gene, mutations
would be expected to be present in the remaining allele of these lesions exhibiting LOH.
Therefore, it would be interesting to screen the remaining allele for mutations. The
M6P/IGF2R gene is a very large gene (>100 kb, 9.1 kb message, Hankins et al., 1996) and
consequently a major task to sequence. This is exacerbated by the fact that only short DNA
fragments can be isolated from paraffin embedded fixed tissue only allowing short regions
to be sequenced at any one time. Consequently it would be easier to isolate RNA from
frozen tissue, although microdissecting individual lesions of DCIS from frozen tissue is not
easily achieved.
The region in which the ER gene is mapped (6q25.1) has been identified in this study as a
major site o f genetic alteration at this stage of the disease indicting the possible involvement
of as o f yet unidentified tumour suppressor genes. It would be useful therefore to carry out a
more detailed analysis o f this region with a higher density of microsatellites to further define
regions o f allelic loss. In addition, it would be interesting to extend this analysis to nonneoplatic tissue, e.g., hyperplasia and its apparently normal surrounding tissue, in order to
determine the earliest stage at which these alterations occur.
215
APPENDIX I
Asymptomatic Breast tumour Clinicopathological Raw Data - ER Variant Analysis
T um our
No.
3
4
5
6
8
9
10
11
12
13
14
15
17
18
19
20
21
22
24
25
26
27
28
29
30
31
33
34
35
36
37
38
39
40
41
42
43
44
51
52
53
54
55
56
Type
Grade
Size (mm)
ER
PgR
MIB-1
ID
ID
Tub
ID
ID
ID
Tub
Tub
ID
Tub
ID
ID
Tub
ID
ID
ID
Tub
ID
Lob/Tub
ID
ID
ID
ID
ID
ID
ID
ID
ID
ID
ID
ID
ID\IL
ID
ID
ID
ID
Tub
ID
ID
ID
ID
ID
ID
ID
II
II
I
I
II
II
I
I
III
I
II
III
I
I
II
II
I
II
I
II
III
I
II
II
I
I
III
II
III
III
I
II
II
I
I
II
I
I
II
I
II
II
II
III
15
15
14
14
15
10
15
12
15
10
15
12
13
8
15
12
15
15
10
13
11
14
13
15
17
15
15
15
15
15
13
14
11
12
12
15
12
15
10
13
14
14
12
11
117
247
111
165
188
160
154
197
175
178
191
124
193
80
196
80
151
123
68
130
176
231
261
232
80
210
0
0
0
0
225
112
151
194
173
183
171
132
232
136
76
130
60
0
98
31
0
98
0
0
153
202
107
178
0
0
64
152
0
72
123
0
0
62
0
0
0
0
160
180
0
0
0
0
190
135
92
32
192
234
12
169
62
10
22
15
145
0
3.2
19.8
5.5
7.2
12.5
3
9.7
2.7
7.6
6.7
27.3
35.7
6.8
4.4
7.5
11
4.8
2.3
3.1
6.6
17.6
5.1
12.6
12.2
8
5.5
60
30
80
1
8
5
15.4
3
15.4
3.3
10
12.2
1.4
1
12
5
8
30
APPENDIX II
Human 5’Flanking Region O f The Oestrogen Receptor Gene (Keaveney et al., 1992)
1
CAAT Box
ggatccatgt gaacgccact gggaaatgag agacctcgtt cccaatcacg gtcagtgcaa
61
ctcgaaagcc taaaatcagt ttaaaacaaa ggtatctacc tttatcttat gttcatatcc
TATA Box
121 taggctttta ataatacgta tttttcacat gtttacagaa agcagtcaac tgagctattc
181 atggaaaggt ttgtgggttt ggttaacgaa gtgaggagta ttacatttca gctggaaaca
241 catccctaga atgccaaaac atttattcca aagtctggtt tcctggtgca atcggaggca
301 tggcaatgcc tctgttcaga gactgggggc tagggccagt aaggcatttg atccacatgt
361 atcccagaag gcttttattg ttaaattata ttctttcgga aaaaccaccc atgtcctatt
421 ttgtaaactt gatatccata cacttttgac tggcattcta ttttagccgt aagactatga
481 ttcacagcaa gcctgttttt cctcttgctt ggggtggcag cagaaagcat agggtacttt
541 ccagcctcca agggtagggg caaaggggct ggggtttctc ctccccagta cagctttctc
INR
601 tggctgtgcc acactgctcc ctgtgagcag acagcaagtc tcccctcact ccccactgcc
661 attcatccag cgctgtgcag tagcccagct gcgtgtctgc cgggaggggc tgccaagtgc
721 cctgcctact ggctgcttcc cgaatccctg ccattccacg cacaaacaca tccacacact
781 ctctctgcct agttcacaca ctgagccatc gcacatgcga gcacattcct tccttccttc
INR
841 tcactctctc ggcccttgac ttctacaagc ccatggaaca tttctggaaa gacgttcttg
901 atccagcagg gtaggcttgt tttgatttct ctctctgtag ctttagcatt ttgagaaagc
961
aacttacctt tctggctagt gtctgtatcc tagcagggag atgaggattg ctgttctcca
1021 tgggggtatg tgtgtgtctc ctttttcttt caggacttgt aggattcttt gtgccatttg
1081 catataattt ggcaggttca cattttttaa gagccctatg aagtgctttt tgcatgtgtt
1141 ttaaaaaggc atttgaaaat tgaaagtgtg atttatggaa attaaatcat ctgtaaaaaa
1201 ttgctttgga aagtaatgat tgctggccat aaagggaaat atctgcgatg cacctaatgt
1261 gtttttaacc ctttatttgc tgacaatcta tagtcattaa tgctaaactc gattttggct
1321 tcagctacat ttgcatattg tccaacaatg gtctattttt gtaagaatta gataaaatgt
1381 atacttgata taaaatagtc aaaaatgtaa ctcttagtaa cagtaagctt ggcatttaga
1441 tagaccatga accacttcgt cagatactct gttgggtgtt tgggatagca attaaaacaa
1501 agtattgata gttgtatcag agtctattag gctgcagcaa aggaagttta ttcaaaagta
g acgcatgata tacttcacc (ESR !)->
1561 taaactatcc aagattatag acgcatgata tacttcacct attttttgtc tccttaatat
1621 gtatatatat atatatatat atatatatat acacatatat gtgtgtgtgt atgtgcgtgt
217
1681 gcatgtttaa cttttaattc agttaaaaac ttttttctat ttgtttttca tctggatatt
4“ (ESR 2) gt agacctataa
1741 tgattctgca tatcctagcc caagtgaacc gagaagatcg agttgtagga ctaaaggata
actaagacg
1801 gacatgcaga aatgcatttt aaaaatctgt tagctggacc agaccgacaa tgtaacataa
Half Palindromic ERE
Half Palindromic ERE
1861 ttgccaaagc tttggttcgt cracctqaggt tatgtttggt ataaaaaqrcrt cacattttat
1921 attcagtttt ctgaagtttt ggttgcataa
Hhal
1981 crccrccctaac caaacrcrtttt tctcraatcat
SV40 Enhancer Core
2041 gtacagtact crtcrortccaac ataaacacac
4- {ER Meth
gggagcattt
2101 tggaagggtc tatctacttt
ccaacctgtg gaaggcatga acacccatgt
ccttcacatg agaattccta atgggaccaa
aagtcaggct gagagaatct cagaaggttg
2 )agtccga ctctcttaga g
tgcagaggaa gaaactgagg tcctggcagg
Spl
atct
c
cqccc
2161 ttgcattctc ctgatggcaa aatgcagctc ttcctatatg tataccctga
2221 ccttcccctc agatgccccc tgtcagttcc cccagctgct aaatatagct gtctgtggct
2281 ggctgcgtat gcaaccgcac accccattct atctgcccta tctcggttac agtgtagtcc
2341 tccccagggt catcctatgt acacactacg tatttctagc caacgaggag ggggaatcaa
2401 acagaaagag agacaaacag agatatatcg gagtctggca cggggcacat aaggcagcac
Spl
2461 attagagaaa Qcccrcrcccct ggatccgtct ttcgcgttta ttttaagccc agtcttccct
INR
2521 gggccacctt tagcagatcc tcgtgcgccc ccgccccctg gccgtgaaac tcaacctcta
2581 tccagcagcg acgacaagta aagtaaagtt cagggaagct gctctttggg atgctcaaat
2641 c
Human Oestrogen Receptor Gene Coding Sequence (Green et al., 1986)
CAAT Box
gagttgtgcc tggagtgatg tttaagccaa tgtcagggca aggcaacagt ccctggccgt
TATA Box
61 cctccagcac ctttgtaatg catatgagct cgggagacca gtacttaaag ttggaggccc
1
121 gggagcccag gagctggcgg agggcgttcg tcctgggagc
tgcacttgct ccgtcgggtc
181 gccggcttca ccggaccgca ggctcccggg gcagggccgg
ggccagagct cgcgtgtcgg
241 cgggacatgc gctgcgtcgc ctctaacctc gggctgtgct
ctttttccag|gtggcccgcc
| Exon 1
3 01 ggtttctgag ccttctgccc tgcggggaca cggtctgcac cctgcccgcg gccacggacc
|->A Domain
3 6 1 1afccraccatga ccctccacac caaagcatct gggatggccc tactgcatca gatccaaggg
|~>B Domain
421 aacgagctgg agcccctgaa ccgtccgcag ctcaagatcc ccctggagcg gccc|ctgggc
4 81
gaggtgtacc tggacagcag caagcccgcc gtgtacaact
accccgaggg cgccgcctac
541
gagttcaacg ccgcggccgc cgccaacgcg caggtctacg
gtcagaccgg cctcccctac
6 01
ggccccgggt ctgaggctgc ggcgttcggc tccaacggcc
tggggggttt ccccccactc
661
aacagcgtgt ctccgagccc gctgatgcta ctgcacccgc
cgccgcagct gtcgcctttc
218
721
(739) ca ggtgccctac tacctgga ->
ctgcagcccc acggccagca ggtgccctac tacctggaga acgagcccag cggctacacg
781
(819) tt cagataatcg acgccagg
gtgcgcgagg ccggcccgcc ggcattctac ag|gcaaatt cagataatcg acgccagggt
841
| Exon 2
|
ggcagagaaa gattggccag taccaatgac aagggaagta tggctatgga atctgcca|ag
ct tagacggt tc
DNA Binding Domain
901 gagactcgct actgtgcagt gtgcaatgac tatgcttcag gctaccatta tggagtctgg
ctctgagc (908)
|
Exon 3
961 tcctgtgagg gctgcaaggc cttcttcaag agaagtattc aag|gacataa cgactatatg
|
Exon 4
1021 tgtccagcca ccaaccagtg caccattgat aaaaacagga| ggaagagctg ccaggcctgc
1081 cggctccgca aatgctacga agtgggaatg atgaaaggtg ggatacgaaa agaccgaaga
(1160) a caagcgccag agagatgat ->
1141 ggagggagaa tgttgaaaca caagcgccag agagatgatg gggagggcag gggtgaagtg
tt acaactttgt gttcgcgg (1168)
12 01 gggtctgctg gagacatgag agctgccaac ctttggccaa gcccgctcat gatcaaacgc
(1210) tg gagacatgag agctgcca
1261 tctaag|aaga acagcctggc cttgtccctg acggccgacc agatggtcag tgccttgttg
| Hormone Binding Domain
1321 gatgctgagc cccccatact ctattccgag tatgatccta ccagaccctt cagtgaagct
tgggaa gtcactacga
1381 tcgatgatgg gcttactgac caacctggca gacagggagc tggttcacat gatcaactgg
agct (1384)
|
Exon 5
1441 gcgaagaggg tgccag|gctt tgtggatttg accctccatg atcaggtcca ccttctagaa
1501 tgtgcctggc tagagatcct gatgattggt ctcgtctggc gctccatgga gcacccagtg
|
Exon 6
1561 aagctactgt ttgctcctaa cttgctcttg gacag|gaacc agggaaaatg tgtagagggc
1621
(1628) ga tcttcgacat gctgctgg ->
atggtggaga tcttcgacat gctgctggct acatcatctc ggttccgcat
ct agaagctgta cgacgacc (1648)
gatgaatctg
1681
|
Exon 7
cagggagagg agtttgtgtg cctcaaatct attattttgc ttaattctg|g agtgtacaca
1741
tttctgtcca gcaccctgaa gtctctggaa gagaaggacc atatccaccg
agtcctggac
1801
aagatcacag acactttgat ccacctgatg gccaaggcag gcctgaccct
gcagcagcag
219
|
Exon 8
1861 caccagcggc tggcccagct cctcctcatc ctctcccaca tcaggcacat gag|taacaaa
1921 ggcatggagc atctgtacag catgaagtgc aagaacgtgg tgcccctcta tgacctgctg
1981 ctggagatgc tggacgccca ccgcctacat gcgcccacta gccgtggagg ggcatccgtg
2041 gaggagacgg accaaagcca cttggccact gcgggctcta cttcatcgca ttccttgcaa
< cctctgcc tggtttcggt gaa (2063)
cccgtgat gaagtagcga aa (2092)
2101 aagtattaca tcacggggga ggcagagggt ttccctgcca caqtctgracraq ctccctggc
2161 tcccacacgg ttcagataat ccctgctgca ttttaccctc atcatgcacc actttagcca
2221 aattctgtct cctgcataca ctccggcatg catccaacac caatggcttt ctagatgagt
2281 ggccattcat ttgcttgctc agttcttagt ggcacatctt ctgtcttctg ttgggaacag
2341 ccaaagggat tccaaggcta aatctttgta acagctctct ttcccccttg ctatgttact
2401 aagcgtgagg attcccgtag ctcttcacag ctgaactcag tctatgggtt ggggctcaga
2461 taactctgtg catttaagct acttgtagag acccaggcct ggagagtaga cattttgcct
2521 ctgataagca ctttttaaat ggctctaaga ataagccaca gcaaagaatt taaagtggct
2641 tcctatggca atgcatcctt ttatgaaagt ggtacacctt aaagctttta tatgactgta
2701 gcagagtatc tggtgattgt caattcactt ccccctatag gaatacaagg ggccacacag
2761 ggaaggcaga tcccctagtt ggccaagact tattttaact tgatacactg cagattcaga
2821 gtgtcctgaa gctctgcctc tggctttccg gtcatgggtt ccagttaatt catgcctccc
2881 atggacctat ggagagcaac aagttgatct tagttaagtc tccctatatg agggataagt
2941 tcctgatttt tgtttttatt tttgtgttac aaaagaaagc cctccctccc tgaacttgca
3001 gtaaggtcag cttcaggacc tgttccagtg ggcactgtac ttggatcttc ccggcgtgtg
3061 tgtgccttac acaggggtga actgttcact gtggtgatgc atgatgaggg taaatggtag
3121 ttgaaaggag caggggccct ggtgttgcat ttagccctgg ggcatggagc tgaacagtac
3181 ttgtgcagga ttgttgtggc tactagagaa caagagggaa agtagggcag aaactggata
3241 cagttctgag cacagccaga cttgctcagg tggccctgca caggctgcag ctacctagga
3301 acattccttg cagaccccgc attgcctttg ggggtgccct gggatccctg gggtagtcca
3361 gctcttattc atttcccagc gtggccctgg ttggaagaag cagctgtcaa gttgtagaca
3421 gctgtgttcc tacaattggc ccagcaccct ggggcacggg agaagggtgg ggaccgttgc
3481 tgtcactact caggctgact ggggcctggt cagattacgt atgcccttgg tggtttagag
3541 ataatccaaa atcagggttt ggtttgggga agaaaatcct cccccttcct cccccgcccc
3601 gttccctacc gcctccactc ctgccagctc atttccttca atttcctttg acctataggc
3661 taaaaaagaa aggctcattc cagccacagg gcagccttcc ctgggccttt gcttctctag
220
3721 cacaattatg ggttacttcc tttttcttaa caaaaaagaa tgtttgattt cctctgggtg
3781 accttattgt ctgtaattga aaccctattg agaggtgatg tctgtgttag ccaatgaccc
3841 aggtagctgc tcgggcttct cttggtatgt cttgtttgga aaagtggatt tcattcattt
3901 ctgattgtcc agttaagtga tcaccaaagg actgagaatc tgggagggca aaaaaaaaaa
3961 aaaaagtttt tatgtgcact taaatttggg gacaatttta tgtatctgtg ttaaggatat
4021 gcttaagaac ataattcttt tgttgctgtt tgtttaagaa gcaccttagt ttgtttaaga
4081 agcaccttat atagtataat atatattttt ttgaaattac attgcttgtt tatcagacaa
4141 ttgaatgtag taattctgtt ctggatttaa tttgactggg ttaacatgca aaaaccaagg
4201 aaaaatattt agtttttttt tttttttttg tatacttttc aagctacctt gtcatgtata
4261 cagtcattta tgcctaaagc ctggtgatta ttcatttaaa tgaagatcac atttcatatc
4321 aacttttgta tccacagtag acaaaatagc actaatccag atgcctattg ttggatattg
4381 aatgacagac aatcttatgt agcaaagatt atgcctgaaa aggaaaatta ttcagggcag
4441 ctaattttgc ttttaccaaa atatcagtag taatattttt ggacagtagc taatgggtca
4501 gtgggttctt tttaatgttt atacttagat tttcttttaa aaaaattaaa ataaaacaaa
4561 aaaaatttct aggactagac gatgtaatac cagctaaagc caaacaatta tacagtggaa
4621 ggttttacat tattcatcca atgtgtttct attcatgtta agatactact acatttgaag
4681 tgggcagaga acatcagatg attgaaatgt tcgcccaggg gtctccagca actttggaaa
4741 tctctttgta tttttacttg aagtgccact aatggacagc agatattttc tggctgatgt
4801 tggtattggg tgtaggaaca tgatttaaaa aaaaaactct tgcctctgct ttcccccact
4861 ctgaggcaag ttaaaatgta aaagatgtga tttatctggg gggctcaggt atggtgggga
4921 agtggattca ggaatctggg gaatggcaaa tatattaaga agagtattga aagtatttgg
4981 aggaaaatgg ttaattctgg gtgtgcacca aggttcagta gagtccactt ctgccctgga
5041 gaccacaaat caactagctc catttacagc catttctaaa atggcagctt cagttctaga
5101 gaagaaagaa caacatcagc agtaaagtcc atggaatagc tagtggtctg tgtttctttt
5161 cgccattgcc tagcttgccg taatgattct ataatgccat catgcagcaa ttatgagagg
5221 ctaggtcatc caaagagaag accctatcaa tgtaggttgc aaaatctaac ccctaaggaa
5281 gtgcagtctt tgatttgatt tccctagtaa ccttgcagat atgtttaacc aagccatagc
5341 ccatgccttt tgagggctga acaaataagg gacttactga taatttactt ttgatcacat
5401 taaggtgttc tcaccttgaa atcttataca ctgaaatggc cattgattta ggccactggc
5461 ttagagtact ccttcccctg catgacactg attacaaata ctttcctatt catactttcc
221
5521 aattatgaga tggactgtgg gtactgggag tgatcactaa caccatagta atgtctaata
5581 ttcacaggca gatctgcttg gggaagctag ttatgtgaaa ggcaaataaa gtcatacagt
5641 agctcaaaag gcaaccataa ttctctttgg tgcaagtctt gggagcgtga tctagattac
5701 actgcaccat tcccaagtta atcccctgaa aacttactct caactggagc aaatgaactt
5761 tggtcccaaa tatccatctt ttcagtagcg ttaattatgc tctgtttcca actgcatttc
5821 ctttccaatt gaattaaagt gtggcctcgt ttttagtcat ttaaaattgt tttctaagta
5881 attgctgcct ctattatggc acttcaattt tgcactgtct tttgagattc aagaaaaatt
5941 tctattcatt tttttgcatc caattgtgcc tgaactttta aaatatgtaa atgctgccat
6001 gttccaaacc catcgtcagt gtgtgtgttt agagctgtgc accctagaaa caacatactt
6061 gtcccatgag caggtgcctg agacacagac ccctttgcat tcacagagag gtcattggtt
6121 atagagactt gaattaataa gtgacattat gccagtttct gttctctcac aggtgataaa
6181 caatgctttt tgtgcactac atactcttca gtgtagagct cttgttttat gggaaaaggc
6241 tcaaatgcca aattgtgttt gatggattaa tatgcccttt tgccgatgca tactattact
6301 gatgtgactc ggttttgtcg cagctttgct ttgtttaatg aaacacactt gtaaacctct
6361 tttgcacttt gaaaaagaat ccagcgggat gctcgagcac ctgtaaacaa ttttctcaac
6421 ctatttgatg ttcaaataaa gaattaaact
222
APPENDIX III
Asymptomatic Breast Tumour Clinicopathological Raw Data - LOH and MSI Analysis
Tumour
No.
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
62
68
70
74
76
78
80
88
90
92
94
96
98
100
102
Type
Grade
IDC
ILC
IDC/ILC
IDC
IDC
IDC
Tub
IDC
Tub
IDC
IDC
IDC
Tub
IDC
Lob/Tub
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
Tub
IDC
IDC
IDC
IDC
Tub
IDC
IDC
IDC
IDC
Tub
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
II
II
II
I
III
I
I
I
I
II
II
I
I
II
I
I
II
II
III
I
II
II
II
I
II
I
I
II
I
I
I
I
I
I
I
I
I
I
II
II
II
II
II
II
II
S ize
(mm)
9
15
15
10
13
10
12
9
8
10
11
13
10
10
15
10
10
15
10
12
15
10
8
10
15
15
12
15
12
11
11
15
14
10
8
5
10
11
13
15
15
15
15
15
15
ER
PgR
MIB-1
96
175
215
182
124
60
142
181
195
162
151
175
162
172
196
10
211
29
22
204
192
174
168
45
215
132
171
183
173
194
179
151
231
148
219
203
162
80
65
0
123
232
188
196
142
130
106
49
49
100
109
92
97
3
168
92
31
13
105
0
115
187
43
11
162
170
110
49
1
46
169
12
234
192
ND
133
123
0
157
0
85
83
152
30
0
0
0
0
0
0
6.6
20.8
15.6
NTC
6.3
0
0.5
6.3
1.5
2.5
15.4
12.3
8.2
10.4
1.9
8
13.2
6.9
29.7
18.2
6.8
0
1
15
23.7
12.2
10
3.3
15.4
ND
6.8
4.8
5.1
16.2
8.2
8.9
15.4
4.4
6.6
10
2.3
12.2
12.5
7.5
15.8
Node
Status
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
pos
neg
neg
neg
neg
neg
neg
neg
pos
neg
neg
neg
neg
Tum our
No.
104
106
108
110
112
114
116
118
120
122
124
126
132
134
136
Type
G rade
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
Tub
IDC
IDC
IDC
IDC
IDC
IDC
II
II
III
II
II
II
II
II
I
II
II
I
II
II
III
Sixe
(mm)
9
15
8
10
12
15
8
15
14
10
15
15
15
15
11
ER
PgR
MIB-1
221
195
0
216
182
237
245
214
75
160
196
103
0
123
176
0
34
0
104
50
77
60
212
15
0
0
0
0
0
0
11.3
7
25.8
7.4
59.9
11.9
24.7
28.3
16
3
7.5
5.6
10
2.3
17.6
Node
Status
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
neg
pos
neg
PUBLICATIONS AND ABSTRACTS
Publications
Loss of Heterozygosity At Chromosome 6q In Early Invasive And Preinvasive Breast
Carcinomas
SA Chappell, T Walsh, JA Shaw, RA Walker
British Journal o f Cancer 75: 1324-1329 (1997)
Loss of heterozygosity at the mannose 6-phosphate/insulin-like growth factor 2 receptor
gene correlates with poor differentiation in early breast carcinomas.
SA Chappell, T Walsh, JA Shaw, RA Walker
British Journal o f Cancer - In Press
Microsatellite Instability In Early Invasive Sporadic Breast Cancer
JA Shaw, T Walsh, SA Chappell, N Carey, K Johnson, RA Walker
British Journal o f Cancer 73: 1393-1397 (1996)
Microsatellite Instability In Ductal Carcinoma In Situ
T Walsh, SA Chappell, RA Walker, JA Shaw
Journal O f Pathology - In Press
Molecular Pathology O f Breast Cancer And Its Application To Clinical Management
RA Walker, JL Jones, S Chappell, T Walsh, JA Shaw
Cancer and Metastasis Reviews 16: 5-27 (1997)
Abstracts
Loss O f Heterozygosity At Chromosome 6q In Early Invasive Breast Cancer
SA Chappell, T Walsh, JA Shaw, RA Walker
Journal o f Pathology 178, page 6a (1996)
Loss Of Heterozygosity At Chromosome 6q In Non-Invasive Breast Cancer
SA Chappell, T Walsh, JA Shaw, RA Walker
British Journal o f Cancer 74, suppl. XXVI page 47 (1996)
Microdissection-Microsatellite Analysis Of Early Invasive Breast Cancers - Lack Of
Instability In Tubular Carcinomas
T Walsh, SA Chappell, RA Walker, JA Shaw
Journal o f Pathology 178, page 6a (1996)
Identification O f A Mutator Phenotype In Non-Invasive Breast Cancer
T Walsh, SA Chappell, RA Walker, JA Shaw
British Journal o f Cancer 74, suppl. XXVI page 17 (1996)
Microsatellite Instability In Ductal Carcinoma In Situ
T Walsh, SA Chappell, RA Walker, JA Shaw
Proceeding o f the American Association For Cancer Research
Analysis O f The E-Cadherin Gene In Infiltrating Ductal Carcinoma And Ductal
Carcinoma In Situ O f The Breast
T Walsh, SA Chappell, RA Walker, JA Shaw
American Journal O f Human Genetics
225
REFERENCES
Aaltonen, L.A., Peltomaki, P., Leach, F.S. Sistonen, P., Pylkkanen, L., Mecklin, J.P.,
Jarvininen, H., Powell, S.M., Jen, J., Hamilton, S.R., Petersen, L., Kinzler, K.W.,
Vogelstein, B., De La Chapelle. Clues to the pathogenesis of familial colorectal cancer.
Science, 260: 812-816 (1993)
Adnane, J., Gaudray, P., Simon, M.P., Simony-Lafontaine, J., Jeanteur, P., Theillet, C.
Proto-oncogene amplification and human breast cancer phenotype. Oncogene, 4: 1389-1395
(1989)
Aldaz, C. M., Chen, T., Sahin, A., Cunningham, J., and Bondy, M. Comparative allelotype
of in situ and invasive human breast cancer: high frequency of microsatellite instability in
lobular breast carcinomas. Cancer Res., 55: 3976-3081 (1995)
Ali, I.U., Lidereau, R., Theillet, C., Callahan, R. Reduction to homozygosity of genes on
chromosome 11 in human breast neoplasia. Science, 238: 185-188 (1987)
Ali, I.U., Campbell, G., Lidereau, R., Callahan, R. Lack of evidence for the prognostic
significance o f c-erb-B2 amplification in human breast carcinoma. Oncogene Res., 3: 139146(1988)
Ali, I.U., Merlo, G., Callahan, R., Lidereau, R. The amplification unit on chromosome
11 q l 3 in aggressive primary human breast tumours entails the bcl-1, int-2 and hst loci.
Oncogene, 4: 89-92 (1989a)
Ali, I.U., Lidereau, R., Callahan, R. Presence of two members of c-erbA receptor family (cerbA p and c-erbA2) in smallest region of somatic homozygosity in chromosome 3p21-p25
in human breast carcinoma. J. Natl. Cancer Inst., 81: 1815-1820 (1989b)
Ali, S, Metzger, D., Bomert, J-M, Chambon, P. Modulation of transcriptional activation by
ligand dependant phosphorylation o f the human oestrogen receptor A/B region. EMBO J.,
12: 1153-1160(1993)
Allegra, J.C., Lippman, M.E., Thompson, E.B. Estrogen receptor status. An important
variable in predicting response to endocrine therapy and cytotoxic chemotherapy in
advanced breast cancer. Cancer 46: 2961 (1980)
Allen, G.F., Leng, X., Tsai, S.Y., Weigel, N.L., Edwards, D.P., Tsai, M-J., O’Mally, B.W.
Hormone and antihormone induce distinct conformational changes which are central to
steroid receptor activation. J. Biol. Chem., 267: 19513 (1992)
Ambrosini, C., Cicatiello, L., cobellis, G., Addeo, R., Silc, V., Bressciani, F., Weiz, A.
Functional antagonism between the estrogen receptor and Fos in the regulation of c-fos
protooncogene transcription. Mol. Endocrinol, 7: 1472-1483 (1993)
Anderson, D.E. Genetic study o f breast cancer: Identification of a high risk group. Cancer,
34: 1090(1974).
226
Anderson, J., Poulsen, H.S. Immunohistochemical oestrogen receptor determination in
paraffin embedded tissue: Prediction o f response to hormone treatment in advanced breast
cancer. Cancer, 43: 1901-1908 (1989)
Andersen, T.I., Holm, R., Nesland, J.M., Heimdel, K.R., Ottestad, L., Borresen, A.L.
Prognostic significance o f the TP53 alterations in breast carcinoma. Br J. Cancer, 68: 540548 (1993)
Anolik, J.H., Klinge, C.M., Hilf, R., Bambara, R.A. Co-operative binding of estrogen
receptor to DNA depends on spacing of binding sites, flanking sequence and ligand.
Biochem., 34: 2511-2520 (1995)
Apter, D., Vihko, R. Early menarche, a risk factor for breast cancer, indicates early onset of
ovulatory cycles. J. Clin. Endocrinol Metab., 57: 82-86 (1983)
Arnold, S.F., Oboum, J.D., Jaffe, H., Notides, A.C. Serine-167 is the major estradiolinduced phosphorylation site on the human estrogen receptor. Mol. Endocrinol, 8: 12081214(1994)
Arnold, S.F., Oboum, J.D., Jaffe, H., Notides, A.C. Phosphorylation of the human estrogen
receptor on trysoine 537 in vivo and by Src family tyrosine kinases in vitro. Mol.
Endocrinol., 9:24: 24-33 (1995a)
Arnold, S.F., Aorojekina, D.P., Notides, A.C. Phosphorylation of tyrosine 537 on the human
estrogen receptor is required for binding to an estrogen response element. J. Biol. Chem.,
270: 30205-30212 (1995b)
Aronica, S.M., Katzenellenbogen, B.S. Progesterone regulation in uterine cells: stimulation
by estrogen, cyclic adenosine 3’,5’-monophosphate and insulin-like growth factor-I and
suppression by antiestrogens and protein kinase inhibitors. Endocrinology, 128: 2045-2052
(1991)
Aronica, S.M., Katzenellenbogen, B.S. Stimulation of estrogen receptor-mediated
transcription and alteration in the phosphorylation state of the rat uterine estrogen receptor
by estrogen, cyclic adenosine monophosphate and insulin-like growth factor-I. Mol.
Endocrinol., 7: 743-752 (1993)
Aurrichio, F. J., Migliaccio, A., Castoria, G., Lastoria, S., Schiavone, E. ATP-dependent
enzyme activating hormone binding o f the estradiol receptor. Biochem. Biophys. Res.
Commun., 131: 378-385 (1981a)
Aurrichio, F. J., Migliaccio, A., Rotondi, A., Inactivation of oestrogen receptor in vitro by
nuclear dephosphorylation. Biochem J., 194: 569-574 (1981b)
Aurrichio, F. J., Migliaccio, A., Di Domenico, M., Nola, E. Oestradiol stimulates tryosine
phosphorylation and hormone binding activity of its own receptor in a cell-free system.
EMBOJ., 6: 2923-2929 (1987)
Aurrichio, F. Phosphorylation o f steroid receptors. Steroid Biochem., 32: 613-622 (1989)
227
Bagachi, M.K., Tsai, S.Y., Tsai, M.-J., O’Malley, B.W. Ligand and DNA dependent
phosphorylation o f human progesterone receptor in vitro. Proc. Natl Acad. Sci. USA, 89:
2664-2668 (1992)
Baker, B.S., Tata, J.R. Accumulation of proto-oncogene c-erb A related transcripts during
Xenopus development: association with early acquisition of response to thyroid hormone
and estrogen. EM BOJ., 9: 879-885 (1990)
Barnes, D.M., Dublin, E.H., Fisher, C J., Levison, D.A., Millis, R.R. Immunohistochemical
detection o f p53 protein in mammary carcinoma: an important new independent indicator of
prognosis? Human Pathol, 24: 469-476 (1993)
Barrett-Lee, P.J., Travers, M.T., McClelland, R.A., Luqmani, Y., Coombes, R.C.
Characterization o f estrogen receptor messenger RNA in human breast cancer. Cancer Res.,
47: 6653-6659(1987)
Bartkova, J., Barnes, D.M., Millis, R.R., Gullick, W.J. Immunohistochemical demonstration
o f c-erb-B2 protein in mammary ductal carcinoma in situ. Hum. Pathol., 21: 1164-1167
(1990)
Bartkova, J., Lukas, J., Muller, H., Lutzhoft, D., Strauss, M., Bartek, J. Cyclin D1 protein
expression and function in human breast cancer. Int. J. Cancer, 57: 353-361 (1994)
Baulieu, E.E. Steroid hormone antagonists at the receptor level - a role for the heat shock
protein mw 90,000 (hsp90). J. Cell. Biochem., 35: 161-174 (1987)
Beato, M. Gene regulation by steroid hormones. Cell, 56: 335-344 (1989)
Beatson, G.T. On the treatment o f inoperable cases of carcinoma of the mamma:
Suggestions for a new method o f treatment with illustrative cases. Lancet 3: 104 (1896)
Bello, M.J., Rey, J.A. Cytogenetic analysis of metastatic effusions from breast tumours.
Neoplasma, 35: 71-81 (1989)
Beral, V., Hermon, C., Reeves, G., Peto, R. Sudden fall in breast cancer death rates in
England and Wales (letter). Lancet, 345: 1642-1643 (1995)
Bergh, J., Norbert, T., Sjogren, S., Lindgren, A., Holmberg, L. Complete sequencing of the
p53 gene provides prognostic information in breast cancer patients, particularly in relation to
adjuvant systemic therapy and radiotherapy. Nature Med, 1: 1029-1034 (1995)
Bems, P.M.J.J., Krijn, J.G.M., Van Putten, W.L.J., Van Steveren, I.L., Portengen, H.,
Foekens, J.A. C-myc amplification is a better prognostic factor than HER2/new
amplification in primary breast cancer. Cancer Res., 52: 1107-113 (1992)
Berry, M., Metzger, D., Chambon, P. Role of the two activating domains of the oestrogen
receptor in the cell-type and promoter-context dependant antagonist activity of the anti­
oestrogen 4-hydroxytomoxifen. EMBOJ., 9: 2811- 2818 (1990)
228
Berx, G., Stakes, K., Vanhengel, J., Molemans, F., Bussemakers,
Vanbokhoven, A.,
Vanroy, F. Cloning and characterization of the human invasion suppressor gene E-cadherin
(CDH1). Genomics, 26: 281-289 (1995a)
Berx, G., Cleton-Jansen, A-M., Nollet, F., de Leeuw, J.F., van der Vijver, M.J., Comelisse,
C., van Roy, F. E-cadherin is a tumour/invasion suppressor gene mutated in human lobular
breast cancers. EM BOJ., 14: 6107-6115 (1995b)
Bicknell, G.R., Shaw, J.A., Pringle, J.H., Furness, P.N. Amplification o f specific mRNA
from single human renal glomerulus, with an approach to the seperation of epithelial cell
mRNA. J. Pathol., 180: 188-193 (1996)
Bieche, I.M., Champeme, F., Matifas,L., Callahan, R., Lidereau, R. Loss of heterozygosity
on chromosome 7q in aggressive breast cancer. Lancet, 339: 139-143 (1992)
Bieche, I., Champene, M.H., Lideveau, R. Loss and gain of distinct regions of chromosome
lq in primary breast cancer. Clin. Cancer Res., 1: 123-127 (1995)
Bird, A.P. CpG-rich islands and the function of DNA methylation. Nature, 321: 209-213
(1986)
Bishop, J.M. Molecular themes in oncogenesis review. Cell, 64: 235-248 (1991)
Blot, W.J., Devesa, S.S., Fraumeni, J.F. Declining breast cancer mortality among young
American women. J. Natl. Cancer Inst., 78: 451-454 (1987)
Bobrowm L.G., Happerfleld, L.C., Grewgory, W.M., Millis, R.R., Ductal carcinomas in
situ: Assessment o f necrosis and nuclear morphology and their association with biological
markers. J. Pathol., 171: 333-341 (1995)
Bocquel, M.-T., Kumar, V., Strieker, C., Chambon, P., Gronmeyer, H. The contribution of
the N- and C-terminal regions o f steroid receptors to activation of transcription is both
receptor- and cell specific. Nucleic Acids Res., 17: 2581-2595 (1989)
Bodmer, W., Bailey, C., Bodmer, J., Bussey, H., Ellis, A., Gorman, P., Luibello, F.,
Murday, V., Rider, S., and Scambler, P. Localisation of the gene for familial adenomatous
polyposis on chromosome 5. Nature, 328: 614-616, 1987
Bonilla, M.,Ramirez, M., Lopez-Cueto, J., Gariglio, P. In vivo amplification and
rearrangements o f c-myc oncogene in human breast tumours. J. Natl. Cancer Inst., 80: 665671 (1988)
Bonnier, P., Romain, S., Carpin, C., Lejeune, C., Tubiana, N., Martin, P,M., Piana, L. Age
as a prognostic factor in breast cancer: relationship to pathologic and biologic features. Int.
J. Cancer, 17: 138-144(1995)
Borg, A., Zhang, Q-X., Aim, P., Olsson, H., Sellberg, G. The retinoblastoma gene in breast
cancer: allele loss is not correlated with loss of gene protein expression. Cancer Res., 52:
2991-2994(1992)
229
Boring, C.C., Squires, T.S., Tong, W., Noxtra, A.V. Cancer Statistics 1993. Cancer, 43: 726(1993)
Borresen, A-L., Ottestad, L., Gaustad, A., Andersen, T.I., Heikkila, R., Jahnsen, T., Tveit,
K.M., Nesland, J.M. Amplification and protein over-expression of the wew/HER-2/c-er6B2
proto-oncogene in human breast carcinomas: relationship to loss of gene sequence on
chromosome 17, family history and prognosis. Br. J. Cancer, 62: 585-590 (1990)
Bouzubar, N., Walker, K.J., Griffiths, K., Ellis, I.O., Elston, C.W., Robertson, J.F.R.,
Blarney, R.W., Nicholson, R.I. Ki67 immunostaining in primary breast cancer: Pathological
and clinical associations. Br. J. Cancer. 59: 943-947 (1989)
Breathnach, R., Chambon, P. Organisation and expression of eukaryotic split genes-coding
for proteins. Annu. Rev. Biochem., 50: 349-383 (1981)
Brenner, A.S., Aldaz., M.C. Chromosome 9p allelic loss and pl6/CDKN2 in breast cancer
and evidence o f p i 6 inactivation in immortal breast epithelial cells. Cancer Research, 55:
2892-2895(1995)
Bronner, C.E., Baker, S.M., Morrison, P.T., Warren, G., Smith, L.G., Lescoe, M.K., Kane,
M., Earebino, C., Lipford, J., Lindblom, A. Mutations in the mismatch repair gene
homologue hM LHl is associated with HNPCC. Nature, 368: 258-261 (1994)
Brook, J.D., McCurrach, M.E., Harley, H.G., Buckler, A.J., Church, D., Aburatani, H.,
Hunter, K., Stanton, V.P., Thirion, J.P., Hudson, T., Sohn, R., Zemelman, B., Snell, R.G.,
Rundle, S.A., Crow, S., Davies, J., Shelboume, P., Buxton, J., Jones, C., Juvonen, V.,
Johnson, K., Harper, P.S., Shaw, D.J., Houseman, D.E. Molecular basis of myotonic
dystrophy: expansion o f a trinucleotide (CTG) repeat at the 3' end of a transcript encoding a
protein kinase family member. Cell, 68: 799-808 (1992)
Brugarolso, J., and Jacks, T. Double indemnity: p53, BRCA and cancer. Nat. Med., 3: 721 722(1997)
Bruun Rasmussen, B., Rose, C., Thorpe, S.M., Hou-Jensen, K., Daehnfeldt, J.L., Palshof, T.
Histopathological characteristics and oestrogen receptor content in primary breast
carcinoma. Virchows Arch. Pathol. Anat., 390: 347-351 (1981)
Buckley, M.F., Sweeney, K.J.E., Hamilton, J.A., Sini, R.L., Manning, D.L., Nicholson, R.I.,
de Fazio, A., Watts, C.K.W., Musgrove, E.A., Sutherland, R.L. Expression and
amplification o f cyclin genes in human breast cancer. Oncogene, 8: 2127-2133 (1993)
Bunone, G., Briand, P-A., Miksicek, R.J., Picard, D. Activation of the unliganded estrogen
receptor by EGF involves the MAP kinase pathway and direct phosphorylation. EMBO J.,
15:2174-2183 (1996)
Buratowski, S., Hahn, S., Guarente, L., Sharp, P.A. Five intermediate complexes in
transcription initiation by RNA polymerase II. Cell, 56: 549-561 (1989)
Buratowski, S., Sopta, M., Greenblatt, Sharp, P.A. RNA polymerase II-associated proteins
required for DNA conformation change in the transcription initiation complex. P.N.A.S.
USA, 88: 7509-7513 (1991)
230
Byar, D.P., Sears, M.E., McGuire, W.L. Relationship between estrogen receptor values and
clinical data in predicting the response to endocrine therapy for patients with advanced
breast cancer. Eur. J. Cancer, 15: 299 (1979)
Cailleau, R., Young, R., Olive, M., and Reeves, W. J. Breast tumour cell lines from pleural
effusions. J. Natl. Cancer Inst., 53: 661-674 (1974)
Califfi, M., Teague, M.W., Jensen, R.A., Vzencak-Jones, C.L., Dupont, W.D., Pari, F.F. p53
gene mutations and steroid receptor status in breast cancer. Cancer, 73: 1247-1256 (1994)
Caput, D., Beutler, B., Hartog, K., Thayer, R., Brown-Shimer, S., Cerami, A. Identification
o f a common nucleotide sequence in the 3’-untranslated region of mRNA molecules
specifying inflammatory mediators. Proc. Natl Acad.Sci. USA, 83: 1670-1674 (1986)
Carter, C.L., Allen, C., Henson, D.E. Relation of tumour size, lymph node status, and
survival in 24,740 breast cancer cases. Cancer, 63: 181-187 (1989)
Castles, C.G., Fuqua, S.A.W., Klotz, D.M., Hill, S.M. Expression of a constitutively active
estrogen receptor variant in the estrogen-negative BT-20 human breast cancer cell line.
Cancer Res., 53: 5934-5939 (1993)
Castles, C.G., Klotz, D.M., Fuqua, S.A.W., Hill, S.M. Coexpression of wild-type and
variant oestrogen receptor mRNAs in a panel of human breast cancer cell lines. Br. J.
Cancer, 71: 974-980 (1995)
Castoria, G., Migliaccio, A., Green, S., Didmenico, M., Chambon, P. Properties of a
purified estradiol-dependant calf uterus tyrosine kinase. Biochemistry, 32: 1740-1750 (1993)
Catelli, M.G., Binart, N., Jung, T.I., Renoir, J.M., Baulieu, E.E., Feramisco, J.R., Welch,
W.J. The common 90-kD protein component of non-transformed ‘8S’ steroid receptors is a
heat-shock protein. EM BOJ., 4: 3131-3135 (1985)
Catherino, W.H., Jordan, V.C. The biological action of cDNAs from mutated estrogen
receptors transfected into breast cancer cells. Cancer Letters, 90: 35-42 (1995)
Cattoretti, G., Pileri, S., Parravicini, C., Becker, M. H. G., Poggi, S., Bifulco, C., Keys, G.,
D'Amato, L., Sabattini, E., Feudale, E., Reynolds, F., Gerdes, J., Rilke, F. Antigen
unmasking on formalin-fixed paraffin-embedded tissue sections. J. Pathology, 171: 83-98
(1993)
Cavailles, V., Augereau, P., Garcia, M., Rochefort, H. Estrogens and growth factors induce
the mRNA o f the 52K-pro-cathepsin D secreted by breast cancer cells. Nucl, Acids Res., 16:
1903-1919(1988)
Cavailles, V., Garcia, M., Rochefort, H. Regulation of cathepsin D and pS2 gene expression
by growth factors in MCF-7 human breast cancer cells. Mol. Endocrinol., 3: 552-558 (1989)
Cavenee, W.K., Dryja, T.P., Philips, R.A., Benedict, W.F., Dogbout, R., Bailie, B.L.,
Murphee, A.L., Strong, L.C., and White, R.L. Expression of recessive alleles by
chromosomal mechanisms in retinoblastoma. Nature, 305: 779-784 (1983)
231
Ceder, R., Razin, A. DNA methylation and development. Biochem. Biophys. Acta, 1049: 1-8
(1990)
Chalbos, D., Philips, A., Rochefort, H. Genomic cross-talk between the estrogen receptor
and growth factor regulatory pathways in estrogen target tissues. Seminars in Cancer
Biology, 5: 361-368 (1994)
Chalepakis, G., Postma, J.P.M., Beato, M. A model for hormone receptor binding to the
mouse mammary tumour virus regulatory element based on hydroxyl radical footprinting.
Nucleic Acids Res., 16: 10237-10247(1988)
Chambraud, B., Berry, M., Redeuilih, G., Chambon, P., Baulieu, E.E. Several regions of
human estrogen receptor are involved in the formation of receptor-heat shock protein 90
complexes. J. Biol. Chem., 265: 20686-20691 (1990)
Chellapan, S.P., Hiebert, S., Mudryj, M., Horowitz, J.M., Nevins, J.R. The E2F transcription
factor is a cellular target for the RB protein. Cell, 65: 1053-1061 (1991)
Chelly, J., Concordet, J.P., Kaplan, J.C., Kahn, A. Illegitimate transcription: transcription of
a gene in any cell type. Proc. Natl. Acad. Sci. USA, 86: 2617-2621 (1989)
Chen, L.C., Dolbaum, C., Smith, H.S. Loss of heterozygosity on chromosome lq in human
breast cancer. Proc Natl. Acad. Sci. USA, 86: 7204-7202 (1989)
Chen, L.C, Kurisu, W., Ljung, B. M., Goldman, E.S., Morre II, D., and Smith, H. S.
Heterogeneity for allelic loss in human breast cancer. J. Natl. Cancer Inst., 84: 506-510
(1992)
Chen, X.B., Farmer, G., Zhu, H., Prywes, R., Prives, C. Co-operative DNA-binding of p53
with TFIID (TBP) - A possible mechanism for transcription activation. Genes Develop., 7:
1837-1849(1993)
Chen, T., Sahin, A., and Aldaz, M. C. Deletion map of chromosome 16q in ductal
carcinoma in situ o f the breast: refining a putative tumour suppressor gene region. Cancer
Res., 56: 5605-5609 (1996)
Chevallier, B., Heintzman, F., Mosseri, V., et a l Prognostic value of oestrogen and
progesterone receptor assays in operable breast cancer. Cancer, 62: 2517-2524 (1988)
Cho, H., Katzenellenbogen, B.S. Synergistic activation of estrogen receptor-mediated
transcription by estradiol and protein kinase activators. Mol. Endocrinol., 7: 441-452 (1993)
Cho, H., Aronica, S.M., Katzenellenbogen, B.S. Regulation of progesterone receptor gene
expression in MCF-7 human breast cancer cells: comparison of the effects of cyclic AMP,
estradiol, IGF-1, and serum factors. Endocrinology, 134: 658-664 (1994)
Choi, W., Howe. G.R., Miller, A.B. An epidemiological study of breast cancer. Am. J.
Epidemiol, 510: 107 (1978)
Chomczynski, P., Sacchi, N. Single-Step Method of RNA isolation by acid guanidinium
thiocyanate-phenol-chloroform extraction. Analytical Biochem., 162: 156-159 (1987)
232
Chomczynski, P. A reagent for the single-step simultaneous isolation of RNA, DNA and
proteins from cell and tissue samples. BioTechniques, 15(3): 532-537 (1993)
Chuqui, R. F., Zhuang, Z., Emmert-Buck, M. R., Liotta, L. A., and Merino, M. J. Analysis
o f loss of heterozygosity on chromosome 1 I q l 3 in atypical ductal hyperplasia and in situ
carcinoma o f the breast. Am. J. Pathol., 150: 297-303 (1997)
Church, G.M., Gilbert, W. Genomic sequencing. Proc. Natl. Acad. Sci., 81: 1991-1995
(1984)
Clark, G.M., McGuire, W.L.: Follow-up study of HER-2Ineu amplification in primary
breast cancer. Cancer Res., 51: 944-948 (1991)
Claus, E.B., Risch, N.J., Thompson, W.D. Age at onset as an indicator of familial risk of
breast cancer. Am. J. Epidemiol., 131: 961-972 (1990)
Claus, E. B., Risch, N, and Thompson, W. D. Genetic analysis of breast cancer in the cancer
and steroid hormone study. Am. J. Hum. Genet., 48: 232-241 (1991)
Cleton-Jansen, A-M., Moerland, E. W., Kuipers-Dijkshoom, N. J., Callen, D., Sutherland,
G. R., Hansen, B., Devilee, P., and Comelisse, C. J. At least two different regions are
involved in allelic imbalance on chromosome arm 16q in breast cancer. Genes Chrom.
Cancer, 9: 101-107(1994)
Colditz, G.A., Hankinson, S.E., Hunter, D.J., Willet, W.C., Manson, J.E., Stampfer, M.J.,
Hennekens, C., Rosner, B., Speizer, F.E. The use of oestrogens and progestins and the risk
o f breast cancer in post menopausal women. N. Eng. J. Med., 332: 1589-1593 (1995)
Coleman, M.P., Esteve, J., Damiecki, P. Trends in cancer incidence and mortality. Chapter
17, Lyon: IARC Scientific Publications No. 121, pp 411-432 (1993)
Coles, C., Condie, A., Chetty, U., Steel, C.M., Evans, H.J., Prosser, J. p53 mutations in
breast cancer. Cancer Res., 52: 5291-5298 (1992)
Contegiacomo, A., Palmirotta, R., DeMarchis, L., Pizzi, C., Mastranzo, P., Delrio, P.,
Petrella, G., Figliolini, M., Bianco, A. R., Frati, L., Cama, A., and Mariani-Costantini, R.
Microsatellite instability and pathological aspects of breast cancer. J. Int. Cancer, 64: 264268(1995)
Contessa, G., Mouriesse, H., Friedman, S., Genin, G., Sarrazin, D., Rousesse, J. The
importance o f histological grade in long term prognosis: a study of 1010 patients, uniformly
treated at the Institute Gustave-Roussy. J. Clin. Oncol., 5: 1378-1386 (1987)
Costello, P., McCann, A., Carney, D.N., Dervan, P.A. Prognostic significance of
micro vessel density in lymph node negative breast carcinoma. Hum. Pathol., 26: 1181-1184
(1995)
Coussens, L., Yang-Feng, T.L., Chen, Y-C.L.E., Gray, A., McGrath, J., Seeburg, P.H.,
Libermann, T.A., Schlessinger, J., Francke, U., Levinson, A., Ullrich, A. Tyrosine kinase
receptor with extensive homology to EGF receptor shares chromosomal location with neu
oncogene. Science, 230: 1132-1139 (1985)
233
Cruciger, Q.V., Pathak, S., Cailleau, R. Human breast carcinomas: marker chromosomes
involving lq in seven cases. Cytogenet. cell. Genet., 17: 231-235 (1976)
Cullen, K.J., Yee, D., Sly, W.S., Pardue, J., Hampton, B., Lippman, M.E., Rosen, N.
Insulin-like growth factor expression and function in human breast cancer. Cancer Res., 50,
48-53 (1990)
Cuput, D., Beutler, B., Hartog, K., Thayer, R., Brown-Shimer, S., Cerami, A. Identification
of a common nucleotide sequence in the 3’-untranslated region of mRNA molecules
specifying inflammatory mediators. Proc. Natl. Acad Sci., USA, 83: 1670-1674 (1986)
Curtis, S.W., Korach, K.S. Uterine estrogen-receptor interaction with estrogen-responsive
DNA-sequence in vitro - effects o f ligand-binding on receptor-DNA complexes. Mol.
Endocrinol., 4: 276-286 (1990)
Daffada, A.A.I., Johnston, S.R.D., Nicholls, J., Dowsett, M. Detection of wild type and exon
5-deleted splice variant oestrogen receptor (ER) mRNA in ER-positive and -negative breast
cancer cell lines by reverse transcription/polymerase chain reaction. J. Mol. Endocrinol., 13:
265-273 (1994)
Daffada, A.A.I., Dowsett, M.D. Tissue-dependent expression of a novel slice variant of the
human oestrogen receptor. J. Steroid Mol., Biochem. Mol. Biol., 55: 413-421 (1995)
Dahms, N.M., Lobel, P., Komfeld, S. Mannose 6-phosphate receptors and lysosomal
enzyme targeting. J. iol. Chem., 264: 12115-12118 (1989)
Dana, S.L., Hoener, P.A., Wheeler, D.L., Lawrence, C.L., McDonnell, D.P. Novel estrogen
response element identified by genetic selection in yeast are differentially responsive to
estrogens and antiestrogens in mammalian cells. Mol. Endocrinol., 8: 1193-1207 (1994)
Danielian, P.S., White, R., Lees, J.A., Parker, M.G. Identification of a conserved region
required for hormonal dependant transcriptional activation by steroid hormone receptors.
EMBOJ., 11: 1025-1033 (1992)
deConinck, E.C., McPherson, L.A., Weigel, R.J. Transcriptional regulation of estrogen
receptor in breast carcinomas. Mol. Cell. Biol., 15: 2191-2196 (1995)
De Leon, D.D., Bakker, B., Wilson, D.M., Hintz, R.L., Rosenfeld, R.G. Demonstration of
insulin-like growth factor (IGF-I and -II) receptors and binding protein in human breast
cancer cell lines. Biochem. Biophys. Res. Commun., 152, 398-405 (1988)
Del Senno, L., Aguiari, G.L., Piva, R. Dinucleotide repeat polymorphism in the human
estrogen receptor (ESR) gene. Hum. Mol. Genet., 1, 354 (1992)
De Marchis, L., Contegiacomo, A., D ’Amico, C., Palmrotta, R., Pizzi, C., Ottini, L.,
Mastranzo, P., Figliolini, M., Petrella, G., Amanti, C., Battista, P., Bianco, A. R., Frati, L.,
Cama, A., and Mariani-Costantini, R. Microsatellite instability is correlated with lymph
node-positive breast cancer. Clin. Cancer Res., 3: 241-248 (1997)
Deng, G., Lu, Y., Zlotnikow, G., Thor, A. D., and Smith, H. S. Loss of Heterozygosity in
normal tissues adjacent to breast carcinomas. Science, 274: 2057-22059 (1996)
234
Dennis, P.A., Rifkin, D.B. Cellular activation of the latent transforming growth factor p
requires binding to the cation-independant mannose 6-phosphate/insulin-like growth factor
type II receptor. P. Nat. Acad. Sci. USA, 88: 580-584 (1991)
Denton, R.R. Koszewski, N.J., Notides, A.C. Estrogen receptor phosphorylation. Hormonal
dependence and consequence of specific DNA binding. J. Biol. Chem., 267: 7263-7268
(1992)
Desai, A.J., Luqmani, Y.A., Walters, J.E., Coope, R.C., Dagg, B., Gomm, J.J., Pace, P.E.,
Rees, C.N., Thirunavukkarasu, V., Shousha, S., Groome, N.P., Coombs, R., Ali, S. Presence
o f exon 5-deleted oestrogen receptor in human breast cancer: functional analysis and clinical
significance. Br. J. Cancer, 75: 1173-1184 (1997)
De Souza, A., Hankins, G.R., Washington, M.K., Fine, R.L., Orton, T.C., Jirttle, R.L.
Frequent loss o f heterozygosity on 6q at the mannose 6-phosphate/insulin like growth factor
II receptor locus in human hepatocelluar tumours. Oncogene, 10: 1725-1729 (1995a)
De Souza, A.T., Hankins, G.R., Washington, M.K., Fine, R.L., Orton, T.C., Jirtle, R.L.
M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of
heterozygosity. Nature Genet., 11,447-449 (1995b)
Devilee, P., Van Den Broek, M., Kuipers-Dijkshoom, N., Kollouri, R., Khan, P.M.,
Pearson, P.L., Comelisse, C.J. At least four different chromosomal regions are involved in
loss o f heterozygosity in human breast carcinoma. Genomics, 5: 554-560 (1989)
Devilee, P., van Vilet, M., van Sloun, P., Kuipers Dijksoom, N., Hermans, J., Pearson, P.L.,
Comelisse, C.J. Allelotype o f human breast carcinoma: a second major site for loss of
heterozygosity is on chromosome 6q. Oncogene, 6: 1705-1711 (1991)
Devilee, P., Schuuring, E., van de Vijver, M. J., and Comelisse, C. J. Recent developments
in the molecular genetic understanding of breast cancer. Crit. Rev. Oncog., 5: 247-270
(1994)
Devilee, P., and Comelisse, C.J. Somatic genetic changes in human breast cancer. Biochem.
Biophys. Acta., 1198: 113-130 (1994)
Dillon, E. K., de Boer, W. B., Papadimitriou, J. M., and Turbett, G. R. Microsatellite
instability and loss of heterozygosity in mammary carcinoma and its probable precursors.
Br. J. Cancer, 76: 156-162 (1997)
Dixon, J.M. The ABC of breast diseases. British Medical Journal, London. ppl9, (1995)
Doerfler, W. DNA methylation and gene activity. Ann. Rev. Biochem., 52: 93-124 (1983)
Dookeran, K.A., Rye, P.D., Dearing, S.J., Walker, R.A. Expression of the pS2 peptide in
primary breast carcinomas: comparison of membrane and cytoplasmic staining patterns. J.
Pathol., 171: 123-129(1993)
Dotzlaw, H., Alkhalaf, M., Murphy, L.C. Characterization of estrogen receptor variant
mRNAs from human breast cancers. Mol. Endocrinol, 6: 773-785 (1992)
235
Dowsett, M., Daffada, A., Chan, C.M.W., Johnston, S.R.D. Oestrogen receptor mutants and
variants in breast cancer. Eur. J. Cancer, 33: 1177-1183 (1997)
Duffy, M.J., Brouillet, J-P., Reilly, D., McDermott, E., O’Higgins, N., Fenelly, J.,
Maudelonde, T., Rochefort, H. Cathepsin D concentration in breast cancer cytosols:
correlation with biochemical, histological and clinical findings. Clin. Chem., 37: 101-104
(1991)
Duggan, B.D., Felix, J.C., Muderspach, L.I., Tourgeman, D., Zheng, J., Shibata, D.
Microsatellite instability in sporadic endometrial carcinoma. J. Nat. Cancer Inst., 86: 12161221(1994)
Dunn, J.E. Breast cancer among American Japanese in the San Francisco Bay area. Natl.
Cancer Inst. Monogr., 47: 157-160(1977)
Dupont, W. D., Page, L.W. Risk factors for breast cancer in women with proliferative breast
disease. N. Engl. J. Med., 312, 146-151 (1985)
Dutrillaux, B., Gerbault-Seureau, M., and Zaffani, B. Characterization of chromosomal
abnormalities in human breast cancer. Cancer Genet. Cytogenet., 49,203-217 (1990)
Edwards, D.P., Adams, D.J., McGuire, W.L. Specific protein synthesis regulated by
estrogen in human breast cancer. J. Steroid Biochem., 15: 237-259 (1981)
Ehlen, T., Dubeau, L. Loss o f heterozygosity on chromosomal segments 3p, 6q, and l i p in
human ovarian carcinomas. Oncogene, 5: 219-223 (1990)
El-Deiry, W.S., Tokino, T., Velculescu, V.E., Levy, D.B., Parsons, R., Trent, J.M., Lin, D.,
Mercer, W.E., Kinzler, K.W., Volgestein, B. Waf-1 a potential mediator of p53 tumour
suppression. Cell, 75: 817-820 (1993)
Elledge, R.M., Fuqua, S.A.W., Clark, G.M., Pujol, P., Allred, D.C., McGuire, W.L.
Prognostic significance o f p53 alterations in node-negative breast cancer. Breast Cancer
Res. Treat., 26: 225-235 (1993)
Ellis, I.O., Galea, M., Broughton, N., Locker, A., Blarney, R.W., Elston, C.W. Pathological
prognostic factors in breast cancer, II: Histological type. Relationship with survival in a
large study with long term follow up. Histopathology, 20: 479-489 (1992)
Ellis, M.J.C., Singer, C., Hornby, A., Rasmussen, A., Cullen, K.J. Insulin-like growth factor
mediated stromal-epithelial interactions in human breast cancer. Breast Cancer Res. Treat.,
31,249-261 (1994)
Elston, C.W., Ellis, I.O. Pathological prognostic factors in breast cancer. The value of
histological grade in breast cancer: experience from a large study with long term follow up.
Histopathology, 19: 403-410(1991)
Emmet-Buck, M. R., Bonner, R. F., Smith, P. D., Chuaqui, R. F., Zhuang, Z., Goldstein, S.
R., Weiss, R. A., and Liotta, L. A. Laser capture microdissection. Science, 274: 998-1001
(1996)
236
Escot, C., Theillet, C., Lidereau, R., Spyratos, F., Champeme, M-H., Gest, J., Callahan, R.
Genetic alterations o f the c-myc proto-oncogene (MYC) in human breast carcinomas. Proc.
Natl. Acad. Sci. USA. 83: 4834-4838 (1986)
Eshelman, J. R., and Markowitz, S.D. Microsatellite instability in inherited and sporadic
neoplasms. Curr. Opin. Oncol., 7: 83-89 (1995)
Evans, R.M. The steroid and thyroid hormone superfamily. Science, 240: 889-895 (1988)
Evan, G.I., Wyllie, A.H., Gilbert, C.S., Littlewood, T.D., Land, H., Brooks, M., Waters,
C.M., Penn, L.Z., Hancock, D.C. Induction of apoptosis in fibroblasts by c-myc protein.
Cell, 69: 1-20(1992)
Eyfjord, J.E., Thorlacius, S., Steinarsdottir, M., Valgardottir, R., Ogmundsdottir, H.M.,
Anamthawat-Jonsson, K. p53 abnormalities and genomic instability in primary human
breast carcinomas. Cancer Res., 55: 646-651 (1995)
Fahmer, T.J., Carrol, S.L., Milbrandt, J. The NGF1-B protein, an inducible member of the
thyroid steroid-receptor family, is rapidly modified posttranslationally. J. Mol. Cell Biol.,
10: 6454-6459 (1990)
Fantl, V., Richards, M.A., Smith, R., Lammie, G.J.A., Johnstone, G., Allen, D., Gregory,
W., Peters, G., Dickson, C., Barnes, D.M. Gene amplification on chromosome band 1l q l 3
and oestrogen receptor status in breast cancer. Eur. J. Cancer, 26: 423-429 (1990)
Fawell., S.E., Lees, J.A., White, R., Parker, M.G. Characterisation and co-localization of
steroid binding and dimerisation activities in the mouse estrogen receptor. Cell, 60: 953-962
(1990)
Fazeli, A., Dickinson, S. L., Hermiston, M. L., Tighe, R. V., Steen, R. G., Small, C. G.,
Stoeckli, E. T., Keino-Masu, K., Masu, M., Rayburn, H., Simons, J., Bronson, R. T.,
Gordon, J. I., Tessier-Lavigne, M., and Weinberg, R. A. phenotype of mice lacking
functional deleted in colorectal cancer (Dec) gene. Nature, 386: 796-803 (1997)
Fearon, E. R., and Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell, 54:
3011-3020(1990)
Fearon, E.R. Genetic alterations underlying colorectal tumourogensis. Cancer Surv., 12:
119-136(1992)
Feinberg, A.P. and Vogelstein, B. A Technique for radiolabelling DNA restriction
endonuclease fragments to high specific activity. Analytical Biochem., 132: 6-13 (1983)
Fereti-Passantonopoulou, A.D., Panani, A.D. Common cytogenetic findings in primary
breast cancer. Cancer Genet. Cytogenet., 27: 289-298 (1987)
Ferguson, D.J.P., Anderson, T.J. Morphological evaluation of cell turnover in relation to
menstrual cycle in the “resting” human breast. Br. J. Cancer, 44: 177-181 (1981)
237
Ferguson, D.J.P, Meier, P., Karrison, T., Dawson, P.J., Straus, F.H., Lowenstein, F.E.
Staging o f breast cancer and survival rates. An assessment based on 50 years of experience
with radical mastectomy. J.A.M.A. 248: 1337-1341 (1982)
Ferguson, A.T., Lapidus, R.G., Baylin, S.B., Davidson, N.E. Demetylation of the estrogen
receptor gene in estrogen receptor-negative breast cancer cells can reactivate estrogen
receptor gene expression. Cancer Res., 55: 2279-2283 (1995)
Fidler, I.J., Hart, I.R. Biological diversity in metaplastic neoplasms - origins and
implications. Science, 217: 998-1001 (1982)
Fishel, R., Lescoe, M. K., Rao, M. R., Copeland, N. G., Jenkins, N. A., Garber, J., Kane, M.,
and Kolodner, R. The human mutator gene homologue MSH2 and its association with
hereditary nonpolyposis colon cancer. Cell, 75: 1027-1038 (1993)
Fishel, R., Ewel, A., Lescoe, M.K. Purified human MSH2 protein binds to DNA
mismatched nucleotides. Cancer Res., 54: 5539-5542 (1994)
Fisher, B., Slack, N.H.D., Bross, I.D.J. Cancer of the breast - size of neoplasm and
prognosis. Cancer, 24: 1071-1080(1969)
Fisher. B., Slack, N.H., Katrych, D. Ten year follow up results of patients with carcinoma of
the breast in a co-operative clinical trial evaluating surgical adjuvent chemotherapy. Surg.
Gynecol. Obstet., 140: 528-534 (1975)
Fisher, B., Redmon, D.C., Fischer, E.R., Caplan, R. Relative worth of oestrogen and
progesterone receptor and pathologic characteristic of differentiation as indicators of
prognosis in node negative breast cancer patients: findings from the national surgical
adjuvant breast and bowel project protocol B-06. J. Clin. Oncol., 6: 1076-1087 (1988)
Folcan, P.J., Glascock, R.F., Irvine, W.T. Studies with tritium labelled hexoestrol in
advanced breast cancer. Lancet 2: 796 (1961)
Friend, S.H., Bernards, R., Rogelj, S., Weinberg, R.A., Rapaport, I.M., Albert, D.M., Dryja,
T.P. A human DNA segment with properties of the gene that predispose to retinoblastoma
and osteosarcoma. Nature, 323: 643-646 (1986)
Fujii, H., Marsh, C., Cairns, P., Sidransky, D., and Gabrielson, E. Genetic divergence in the
clonal evolution of breast cacner. Cancer Res., 56: 1493-1497 (1996a)
Fujii, H., Szumel, R., Marsh, C., Zhou, W., and Gabrielson, E. Genetic progression,
histological grade, and allelic loss in ductal carcinoma in situ. Cancer Res., 56: 5260-5265
(1996b)
Fujii, H., Zhou, W., Gabrielson, E. Detection of frequent allelic loss of 6q23-25.2 in
microdissected human breast cancer tissues. Genes Chromosomes Cancer, 16: 35-39
(1996c)
Fuqua, S.A.W., Falette, N.F., McGuire, W.L. Sensitive detection of estrogen receptor RNA
by polymerase chain reaction assay. J. Nat. Cancer Inst., 82: 858-861 (1990)
238
Fuqua, S.A.W., Fitzgerald, S.D., Chamness, G.C. Variant human breast tumour estrogen
receptor with constitutive transcriptional activity. Cancer Res., 51: 105-109 (1991)
Fuqua, S.A.W., Fitzgerald, S.D., Allred, D.C., Elledge, R.M., Nawaz, Z., McDonnell.,D.P.,
O ’Malley, B.W., Greene, G.L., McGuire, W.L. Inhibition of estrogen receptor action by a
naturally occuring variant in human breast tumours. Cancer Res., 52: 483-486 (1992)
Fuqua, S.A.W., Allred, D.C., Elledge, R.M., Krieg, S.L., Bendix, M.G., Nawaz, Z.,
O ’Malley, B.W., Greene, G.L., McGuire, W.L. The ER-positive/PgR negative breast cancer
phenotype is not associated with mutations within the DNA binding domain. Br. Cancer.
Res. Treat., 26: 191-202 (1993)
Fuqua, S.A.W., Wolf, D.M. Molecular aspects of estrogen receptor variants in breast cancer.
Br. Cancer. Res. Treat., 35: 233-241 (1995)
Fuqua, S.A.W., Wiltschke, C., Lemieux, P., Wolf, D.M. A hyperactive estrogen receptor
variant in hyperplastic breast-ER82 breast lesions. Lab Investigation, 74: 79 (1996)
Garcia, I., Dietrich, P.Y., Aapro, M., Vauthier, G., Vadas, L., Engel, E. Genetic alterations
of c-myc, c-erb-Bl, and c-Ha-ras protooncogenes and clinical associations in human breast
carcinomas. Cancer Res., 49: 6675-6679 (1989)
Garcia, T., Sanchez, M., Cox, J.L., Shaw, P.A., Ross, J.B., Lehrer, S., Schachter, B.
Identification o f a variant form o f the human estrogen receptor with an amino acid
replacement. Nucleic Acids Res., 17: 8364 (1989)
Garcia, M., Derocq, P., Pujol, P., Rochefort, H. Overexpression of transfected cathepsin D
in transformed cells increased their malignant phenotype and metastatic potency. Oncogene,
5: 1809-1814(1990)
Garcia, M., Derocq, D., Freiss, G., Rochefort, H. Activation of estrogen receptor transfected
into a receptor negative breast cancer cell line decreases the metastatic and invasive
potential o f the cells. Proc. Natl. Acad. Sci. USA, 89: 11538-11542 (1992)
Gaub, M-P., Bellard, M., Scheurer, I., Chambon, P., Sassone-corsi, P. Activation of the
ovalbumin gene by the estrogen receptor involves the Fos-Jun complex. Cell, 63: 1267-1276
(1990)
Gaub, M. -P., Rochette-Egly, C., Lutz, Y., Ali, S., Matthes, H., Scheuer, I., Chambon, P.
Immunodetection o f multiple species of retanoic acid receptor-alpha-evidence for
phosphorylation. Exp. Cell Res., 201: 335-346 (1992)
Gebhart, E., Bruderlein, S., Augustus, M., Siebert, E., Feldner, J., Schmidt, W. Cytogenetic
studies on human breast carcinomas. Breast Cancer Res. Treat., 8: 125-138 (1986)
Genuardi, M., Tsihira, H., Anderson, D.E., Saunders, G.F. Distal deletion of chromosome
lp in ductal carcinoma o f the breast. Am. J. Hum. Genet., 45: 73-82 (1989)
Gillet, C., Fantl, V., Smith, R., Fisher, C., Bartek, J., Dickson, C., Barnes, D.M., Peters, G.
Amplification and overexpression of cyclin D1 in breast cancer detected by
immunocytochemistry staining. Cancer Res., 54: 1812-1817 (1994)
239
Gillet, C., Smith, P., Gregory, W., Richard, M., Millis, R., Peters, G., Barnes, D. Cyclin D1
and prognosis in human breast cancer. Int. J. Cancer, 69: 92-99 (1996)
Gilneur, C., Zenke, M., Beug, H., Ghysdeal, J. Phosphorylation of the v-erb-A protein is
required for its function as an oncogene. Genes Dev., 4: 1663-1676 (1990)
Giochonmantel, A., Loosfelt, H., Lescop, P., Saar, S., Atger, M., Perrot-Applanat, M.,
Milgrom, E. Mechanism o f nuclear localisation of the ‘progesterone’ receptor: evidence for
interation between monomers. Cell, 57: 1147-1154 (1989)
Glascock, R.F., Hoeskstron, V.G. Selective accumulation of tritium-labeled hexoestrol by
the reproductive organs o f immature female goats and sheep. J. Steroid Biochem., 72: 673
(1959)
Glass, C.K., Holloway, J.M., Devary, O.V., Rosenfeld, M.G. The thyroid hormone receptor
binds with opposite transcriptional effects to a common sequence in thyroid hormone and
oestrogen response elements. Cell, 54: 313-323 (1988)
Glavac, D., Dean, M. Potimization o f the single-strand conformation polymorphism (SSCP)
techchnique for detection o f point mutations. Human Mutation, 2: 404-4141 (1993)
Glebov, O. K., McKenzie, K. E., White, C. A., and Sukumar, S. Frequent p53 gene
mutations and novel alleles in familial breast cancer. Cancer Res., 54: 3703-3709 (1994)
Goelz, S.E., Vogelstein, B., Hamilton, S.R., Feinberg, A.P. Hypomethylation of DNA from
benign and malignant human colon neoplasms. Science, 228: 187-190 (1985)
Gordon, M.S., Notides, A.C. Computer modelling of estradiol interactions with the estrogen
receptor. J. Steroid Biochem., 25: 177- 181 (1986)
Gorski, J., Toft, D., Shyamala, G. Hormone receptors: Studies on interaction of estrogen
with the uterus. Recent Prog. Horm. Res., 24: 45 (1968)
Gorski, J., Gannon, F. Current models of steroid hormone action: A Critique. Ann. Rev.
Physiology p425 (1976)
Gotteland, M., Desauty, G., Delarue, J-C., Liu, L., May, E. Human estrogen receptor
messenger RNA variants in both normal and tumour breast tissues. Mol. Cell. Endocrinol.,
112:1-13(1995)
Goulding , H., Nik Abdul Rashid, N.F., Roberston, J.F. Bell, J.A., Elston, C.W., Blarney,
R.W., Ellis, I.O. Assessment o f angiogenesis in breast carcinoma. An important factor in
prognosis? Hum. Pathol., 26: 1196-1200(1995)
Graff, J.R., Herman, J.G., Lapidus, R.G., Chopra, H., Xu, R., Jarrard, D.F., Isaacs, W.B.,
Pitha, P.M., Davidson, N.E., Baylins, S.B. E-cadherin expression is silenced by DNA
hypermethylation in human breast and prostate carcinomas. Cancer Res., 55, 5195-5199
(1995)
240
Graham, R.C. Jr., Kamovsky, M J. The histochemical demonstration of uricase activity J.
His toe hem. Cytochem., 14: 291-302 (1966)
Graham, M.L., Krett, N.L., Miller, L.A., Leslie, K.K., Gordon, D.F., Wood, W.M., Wei,
L.L., Horwitz, K.B. T47D C0 cells, genetically unstable and containing estrogen receptor
mutations are a model for the progression of breast cancers to hormone resistance. Cancer
Res., 50: 6208-6217(1990)
Grandien K.F.H., Berkenstam, A., Nilssom, S., Gustafsson, J-A. Localisation of DNase I
hypersensitivity sites in the human oestrogen receptor gene correlates with the
transcriptional activity o f two differentially used promoters. J. Mol. Endocrinol., 10: 267277(1993)
Greaves, D.R., Patient, R.K. (AT)n is an interspersed repeat in the Xenopous genome.
EM BOJ., 4: 2617-2626 (1985)
Green, S., Walter, P., Kumar, V., Krust, A., Bomert, J-M., Argos, P., Chambon, P. Human
oestrogen receptor cDNA: sequence, expression and homology to v-erb-A. Nature, 320:
134-139(1986)
Green, S., Chambon, P. Oestradiol induction of a gluticorticoid responsive gene by a
chimeric receptor. Nature, 325: 75-78 (1987)
Greene, G.L., Gilna, P., Waterfield, N. Baker, A., Hort, Y., Shine, J. Sequence and
expression of human estrogen receptor complimentary DNA Science, 231: 1150-1154
(1986)
Greene, S., Kumar, V., Theulaz, I., Wahli, W., Chambon, P. The N-terminal DNA-binding’
zinc-finger’ o f the oestrogen and gluticorticoid receptors determines target gene specificity.
EMBOJ., 320: 3037-3044 (1988)
Gronemeyer, H. Transcription activation by estrogen and progesterone receptors. Ann. Rev.
Genet., 25: 89-123 (1991)
Guerin, M., Barrois, Terrier, M.J., Spielmann, M., Riou, G. Overexpression of either c-myc
or c-erb-B2/neu proto-oncogenes inhuman breast carcinomas: correlation with poor
prognosis. Oncogene Res., 3: 21-31 (1988)
Gullick, W.J., Love, S.B., Wright, C., Barnes, D.M., Gusterson, B., Harris, A.L., Altman,
D.G. C-erbB2 protein overexpression in breast cancer is a risk factor in patients with
involved and uninvolved lymph nodes. Br. J. Cancer, 63: 434-438 (1991)
Gupta, P.W., Sahota, A., Boyadjiev, S.A., Bye, S., Shao, C., O’Neill, J.P, Hunter, T.C.,
Albertini, R.J., Stambrook, P.J., Tischfield, J.A. High frequency in vivo loss of
heterozygosity is primarily a consequence of mitotic recombination. Cancer Res., 57: 11881193 (1997)
Haenszel, W. Cancer mortality rates among the foreign bom in the United States. J. Natl.
Cancer Inst., 26: 37-132 (1961)
241
Han, H. J., Yanagisawa, A., Kato, Y., Park, J. G., and Nakamura, Y. Genetic instability in
pancreatic cancer and poorly differentiated type of gastric cancer. Cancer Res., 53: 50875089(1993)
Hankins, G.R., De Souza, A.T., Bentley, R.C., Patel, M.R., Marks, J.R., Iglehart, J.D., Jirtle,
R.L. M6P/IGF2 receptor: a candidate breast tumour suppressor gene. Oncogene, 12: 20032009(1996)
Harris, J.R., Heilman, S., Henderson, I.C., Kinne, D. W. (eds) Breast Diseases, 2nd Edition,
J.B. Lippincott Company (1991)
Hartveit, F., Naartmann-Moe., Stoa, K.F. Early recurrence in oestrogen receptor negative
breast carcinomas. Acta Scand., 146: 93-95 (1980)
Harwood, J., Tachibana, A., Davis, R., Bhattacharyya, N.P., Meuth, M. High rate of
multilocus deletion in a human tumour cell line. Hum. Mol. Gene., 2: 165-171 (1993)
Hawley, D.K., Roeder, R.G. Separation and partial characterization of three functional steps
in transcription initiation by human RNA polymerase II. J. Biol. Chem., 260: 8163-8172
(1985)
Hebert, E., Herbelin, C., Bougnoux, P. Analysis o f the IGF-II receptor gene copy number in
breast carcinoma. Br. J. Cancer, 69, 120-124 (1994)
Heintz, N.H., Lelie, K.O., Rogers, L.A. Howard, P.L. Amplification of the c-erbR-2
oncogene and prognosis o f breast adenocarcinoma. Arch, Pathol. Lab. Med, 114: 160-163
(1990)
Henderson, B.E., Pike, M.C., Casagrande, J.T. Breast cancer and the estrogen window
hypothesis. Lancet, 2: 363 (1981)
Henry, J.A., Nicholson, S., Hennessym C., Lenard, T.W.J., May, F.E.B., Westley, B.R.
Expression o f oestrogen regulated pNR - 2 mRNA in human breast cancer: relation to
oestrogen receptor mRNA levels and response to endocrine therapy. Br. J. Cancer, 50:
3832-3837(1989)
Henry, J.A., Piggot, N.H., Mallick, U.K., Nicholson, S., Fardon, J.R., Westley, B.R., May,
F.E.B. PNR-2/pS2 immunohistochemical staining in breast carcinoma: correlation with
prognostic factors and endocrine response. Br. J. Cancer., 62: 615-622 (1991)
Herman, J.G., Latif, F., Weng, Y., Lerman, M.I., Zbar, B., Liu, S., Samid, D., Duan, D.S.R.,
Gnarra, J.R., Linehan, W.M., Baylins, S.B. Silencing of the VHL tumour-suppressor gene
by DNA methylation in renal carcinoma. Proc. Natl. Acad. Sci. USA, 91, 9700-9704 (1994)
Hermon, C., Beral. V. Breast cancer mortality rates are levelling off or beginning to decline
in many Western countries: analysis of timed trends, age-cohort and age-period models of
breast cancer mortality in 20 countries. Br. J. Cancer, 73: 955-960 (1996)
Hernandez, N. TBP, a universal eukaryotic transcription factor? Genes Devel., 7: 1291-1308
(1993)
242
Hill, S., Rodgers, C.S., Hullten, M.A. Cytogenetic analysis in human breast carcinoma. II.
Seven cases in the triploid/tetraploid range investigated using direct preparations. Cancer
Genet. Cytogenet., 24: 45-62 (1987)
Hirata, S., Yamada-Mouri, N., Nara, M., Takizawa, M., Ito, H., Kato, J. Presence of
alternatively spliced estrogen receptor mRNA variants in normal human endometrium and
endometrial cancer. Endocrine J., 42: 289-293 (1995)
Hoffman, A., Sinn, E., Yamamoto, T., Wang, J., Roy, A., Horikoshi, M., Roeder, R.G.
Highly conserved core domain and unique N terminus with presumptive regulatory motifs in
a human TATA factor (TFIID). Nature, 346: 387-390 (1990)
Hoi, F.A., Geurds, M.P.A., Hamel, B.C.J., Mariman, E.C.M. Improving the polymorphism
content o f the 3* UTR o f the human IGFR2 gene. Hum. Mol. Genet., 1: 347 (1992)
Houlston, R. S., Collins, A., Slack, J., and Morton, N. E. Dominant genes for colorectal
cancer are not rare. Ann. Hum. Genet., 56: 99-103, 1992
Howell, A., Anderson, E., Laidlaw, I., Schor, A., Schor, S., Potten, C. Cyclic activity and
ageing o f the human breast: clues to assessment of risk and strategies for prevention.
Endocrine. Therapy O f Br. Cancer, ESO Monograph, pp27-46.
Hsu, S.M., Rannie, L., Fanger, H. Use of avidin-biotin-peroxidase technique: a comparison
between ABC and unlabelled antibody (PAP) procedures. Am. J. Clin. Pathol., 75: 734-738
(1981)
Huang, J., Papadpoulos, N., McKinley, A. J., Farrington, S. M., Curtis, L. J., Wyllie, A. H.,
Zheng, S., Willson, J. K. V., Markowitz, S. D., Morin, P., Kinzler, K. W., Vogelstein, B.,
and Dunlop, M. G. APC mutations in colorectal tumours with mismatch repair deficiency.
Proc. Natl. Acad. Sci. USA., 93: 9049-9054 (1996)
Hynes, R.O. Integrins, versatility, modulation and signalling in cell adhesion. Cell, 69: 1125(1992)
Ignor-Trowbridge, D.M., Teng, C.T., Ross, K.A., Parker, M.G., Korach, K.S., McLachlan,
J.A. Peptide growth factors elicit estrogen receptor-dependant transcriptional activation of
an estrogen-responsive element. Mol. Endocrinol., 7: 992-998 (1993)
Ionov, Y., Peinado, M.A. Malkhosyan, S., Shibata, D., Perucho, M. Ubiquitous somatic
mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis.
Nature, 363: 558-561 (1993)
Inoue, S., Orimo, A., Hosoi, T., Kondo, S., Toyoshima, H., Kondo, T., Igegami, A., Ouchi,
Y., Orimo, H., Muramatsu, M. Genomic binding-site cloning reveals an estrogen-responsive
gene that encodes a ring finger protein. Proc. Natl. Acad. Sci. USA, 90: 11117-11121 (1993)
Issa, J-P. J., Vertino, P.M., Wu, J., Sazawal, S., Celano, P., Nelkin, B.D., Hamilton, S.R.,
Baylin, S.B. Increased cytosine DNA metyltransferase activity during colon cancer
progression. J. Natl. Cancer Inst., 85: 1235-1240(1993)
243
Issa, J-P. J., Ottaviano, Y.L., Celano, P., Hamilton, S.R., Davidson, N.E., Baylin, S.B.
Methylation o f the oestrogen receptor CpG islands links ageing and neoplasia in human
colon. Nature Genet., 7: 536-540 (1994)
Issa, J-P. J., Zehnbauer, B.A., Civin, C.I., Collector, M.I., Sharkis, S.J., Davidson, N.E.,
Kaufmann, S.H., Baylin, S.B. The estrogen CpG island is methylated in most hematopoietic
neoplasms. Cancer Res., 56: 973-977 (1996)
Ito, T., Yoshimoto, M., Iwase, T., Watanabe, S., Katagiri, T., Harada, Y., Kasumi, F.,
Yasuda, S., Mitomi, T., Emi, M., Nakamura, Y. Association of genetic alterations on
chromosome 17 and loss o f hormone receptors in breast cancer. Br. J. Cancer, 71: 438-441
(1995)
Iwase, H., Greenman, J.M., Barnes, D.M., Bobrow, L., Hodgson, S., Mathew, C.G. Loss of
heterozygosity o f the oestrogen receptor gene in breast cancer. Br. J. Cancer, 71, 448-450
(1995)
Jacobs, C., Rubsaamen, H. Expression of pp60C-src protein kinase in adult and fetal human
tissue - high activities in some sarcomas and mammary carcinomas. Cancer Res., 43: 16961702(1983)
Jaquuemier, J., Moles, J.P., Penault-Llorca, F., Adelaide, J., Torrente, M., Viens, P.,
Bimbaum, D., Theillet, C. p53 immunohistochemical analysis in breast cancer with four
monoclonal antibodies: comparison o f staining and PCR: SSCP results. Br. J. Cancer, 69:
846-852(1994)
Jeltsch, J.M., Turcotte, B., Gamier, J.M., Lerouge, T., Krowowskim Z., Gronemeyer, H.,
Chambon, P. Characterisation o f multiple mRNAs originating from the chicken
progesterone receptor gene. Evidence for a specific transcript encoding form A. J. Biol.
Chem., 265: 3967-3974 (1990)
Jensen, E.V., Jacobson, H. Basic guide to the mechanism of oestrogen action. Recent Prog.
Hormone Res., 18: 317 (1962)
Jensen, E.V., Suzuki, T., Kuwashima, T. A two step mechanism of estradiol with rat uterus.
Proc. Nat. Acad. Sci., USA 59: 632 (1968)
Jensen, E.V., Mohia, S., Gorell, T. Estrophille to nucleophile in two easy steps. J. Steroid
Biochem., 3: 445 (1972)
Jiang, S.Y., Jodan, V.C. Growth regulation of estrogen receptor-negative breast cancer cells
transfected with complementary DNAs for estrogen receptor. J. Natl. Cancer Inst., 84: 580591 (1992)
Jirtle, R.L., Haag, J.D., Ariazie, A., Gould, M.N. Increased mannose 6 -phosphate/insulin­
like growth factor II receptor and transforming growth factor pi levels during monoterpeneinduced regression o f mammary tumour. Cancer Res., 53, 3849-3852 (1993)
Joel, P.B., Traish, A.M., Lannigan, D.A. Estradiol and phorbol ester cause phosphorylation
of serine 118 in the human estrogen receptor. Mol. Endocrinol., 9: 1041-1052 (1995)
244
Jones, J.L., Critchley, D.R., Walker, R.A. Alterations of stromal protein and integrin
expression in breast - a marker of premalignant change?. J. Pathol, 167: 399-406 (1992)
Jones, J.L., Royall, J.E., Walker, R.A. E-cadherin relates to EGFR expression and lymph
node metastasis in, primary breast carcinoma. Br. J. Cancer, 74: 1237-1241 (1996)
Jonsson, M., Johannsson, O., and Borg, A. Infrequent occurrence of microsatellite instability
in sporadic and familial breast cancer. Eur. J. Cancer, 31A: 2330-2334 (1995)
Kani, Y., Oda, T., Tsuda, H., Ochiai, A., Hirohashi, S. Point mutations of the E-cadherin
gene in invasive lobular carcinoma o f the breast. Jpn. J. Cancer Res., 85: 1035-1039 (1994)
Kariya, Y., Kato, K., Hayashizaki, Y., Himeno, S., Tarui, S., Matsubaru, K. Revision of
consensus sequence o f human Alu repeats - a review. Gene, 53: 1-10 (1987)
Kamik, P.S., Kulkami, S., Liu, X-P., Budd, G.T., Burkowski, R.M. Estrogen receptor
mutations in tamoxifen-resistant cancer. Cancer Res., 54: 349-353 (1994)
Kamick, P., Plummer, S., Casey, G., Myles, J., Tubbs, R., Crowe, J., and Willimams, B. R.
G. Microsatellite instability at a single locus (D11S988) on chromosome 1 l p l 5.5 as a late
event in mammary tumourigenesis. Hum. Mol. Genet., 10: 1889-194 (1995)
Karran, P. Appropriate partners make good matches. Science, 268: 1857-1858 (1995)
Karran, P. Microsatellite instability and DNA mismatch repair in human cancer. Semin.
Cancer. Biol, 7: 15-24 (1996)
Kasami, M., Vnencak-Jones, C. L., Manning, S., Dupont, W. D., and Page, D. L. Loss of
heterozygosity and microsatellite instability in breast hyperplasia. Am. J. Path, 150: 19251932(1997)
Kashiwaba, M., Tamura, G., Suzuki, Y., Maesawa, C., Ogasawara, S.,Sakata, K., Sotadate,
R. Epithelial-cadherin gene is not mutated in ductal carcinoma of the breast. Jpn. J. Cancer
Res., 8 6 : 1054-1059(1995)
Kastner, P., Krust, A., Turcotte, B., Stropp, U., Tora, L., Gronmeyer, H., Chambon, P. Two
distinct estrogen-regulated promoters generate transcripts encoding the two functionally
different human progesterone receptor forms A and B. EMBOJ, 9: 1603-1614 (1990a)
Kastner, P., Krust, A., Mendelson, C., Gamier, J-M., Zelent, A., Leroy, P., Staub, A.,
Chambon, P. Murine isoforms o f the retanoic acid receptor g with specific patterns of
expression. Proc. N atl Acad. Sci. USA, 87: 2700-2704 (1990b)
Katzenellenbogen, B.S, Norman, M.J. Multi-hormone regulation of the progesterone
receptor in MCF-7 human breast cancer cells: intereactionships among insulin/insulin-like
growth factor-I, serum and estrogen. Endocrinology, 126: 891-898 (1990)
Katzenellebogen, B.S. Antiestrogen resistance: mechanisms by which breast cancer cells
undermine the effectiveness o f endocrine therapy. J. Natl. Cancer Inst., 83: 1434-1435
(1991)
245
Keaveney, M., Neilan, J., Gannon, F. An analysis by restriction enzymes of the genomic
structure o f the 3 ’ untranslated region of the human oestrogen receptor gene. Biochem.
Biophys. Acta. 1007: 289-294 (1989)
Keaveney, M., Klug, J., Dawson, M.T., Nestor, P.V., Neilan, J., Forde, R.C., Gannon, F.
Evidence for a previously unidentified upstream exon in the human oestrogen receptor gene.
J. Mol. Endocrin., 6 : 111-115 (1991)
Keaveney, M., Klug, J., Gannon, F. Sequence analysis of the 5’ flanking region of the
human estrogen receptor gene. DNA Sequence, 2: 347-358 (1992)
Keaveney, M., Parker, M.G., Gannon, F. Identification of a functional role for the 3’ region
o f the human oestrogen receptor gene. J. Mol. Endocrin., 10: 143-152 (1993)
Kembler, R. Classical cadherins. Semin. Cell Biol., 3: 149-155 (1992)
Kembler, R. From cadherins to catanins: cytoplasmic protein interactions and regulation of
cell adhesion. Trends Genet., 9: 317-321 (1993)
Kerangueven, F., Essioux, L., Dib, A., Nogauchi, T., Allione, F., Geneix, J., Longy, M.,
Lidereau, R., Eisinger, F., Pebusque, M-J., Jacquemier, J., Bonaiti-Pellie, C., Sobol, H.,
Bimbaum, D. Loss o f heterozygosity and linkage analysis in breast cancer: indication for a
3rd susceptibility gene on the short arm of chromosome 8 . Oncogene, 10: 1023-1026 (1995)
Kern, S.E., Pitenpol, J.A., Thiagalingam, S., Seymour, A., Kinzler, K.W., Vogelstein, B.
Oncogenic forms o f p53 inhibit p53-regulated gene-expression. Science, 256: 827-8630
(1992)
Keydar, I., Chen, L., Karby, S., Weiss, F. R., Delarea, J., Radu, M., Chaitcik, S., and
Brenner, H. J. Establishment and characterization of a cell line of human breast carcinoma
origin. Eur. J. Cancer, 15: 659-670 (1979)
Khorana, S., Gagel, R.F., Cote, G.J. Direct sequencing of PCR products in agarose gel
slices. Nucleic Acids Res., 22(16): 3425-3426 (1994)
King, R.J.B., Cowan, D.M., Lumen, D.R. The uptake of (6,7-^H) oestradiol by dimethylbenzathracene-induced rat mammary tumours. J. Endocrinol., 32: 83 (1963)
King, W.L., DeSombre, E.R., Jensen, E.V., Greene, G.L. Comparison of
immunocytochemical and steroid binding assays for estrogen receptor in human breast
tumours. Cancer Res., 45: 293-304 (1985)
King, B.L., Carcangiu, M-L., Carter, D., Kiechle, M., Pfisterer, J., Pfleiderer, A., Kacinski,
B.M. Microsatellite instability in ovarian neoplasms. B. J. Cancer, 72: 376-382 (1995)
Kinzler, K. W., and Vogelstein, B. Lessons from hereditary colorectal cancer. Cell, 87: 159170(1996)
Kinzler, K. W., and Vogelstein, B. Gatekeepers and Caretakers. Nature, 386: 761-762
(1997)
246
Klein-Hitpass, L., Schorpp, M., Wagner, U., Ryfell, G.U. An estrogen-responsive element
derived from the 5’ flanking region of the Xenoupus vitellogenin A2 gene functions in
transfected human cells. Cell, 46: 1053-61 (1986)
Klein-Hitpass, L., Ryfell, G.U., Heitlinger, E., Cato, A.C.B. A 13 bp palindrome is a
functional estrogen responsive element and interacts specifically with estrogen receptor.
Nucleic Acids Res., 16: 647-663 (1988a)
Klein-Hitpass, L., Kaling, M., Ryffel, G.U. Synergism of closely adjacent oestrogenresponsive elements increase their regulatory potential. J. M ol Biol., 201: 537-544 (1988b)
Klein-Hitpass, L., Tsai, S.Y., Greene, G.L., Clark, J.H., Tsai, M-J., O ’Mally, B.W. Specific
binding o f estrogen-receptor to the estrogen response element. Mol., Cell., Biol., 9: 43-49
(1989)
Klein-Hitpass, L., Tsai, S.Y., Weigel, N.J., Riley, D., Rodriguez, R., Schrader, W.T., Tsai,
M-J., O'Malley, B.W. Native progesterone receptor stimulates cell-free transcription of a
target gene. Cell, 60: 247 (1990)
Klijn, J.G.M., Bems, E.M.M.J., Foekens, J.A. Prognostic factors and response to therapy in
breast cancer. Cancer Surv., 18: 165-198 (1993)
Klinge, C.M., Peale, FF.V., Jr., Hilf, R., Bambara, R.A., Zain, S. Co-operative estrogenreceptor interaction with consensus or variant estrogen responsive elements in vitro. Cancer
Res., 52: 1073-1081 (1992a)
Klinge, C.M., Bambara, R.A., Hilf, R. Antiestrogen-liganded estrogen-receptor interaction
with estrogen responsive elements in vitro. J. Steroid Biochem. Mol. Biol, 43: 249-262
(1992b)
Klotz, D.M., Castles, C.G., Hill, S.M. Differential expression of wild type and variant ER
mRNA clones o f MCF-7 breast cancer cells may account for differences in estrogenresponsiveness (abstract 1070). Proceedings o f the 76th Annual Meeting o f the Endocrine
Society, Anaheim, CA, p. 468. Endocrine Society Press: Bethesda MD (1994)
Klug, A., Rhodes, D. 4,Zinc fingers” : a novel protein motif for nucleic acid recognition.
Trends Biochem. Sci., 12: 464-469 (1987)
Knudson, A.G. Hereditary cancers disclose a class of cancer genes. Cancer, 63: 1888-1891
(1989)
Koemer, F.C., Goldberg, D., Edgerton, S.M., Schwartz, L.H. pS2 protein and steroid
hormone receptors in invasive breast carcinomas. Int. J. Cancer, 52: 183-188 (1992)
Kojima, S., Nara, K., Rifkin, D.B. Requirement for transglutaminase in the activation of
latent transforming growth factor-p in bovine endothelial cells. J. Cell. Biol., 121, 439-448
(1993)
Koreth, J., Bethwaite, P. B. and O'D McGee, J. Mutation at 11 q23 in human non-familial
breast cancer. A microdissection microsatellite analysis. J. Pathol, 176: 11-18 (1995)
247
Korinek,V., Barker, N, Morin, P. J, vanWichen, D., deWeger, R., Kinzler, K. W.,
Vogelstein, B., and Clevers, H. Constitutive transcriptional activation by a beta-catenin-Tcf
complex in APC(-/-) colon carcinoma. Science, 275: 1784-1787 (1997)
Komfeld, S. Structure and function of the mannose 6 -phosphate/insulin like growth factor
receptors. Ann. Rev. Biochem., 61: 307-3030 (1992)
Kovacina, K.S., Steele-Perkins, G., Purchio, A.F., Lioubin, M., Miyazono, K., Heldin, C.H.,
Roth, R.A. Interactions o f recombinant and platelet transforming growth factor beta 1
precursor with the insulin-like growth factor receptor II/mannose 6 -phosphate receptor.
Biochem. Biophy. Res. Commun. 160: 393-403 (1989)
Kozak, M. Compilation and analysis o f sequences upstream from the translational start site
in eukaryotic messenger-RNAs. Nucleic Acids Res., 12: 857-872 (1984)
Kraus, W.L., Mclnemey, E.M., Katzenellenbogen, B.S. Amino-terminal and carboxylterminal regions o f a steroid-hormone nuclear receptor. Proc. Natl. Acad. Sci. USA, 92:
12314-12318(1995)
Krontiris, T.G., Delvin, B., Karp, D.D., Robert, J.J., Risch, N. An association between the
risk of cancer and mutations in the H-ras-1 mini-satellite locus. New Eng. J. Med., 329: 517523 (1993)
Krust, A., Green, S., Argos, P., Kumar, V., Walter, P., Bomert, J.M., Chambon, P. The
chicken oestrogen receptor sequence: homology with v-erbA and the human oestrogen and
gluticorticoid receptors. EM BOJ., 5: 891-897 (1986)
Kuiper, G.G.J.M., Enmark, E., Peltohuikko, M., Nilsson, S., Gustafsson, J.A. Cloning of a
novel estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. U.S.A.,
93: 5925-5930(1996)
Kumar, V., Green, S., Staub, A., Chambon, P. Localisation of the oestradiol-binding and
putative DNA-binding domains o f the human oestrogen receptor. EMBO J., 5: 2231-2236
(1986)
Kumar, V., Green, S, Stack, G., Berry, M., Jin, J.R., Chambon, P. Functional domains of the
human estrogen receptor. Cell, 51: 941-951 (1987)
Kumar, V., Chambon, P. The estrogen receptor binds tightly to its responsive element as a
ligand-induced homodimer. Cell, 55: 145-156 (1988)
Kushner, P.J., Hort, E., Shine, J., Baxter, J.D., Greene, G.L. Construction of cell lines that
express high levels o f human estrogen receptor and are killed by oestrogens. Mol.
Endocrinol., 4: 1465-1473 (1990)
Kushner, P.J., Baxter, J.D., Duncan, K.G., Lopez, G.N., Schaufele, F., Unt, R.M., Webb, P.,
West, B.L. Eukaryotic regulation elements lurking in plasmid DNA - the activator protein-1
site in PUC. Mol., Endocrinol., 8 : 405-407 (1994)
248
Kute, T.E., Muss, H.B., Anderson, D. Relationship of steroid receptor, cell kinetics and
clinical status in patients with breast cancer. Cancer Res., 41: 3524-3529 (1981)
Lagios, M. D., Westadahl, P. R., Margolin, F. R., and Rose, M. R. Duct carcinoma in situ.
Relationship o f extent of non-invasive disease to the frequency of occult invasion,
multicentricity, lymph node metastes and short-term treatment failures. Cancer, 50: 13091314(1982)
Lagios, M. D., Margolin, F., R., Westadahl, P. R., Rose, M. R. Mammographically detected
duct carcinoma in situ: in frequency o f local recurrence following tylectomy and prognostic
effect o f nuclear grade on local recurrence. Cancer, 63: 618-624 (1989)
Lahooti, H., White, R., Danielian, P.S., Parker, M.G. Characterization of ligand dependant
phosphorylation o f the estrogen receptor. Mol. Endorinol., 8 : 182-188 (1994)
Lahooti, H., White, R., Hoare, S.A., Rahman, D., Pappin, D.J.C., Parker, M.G.
Identification o f phosphorylation sites in the mouse oestrogen receptor. J. Steroid Biochem.
Mol. Biol., 55: 305-313 (1995)
Lakhani, S. R., Collins, N., Sloane, J. P., and Stratton, M. R. Loss of heterozygosity in
lobular carcinoma in situ o f the breast. J. Clin. Pathol: Mol. Pathol., 48: M74-M78 (1995a)
Lakhani, S. R., Collins, N., Sloane, J. P., and Stratton, M. R. Atypical ductal hyperplasia o
fthe breast: clonal proliferation with loss of heteerozygosity on chromosomes 16q and 17p.
J. Clin. Pathol., 48: 611-615 (1995b)
Lannigan, D.A., Notides, A.C. A novel mechanism for eukaryotic gene expression: the
involvement o f DNA tertiary structure in oestrogen recognition of its target nucleotide
sequence. Biochem. Pharmacol., 40: 2579-2585 (1990)
Lapidus, R.G., Ferguson, A.T., Ottaviano, Y.L., Pari, F.F., Smith, H.S., Weitzman, S.A.,
Baylin, S.B., Issa, J-P. J., Davidson, N.E. Methylation of estrogen and progesterone receptor
gene 5’ CpG islands correlates with lack of estrogen and progesterone receptor gene
expression in breast tumours. Clin. Cancer. Res., 2: 805-810 (1996)
Larsson, C., Bystrom, C., Skoog, L., Rotstein, S., Nordenskjold. Genomic alterations in
human breast carcinomas. Genes, Chromosomes, and Cancer, 2: 191-197 (1990)
Lasko, D., Cavanee, W., Nordenskjold, M. Loss of constitutive heterozygosity in human
cancer. Ann Rev. Genet., 25: 281-314 (1991)
La Spada, A.R., Wilson, E.M., Lubahn, D.B. Harding, A.E., Fischbeck, K.H. Androgen
receptor mutations in X-linked spinal and bulbar muscular atrophy. Nature, 352: 77-79
(1991)
La Thangue, N. B. DRTF1/E2F: an expanding family of heterodimeric transcription factors
implicated in cell-cycle control. Trends Biochem. Sci., 19: 108-114 (1994)
Laureys, G., Barton, D.E., Ullrich, D.E., Francke, U. Chromosomal mapping of the gene for
the type II insulin-like growth factor receptor/cation-independent mannose 6 -phosphate
receptor in man and mouse. Genomics, 3, 224-229 (1988)
249
La Vecchia, C., Negri, E., Franceschi, S., Favero, A., Nanni, O., Filiberti, R., Conti, E.,
Montella, M., Veronesi, A., Ferraroni, M., Decarli, A. Hormone replacement treatment and
breast cancer risk: a co-operative Italian study. Br. J. Cancer, 72: 244-248 (1995)
Leach, F. S., Nicolaides, N. C., Papadopoulos, N., Liu, Jen, Parsons, R., Peltomaki,
Sistonen, P., Aaltonen, L., Nystrom-Lahti, M., Guan, X. Y., Zhang, J., Meltzer, P., Yu, J.,
Kao, Chen, D., D., Cerosaletti, K., foumier, R., Todd, S., Lewis, T., Leach., R., Naylor, S.,
Weissenbach, J., Mecklin, J., Jarvinen, H., Petersen, G., Hamilton, S., Green, J., Watson, P.,
Lynch, H., Trent, J., de la Chappelle, A., Kinzler, K., and Vogelstein, B. Mutations of a
mutS homologue in hereditary nonpolyposis colorectal cancer. Cell, 75: 1215-1225 (1993)
Lee, S.J., Nathams, D. Proliferin secreted by cultured cells binds to mannose 6 -phosphate
receptors. J. B iol Chem., 263, 3521-3527 (1988)
Lees, J.A., Fawell, S.E., White, R., Parker, M.G. A 22 amino-acid peptide restores DNAbinding activity to dimerisation defective mutants of the estrogen receptor. Mol. Cell. Biol.,
10: 5529-5531 (1990)
Leff, S.E., Rosenfeld, M.G., Evans, R.M. Complex transcriptional units: diversity in gene
expression by alternative RNA processing. Annu. Rev. Biochem., 55: 1091-1117 (1986)
Leong, A.S.Y., Lee, A.K.C. Biological indices in the assessment of breast cancer. J. Clin.
Mol. Pathol, 48: M22-M28 (1995)
Leppert, M., Dobbs, M., Scambler, P., O’Connell, P., Nakamura, Y., Stauffer, D., Woodard,
S., Burt, R., Hughes, J., Gardner, E., and White, R. The gene for familial polyposis coli
maps to the long arm o f chromosome 5. Science, 238: 1411-1443 (1987)
Leroy, P., Krust, A., Zelent, A., Mendelsohn, C., Gamier, J.M., Kastner, P., Dierich, A.,
Chambon, P. Multiple isoforms o f the mouse retanoic acid receptor a are generated by
alternative splicing and differential induction by retanoic acid. EMBOJ., 10: 59-69 (1991)
Leslie, K.K., Tasset, D.M., Horwitz, K.B. Functional analysis of a mutant estrogen receptor
isolated from T47Dco breast cancer cell. Am. J. Obstet. Gynecol., 166: 1053-1061 (1992)
Levine, A.J. Tumour suppressor genes. Bioessays, 12: 60-66 (1990)
Lew, E.A., Garfunkel, L. Variations in mortality by weight among 750, 000 men and
women. J. Chron. Dis., 563: 32 (1979)
Leygue, E.R., Watson, P.H., Murphy, L.C. Estrogen receptor variants in normal human
mammary tissue. J. Natl. Cancer. Inst., 8 8 : 284-290 (1996a)
Leygue, E., Huang, A., Murphy, L.C., Watson, P.H. Prevalence of estrogen receptor variant
messenger RNAs in human breast cancer. Cancer Res., 56: 4324-4327 (1996b)
Li, C.Y., Ziesmer, S.C., Lazcano-Villareal, O. Use of azide and hydrogen peroxide as an
inhibitor for endogenous peroxidase in the immunoperoxidase method. J. Histochem.
Cytochem., 35: 1457-1460(1987)
250
Li, E., Beard, C., Jaenisch, R. Role for DNA methylation in genomic imprinting. Nature,
366: 362-365 (1993)
Lidereau, R., Mathieu-Mahul, D., Escot, C., Theillet, C., Champeme, M.H., Cloe, S.,
Mauchauffe, M., Ali, I., Amione, J., Callahan, R., Larsen, C.J. Genetic variability of proto­
oncogenes for breast cancer risk and prognosis. Biochemie, 70: 951-959 (1988)
Lindblom, A., Rotstein, S., Skoog, L., Nordensjold, M., Larsson, C. Deletions on
chromosome 16 in primary breast carcinomas are associated with development of distant
metastases. Cancer Res., 53: 3707-3711 (1993)
Lindstedt, A.D., West, N.B., Brenner, R.M. Analysis of monomeric-dimeric states of the
estrogen receptor with monoclonal antiestrophilins. J. Steroid Biochem., 24: 677-686 (1986)
Lippman, M. E., Allegra, J.C. Quantitative oestrogen receptor analyses: the response to
endocrine and cytotoxic chemotherapy in human breast cancer and the disease free interval.
Cancer, 46: 2884-2888 (1980)
Lippman, M.E., Dickson, R.B., Bates, S., Knabbe, C., Huff, K., Swain, S., McManaway,
M., Bronzert, D., Kasid, A., Gelmann, E.P. Autocrine and paracrine growth regulation in
human breast cancer. Breast Cancer Res. Treat., 1: 59-70 (1986)
Lipponen, P., Ji, H., Aaltomaa, S., Syrjanen, K. p53 protein expression in breast cancer as
related to histopathological characteristics and prognosis. Int. J., Cancer, 55: 51-56 (1993)
Lipponen, P., Saarelainen, E., Ji, H., Aaltomaa, S., Syijanen, K. Expression of E-cadherin
(E-Cd) as related to other prognostic factors and survival in breast cancer. J. Pathol., 174:
101-109(1994)
Liu, E., Dollbaum, C., Scott, G. Molecular lesions involved in the progression of human
breast cancer. Oncogene, 3,323-327 (1988)
Liu, X., Miller, C.W., Koeffler, P.H., Berk, A.J. The p53 activation domain binds the TATA
box-binding polypeptide in holo-TFIID, and a neighbouring p53 domain inhibits
transcription. Mol. Cell. Biol., 13: 3291-3300 (1993)
Liu, B., Nicolaides, N.C., Markowitz, S., Wilson, J.K., Parsons, R.E., Jen, J., Papadopolous,
N., Peltmomaki, P., de la Chappelle, A., Hamilton, S.R. Mismatch repair gene defects in
sporadic colorectal cancers with microsatellite instability. Nature Genet., 9: 48-55 (1995)
Lobo, R.A. Benefits and risks o f oestrogen replacement therapy. Am. J. Obs. Gynae., 173:
982-989(1995)
Lodata, R.J.F., Maguire, H.C., Greene, M.L., Weiner, D.B., LiVolisi, VIA.
Immunohistochemical evaluation o f c-erb-B2 oncogene expression in ductal carcinoma in
situ and atypical ductal hyperplasia of the breast. Mod, Pathol., 3: 449 (1990)
Loeb, L.A. Microsatellite instability: Marker of a mutator phenotype in cancer. Cancer Res.,
54: 5059-5063 (1991)
251
Loeb, L A. Microsatellite instability: marker of a mutator phenotype in cancer. Cancer Res.,
54: 5059-5063 (1994)
London, S.J., Connolly, J.L. Schnitt, S.J., Colditz, G.A. A prospective study of benign
breast disease and the risk of breast cancer.
267:941-944(1992)
Louis, D.N., von Deimling, A., Seizinger, B.R. A (CA)n dinucleotide repeat assay for
evaluating loss o f allelic heterozygosity in small and archival human brain tumour
specimens Am. J. Path., 141: 777-782 (1992)
Love, R.R.., Mazess, R.B., Barden, H.S., Epstein, S., Newcomb, P.A., Jordan, V.C.,
Carbone, P.P., DeMets, D.L. Effects of tamoxifen on bone-mineral density in
postmenapausal women with breast cancer. N. Engl. J. Med., 326: 852-856 (1992)
Lovekin, C., Ellis, I.O., Locker, A., Robertson, J.F.R., Bell, J., Nicholson, R., Gullick, W.J.,
Elstom, C.W., Blarney, R.W. C-erb-B2 oncoprotein expression and prognosis in operable
breast cancer. Br. J. Cancer, 63: 439-443 (1991)
Lubin, F., Ruder, A.M., Wax, Y., Modan, B. Overweight and changes in weight throughout
adult life in breast cancer aetiology. Am. J. Epidemiol., 122: 579-588 (1985)
Luisi, B.F., Xu, W.X., Otwinowski, Z., Freedman, L.P., Yammamoto, K.R., Sigler, P.B.
Crystallographic analysis o f the interaction of the gluticorticoid receptor with DNA. Nature,
352:497(1991)
Lundberg, C., Skoog, L., Cavanee, W.K., Nordenskjold, M. Loss of heterozygosity in
human ductal breast tumours indicates a recessive mutation on chromosome 13. Proc: Natl.
Acad. Sci. USA, 84: 2372-2376 (1987)
Ma, Z.Q., Santagati, S., Patrone, C., Pollio, G., Vegeto, E., Maggi, A. Insulin-like growth
factors activate estrogen receptor to control the growth and differentiation of the human
neuroblastoma cell line SK-Er3. Mol. Endocrinol., 8 : 910-918 (1994)
Mackay, J., Elder, P.A., Porteous, D.J., Steel, C.M., Hawkins, R.A., Going, J.J., Chetty, U.
Partial deletion of chromosome 1 lp in breast cancer correlates with size of primary tumour
and oestrogen receptor level. Br. J. Cancer, 58: 710-714 (1988a)
Mackay, J., Steel, C.M., Elder, P.A., Forrest, A.P.M., Evans, H.J. Allele loss on short arm of
chromosome 17 in breast cancer. Lancet 2: 1384-1385 (1988b)
MacMahon, B., Cole, P., Lin, T.M. Age at first birth and cancer of the breast. A summary of
an international study. Bull WHO, 43: 209 (1970)
Madar, S., Kumar, V., de Vemeuil, H., Chambon, P. Three amino acids of the oestrogen
receptor are essential to its ability to distinguish an oestrogen from a gluticorticoid
responsive element. Nature, 338: 271-274 (1989)
Magliaccio, A., Rotondi, A., Auricchio, F. Estradiol receptor: phosphorylation on tyrosine
in uterus and interaction with antiphosphotyrosine antibody. EMBOJ., 5: 2867-2872 (1986)
Malakin, D., Li. F.P., Strong, L.C., Fraumeni, J.F. Jr., Nelson, C.E., Kim. D.H., Kassel, J.,
Gryka, M.A., Bischoff, F.Z., Tainsky, M.A., Friend, S.H. Germ line p53 mutations in a
252
familial syndrome of breast cancer, sarcomas, and other neoplasms. Science, 250: 16841689(1990)
Malkhosyan, S., Rampino, N., Yamamoto, H., and Perucho, M. Frameshift mutator
mutations. Nature, 382: 499-500 (1996)
Maminta, M.L., Molteni, A., Rosen, S.T. Stable expression of the estrogen receptor in HeLa
cells by infection: Effect o f estrogen on cell proliferation and c-myc expression. M ol Cell
Endocrinol., 78: 61-69 (1991)
Maniatis, T., Fritsch, E.F., Sambrook, J. Molecular Cloning: A Laboratory Manual. New
York: Cold Spring Harbour, book 3, E37-38, 1989
Manley, J.L. Polyadenylation o f mRNA precursors. Biochem. Biophys. Acta. 950: 1-12
(1988)
Manning, D.L., Robertson, J.F.R., Ellis, I.O., Eslston, C.W., McClelland, R.A., Gee,
J.M.W., Jones, R.J., Green, C.D., Cannon, P., Blarney, R.W., Nicholson, R.I. Oestrogenregulated genes in breast cancer: association of pLIVl with lymph node involvement.
European J. Cancer, 30A: 675-678 (1994)
Markowitz, S. D., Wang, J., Myeroff, L., Parsons, R., Sun, L., Lutterbaugh, J., Fan, R. S.,
Zborowska, E., Kinzler, K. W., Vogelstein, B., Brattain, M., and Willson, J. K. V.
Inactivation o f the type II TGF-p receptor in colon cancer cells with microsatellite
instability. Science, 268: 1336-1338 (1995)
*Marra, G., and Boland, C. R. Hereditary nonpolyposis colorectal cancer (HNPCC): the
syndrome, the genes, and historical perspective. J. Natl. Cancer Inst., 87: 1114-1125
(1995?)
Mars, W.M., Saunders, G.F. Chromosomal abnormalities in human breast cancer. Cancer
Metastasis Rev., 9, 35-43 (1990)
Marsigilante, S., Muscella, A., Resta, L., Storelli, C. Human larynx expresses isoforms of
the oestrogen receptor. Cancer Letters, 99: 191-196 (1996)
Martin, P.M., Rolland, P.H., Jacquemier, J. Multiple steroid receptors in human breast
cancer. II. Estrogen and progestin receptors in 672 primary tumours. Cancer Chemotherapy
Pharmacol. 2: 107 (1979)
Martinez, E., Givel, F., Wahli, W. The oestrogen-responsive element as an inducible
enhancer: DNA sequence requirements and conversion to a gluticorticoid-responsive
element. EMBOJ., 6 : 3719-3727 (1987)
Martinez, E., Wahli, W. Co-operative binding of oestrogen receptor to imperfect oestrogen
responsive DNA elements correlates with their synergistic hormone-dependant enhancer
activity. EMBOJ., 8 : 3781-3791 (1989)
Marx, D., Schauer, A., Reiche, C. C-erbB-2 expression and correlation to other biological
parameters o f breast cancer. J. Cancer Res. Clin. Oncol., 116: 15-20(1990)
breast cancer cells. Cancer Res., 46: 6034-6040 (1986)
253
Masiakowski, P., Breathnach, R., Bloch, Gannon, T., Krust, A., Chambon, P. Cloninf of
cDNA sequences o f hormone regulated genes from the MCF-7 human breast cancer cell
line. EM BO J, 10: 7895-7903 (1982)
Maynard, P.V., Blarney, R.W., Elston, C.W. Oestrogen receptor assay in primary breast
cancer and early recurrence o f the disease. Cancer Res., 38: 4292-4295 (1978)
Mazars, R., Spinardi, L., Ben Cheikh, M., Simony-Lafontaine, J., Jeanteur, P., Theillet, C.
p53 mutations occur in aggressive breast cancer. Cancer Res., 52: 3918-3923 (1992)
McGuire, W.L., Vollmer, E.P., Carbone, P.P. (eds). Estrogen receptor in human breast
cancer. New York, Raven Press. (1975)
McGuire, W.L., Clark, G,M. Prognostic factors for recurrence and survival in axillary node­
negative breast cancer. J. Steroid Biochem., 34, 145-148 (1989)
McGuire, W.L., Chanmess, G.C., Fuqua, S.A.W. Estrogen receptor variants in clinical
breast cancer. Mol. Endocrinol., 5: 1571-1577 (1991)
McGuire, W.L., Chamness, G.C., Fuqua, S.A.W. Abnormal oestrogen receptor in clinical
breast cancer. J. Steroid Biochem., 43: 243-247 (1992)
McIntosh, G.G., Anderson, J.J., Milton, I., Steward, M., Parr, A.H., Thomas, M.D., Henry,
J.A., Angus, B., Lennard, T.W.J., Home, C.H.W. Determination of the prognostic value of
cyclin D1 overexpression in breast cancer. Oncogene, 11: 885-891 (1995)
McLauchlan, J., Gaffney, D., Whitton, J.L., Clements, J.B. The consensus sequence
YGTGTTYY located downstream from the AATAAA signal is required for efficient
formation o f mRNA 3’ termini. Nucleic Acids Res., 13: 1347p-1368 (1985)
Menasce, L.P., White, G.R.M., Harrison, C.J., Boyle, J.M. Localization of the estrogen
receptor locus (ESR) to chromosome 6q25.1 by FISH and simple post-FISH banding
technique. Genomics, 17: 263-265 (1993)
Menasce, L.P., Orphanous, V., Santibanez-Koref, M., Boyle, J.M., Harrison, C.J. Common
region o f deletion on the long arm of chromosome 6 in non-Hodgkin’s lymphoma and acute
lymphoblastoic leukaemia. Genes Chromosomes and Cancer, 10, 286-288 (1994)
Mendel, D.B., Bodwell, J.E., Gametchu, B., Harrison, R.W. Munck, A. Molybdatestabilized nonactivated glucocorticoid-receptor complexes contain a 90-kDa nonsteroidbinding phosphoprotein that is lost in activation. J. Biol. Chem., 261: 3758-3763 (1986)
Merkel, D.E., McGuire, W.L. Ploidy, proliferative activity and prognosis. DNA flow
cytometry o f solid tumours. Cancer, 65: 1194 (1990)
Merlo, G.R., Siddiqui, J., Cropp, C.S., Liscia, D.S., Lidereau, R., Callahan, R., Kufe, D.W.
Frequent alteration o f the DF3 tumour-associated antigen gene in primary human breast
carcinomas. Cancer Res., 49: 6966-6971 (1989)
254
Merlo, G.R., Bemardi, A., Diella, F., Venesio, T., Cappa, A.P.M., Callahan, R., Liscia, D.S.
Primary human breast carcinomas mutations in exons 5 and 6 of the p53 gene are associated
with high S-phase index. Int. J. Cancer, 53: 531-535 (1993)
Merlo, A., Gabrielson, E., Mabry, M., Vollmer, R., Baylin, S.B., Sidransky, D. Homozygous
deletions on chromosome 9p and loss of heterozygosity on 9q, 6 p, and 6 q in primary human
small cell lung cancer. Cancer Res., 54, 2322-2326 (1994)
Merlo, A., Herman, J.G., Mao, L., Lee, D.J., Gabrielson, E., Burger, P.C., Baylin, S.B.,
Disransky, D. 5’ CpG island methylation in associated with transcriptional silencing of the
tumour suppressor pl6/CD K N 2/M TSl in human cancers. Nature Med., 1, 686-692 (1995)
Meshinchi, S., Sanchez, E.R., Martell, K.J., Pratt, W,B. Elimination and reconstitution of
the requirement for hormone in promoting temperature-dependant transformation of
cytosolic glucocorticoid receptors to the DNA-binding state. J. Biol. Chem., 265: 48634870(1990)
Mester, J., Beaulieu, E.E. Dynamics o f estrogen-receptor distribution between the cytosol
and nuclear fraction o f immature rat uterus after oestradiol administration. Bioch. J., 146:
617(1975)
Metzger, D., White, J.H., Chambon, P. The human oestrogen receptor functions in yeast.
Nature, 334: 31-36(1988)
Meyer, J.S., Rao, B.R., Stevens, S.C., White, W.L. Low incidence of estrogen receptor in
breast carcinomas with rapid rates o f cellular replication. Cancer 40: 2290 (1977)
Meyer, M.-E., Gronmeyer, H., Turcotte, B., Bocquel, M.-T., Tasset, D., Chambon, P.
Steroid hormone receptors compete for factors that mediate their enhancer function. Cell,
57:433-442(1989)
Meyn, M. S. Ataxia-Telangiectasia and cellular responses to DNA damage. Cancer Res., 55:
5991-6001 (1995)
Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P.A., Harshman, K., Tavtigian, S., Liu,
Q.Y., Cochran, C., Bennet, L.M., Ding, W., Bell, R., Rosenthal, J., Hussey, C., Tran, T.,
McClure, M., Frye, C., Hattier, T., Phelps, R., Haugenstrano, A., Katcher, H., Yakumo, K.,
Gholami, Z., Shaffer, D., Stone, S., Bayer, S., Wray, C., Bogden, R., Dayananth, P., Ward,
J., Tonin, P., Narod, S., Bristow, P.K., Norris, F.H., Helverring, L., Morrison, P., Reosteck,
P., Lai, M., Barrett, J.C., Lewis, C., Neuhauseen, S., Cannonal-bright, L./, Goldgar, D.,
Wiseman, R., Kamb, A., Skolinick, M.H. A strong candidate for the breast and ovarian
cancer susceptibility gene BRAC1. Science, 266: 66-71 (1994)
Miksicek, R., Lei, Y., Wang. Y. Exon skipping gives rise to alternatively spliced forms of
the estrogen receptor in breast tumour cells. Br. Cancer Res. Treat., 26: 163-174 (1993)
Miller, J., McLauchlan, A.D., Klug, A. Repetitive zinc-binding domains in the protein
transcription factor III A from Xenopus oocytes. EMBOJ., 4: 1609-1614 (1985)
Miller D.R., Rosenberg, I., Kaufmann, D.W. Breast cancer before the age o f 45 and oral
contraceptive use: New findings. Am. J. Epidemiol., 126: 269-280 (1989)
255
Miller, A.B. Incidence and demographics: Radiation risk. In Harris, J.R., Heilman, S.,
Henderson, I.C., Kinne, D.W. Breast Diseases. 2nd ed. Philadelphia: J.B. Lippincott, ppl 19126(1992)
Millikin, D., Meeses, E., Vogelstein, B., Witowski, C., Trent, J. Loss of heterozygosity for
loci on the long arm o f chromosome 6 in human malignant melanoma. Cancer Res., 51,
5 4 4 9 - 5 4 5 3 (1991)
Misrahi, M., Loosfelt, H., Atger, G., Meriel, C., Zerah, V., Dessen. P., Milgrom, E.
Organisation o f the entire rabbit progesterone receptor mRNA and of the promoter and 5’
flanking region o f the gene. Nucleic Acids Res., 16: 5459-5473 (1988)
Mitchell, P.J., Tjian, R. Transcriptional regulation in mammalian cells by sequence-specific
DNA binding proteins. Science, 245: 371-378 (1989)
Modrich, P. Mechanisms and biological effects of mismatch repair. Ann. Rev. Genet., 25:
229-253 (1991)
Moll, R., Mitze, M., Frixen, V.H., Birchmeier, W. Differential loss of E-cadherin in
infiltrating ductal and lobular breast carcinoma. Am. J. Pathol., 143: 1731-1742 (1993)
Montano, M.M., Mller, V., Trobaugh, A., Katzenellenbogen, B.S. The carboxy-terminal
domain of the human estrogen receptor: role in the transcriptional activity of the receptor
and the effectiveness o f antiestrogens as estrogen anatagonists. Mol. Endocinol., 9: 814-825
(1995)
Morin, P. J., Vogelstein, B., and Kinzler, K. W. Apoptosis and APC in colorectal
tumourigenesis. Proc. Natl. Acad. Sci. USA., 93: 7950-7954 (1996)
Morin, P. J., Sparks, A. B., Korinek, V., Barker, N., Clevers, H., Vogelstein, B., and
Kinzler, K. W. Activation o f p-catenin-Tcf signalling in colon cancer by mutations in pcatenin or APC. Science, 275: 1787-1790 (1997)
Morita, R., Saito, S., Ishikawa, J., Ogawa, O., Yoshida, O., Yamakawa, K., Nakamura, Y.
Common regions o f deletion on chromosomes 5q, 6 q, and lOq in renal cell carcinoma.
Cancer Res., 51, 5817-5820 (1991)
Mosselman, S., Polman, J., Dijkema, R. ER-beta -identification and characterization of a
novel human estrogen receptor. FEES Letters, 392: 49-53 (1996)
Munn, K.E., Walker, R.A., Varley, J.M. Frequent alterations of chromosome 1 in ductal
carcinoma in situ o f the breast. Oncogene, 10, 1653-1657 (1995)
Munn, K.E., Walker, R.A., Menasce, L., and Varley, J.M. Mutation of the TP53 gene and
allelic imbalance at chromosome 17pl 3 in ductal carcinoma in situ. Br. J. Cancer, 74: 15781585 (1996a)
Munn, K.E., Walker, R.A., Menasce, L., and Varley, J.M. Allelic imbalance in the region of
the BRCA1 gene in ductal carcinoma in situ of the breast. Br. J. Cancer, 73: 636-639
(1996b)
256
Murdoch, F.E., Meier, D.A., Furlow, J.D., Grunwald, K., Gorski, J. Estrogen-receptor
binding to a DNA response element in vitro is not dependant upon estradiol. Biochemistry.
9: 8377-8385 (1990)
Murphy, L.C., Wang, M., Coutts, A., Dotzlaw, H. Novel mutations in the estrogen receptor
mRNA in human breast cancers. J. Clin. Endorinol. Metab., 81: 1420-1427 (1996)
Myeroff, L. L., Parsons, R., Kim, S-J., Hedrick, L., Cho, R. K., Orth, K., Mathis, M.,
Kinzler, K. W., Lutterbaugh, J., Park, K., Bang, Y-J., Lee, H, Y., Park, J-G., Lynch, H. T.,
Roberts, A, B., Vogelstein, B., and Markowitz, S.D. A transforming growth factor receptor
P type II gene common in colon and gastric but rare in endometrial cancers with
microsatellite instability. Cancer Res., 55: 5545-5547 (1995)
Myers, R.M., Larin, Z., Maniatis, T. Detection of single base substitutions by ribonuclease
cleavage at mismatches in RNA:DNA duplexes. Science, 230: 1242-1246 (1985)
Myers, T.W., Gelfand, D.H. Reverse transcription and DNA amplification by a Thermus
thermophillus DNA polymerase. Biochem., 30: 7661- 7666 (1991)
Nagai, H., Negrini, M., Carter, S.L., Gillum, D.R., Rosenberg, A.L., Schwartz, G.F., Croce,
C.M. Detection and cloning o f loss o f heterozygosity at chromosome lp in breast cancer.
Cancer Res., 55: 1752-1757(1995)
Natali, P.G., Nicotra, M.R., Botti, C., Mottolese, M., Bigothi, A., Segatto, O. Changes in
expression o f a 6 p 4 integrin heterodimer in primary and metastasic breast cancer. Br. J.
Cancer, 6 6 : 318-322 (1992)
Nayar, R., Zhuang, Z., Merino, M. J., and Silverberg, S. G. Loss of Heterozygosity on
chromosome 1 l q l 3 in lobular lesions of the breast using tissue microdissection and the
polymerase chain reaction. Hum. Pathol, 28: 277-282 (1996)
Negrini, M., Sabbioni, S., Possati, L., Corallini, A., Barbanti-Brodano, G., Croce, C.M.
Suppression o f tumourogenicity o f breast cancer cells by microcell-mediated chromosome
transfer studies on chromosome 6 and 11. Cancer Res., 54, 1331-1336 (1994)
Newton, C.J., Burie, R., Trapp, T., Brockmeir, S., Pagotto, U., Stalla, G.K. The unliganded
estrogen receptor (ER) transduces growth factor signals. J. Steroid Biochem. Mol. Biol., 48:
481-486(1994)
Nicholson, R.I., Bouzubar, K.J., Walker, K.J., McMlelland, R., Dixon, A.R., Robertson,
J.F.R., Ellis, I.O., Blarney, R.W. Hormone sensitivity in breast cancer: Influence of
heterogeneity o f oestrogen receptor expression and cell proliferation. Eur. J. Cancer, 27:
908-913 (1991)
Nicholson, R.I., McClelland, R.A., Finlay, P., Elston, C.L., Gullick, W.J., Dixon, A.R.,
Robertson, J.F.R., Ellis, I.O., Blarney, R.W. Relationship between EGF-R, c-erb-B2 protein
expression and Ki67 immunostaining in breast cancer and hormone sensitivity. Eur. J.
Cancer., 29A: 1018-1023 (1993)
257
Nicoladies, N.C., Papadopoulus, N., Liu, B. Wei, Y.F., Carter, K.C., Ruben, S.M., Rosen,
C.A., Haseltine, W.A., Fleshchmann, R.D., Fraser, C.M. Mutations of two PMS
homologues in HNPCC. Nature, 371, 75-80 (1994)
Nicolaides, N. C., Carter, K. C., Shell, B. K., Papadopoulos, N., Vogelstein, B., and Kinzler,
K. V. Genomic organisation o f the human PMS2 gene family. Genomics, 80: 195-206
(1995)
Norris, J., Fan, D., Aleman, C., Marks, J.R., Futreal, P.A., Wisemann, R.W., Iglehart, J. D.,
Deininger, P.L., McDonnell, D.P. Identification of a new subclass of Alu repeats which can
function as estrogen receptor-dependant transcriptional enhancers. J. Biol. Chem., 270:
22777-22782(1995)
Noviello, C., Courjal, F., Theillet, C. Loss of heterozygosity on the long arm of
chromosome 6 in breast cancer: possibly four regions of deletion. Clin. Cancer Res., 2:
1601-1606(1996)
Nowell, P. The clonal evolution o f tumour populations. Science, 194: 23-28 (1976)
Nunez-A-M., Berry, M., Imler, J-L., Chambon, P. The 5’ flanking region of the pS2 gene
contains a complex enhancer region responsive to oestrogens, epidermal growth factor, a
tumour promoter (TPA), the c-H-ras oncoprotein and the c-jun protein. EMBO J., 8 : 823829(1989)
Oka, H., Shiozaki, H., Kobayashi, K., Inove, M., Tahara, H., Kobayashi, T., Takatsuka, Y.,
Matsuyoshi, N., Himo, S., Takeichi, M., Mori, T. Expression of E-cadherin cell adhesion
molecules in human breast cancer tissues and its relationship to metastasis. Cancer Res., 53:
1696-1701 (1993)
O ’Malley, B.W., Tsai, M.-J. Overview of the steroid receptor superfamily of gene
regulatory proteins. In Steroid Hormone Action. Edited by M.G. Parker, IRL Press (1993)
Oltavi, Z. N., Milliman, C. L., and Korsmeyer, S. J. Bcl-2 heterodimerises with in vivo with
a conserved homologue, bax, that accelerates programmed cell death. Cell, 74: 609-619
(1993)
Orita, M., Iwahana, H., Kanazawa, H., Hayashi., Seiya, T. Detection of polymorphisms of
human DNA by gel electrophoresis as single stranded conformational polymorphisms. Proc.
Nat. Acad. Sci. USA, 8 6 : 2766-2770 (1989a)
Orita, M., Suzuki, Y., Seika, T., Hayashi, K. Rapid and sensitive detection of point
mutations and DNA polymorphisms using the polymerase chain reaction. Genomics, 5: 874879 (1989b)
Orphanos, V., McGowen, G., Hey, Y., Boyle, J.M., Santibanez-Koref, M. Proximal 6 q, a
region showing allele loss in primary breast cancer. Br. J. Cancer, 71: 290-293 (1995)
Osborne, C.K., Arteaga, C.L. Autocrine and paracrine growth regulation o f breast cancer:
Clinical implications. Breast Cancer Res. Treat., 15: 3-11 (1990)
258
Osbome, R.J., Merlo, G.R., Mitsudomi, T., Venesio, T., Liscia, D.S., Cappa, A.P.M., Chiba,
I., Takahashi, T., Nau, M.M., Callahan, R., Minna, J.D. Mutations in the p53 gene in
primary human breast cancers. Cancer Res., 51: 6194-6198 (1991)
Ottaviano, Y.L., Issa, J.P.. Pari, F.F., Smith, H.S., Baylin, S.B., Davidson, N.E. Methylation
of the oestrogen receptor gene CpG island marks loss of estrogen receptor expression in
human breast cancer cells. Cancer Res., 54: 2552-2555 (1994)
Ouyang, H., Shiwaku, H. O., Hagiwara, H., Miura, K., Abe, T., Kato, Y., Ohtani, H., Shiiba,
K., Souza, R. F., Meltzer, S. J., and Horrii, A. The Insulin-like Growth Factor II Receptor
gene is mutated in genetically unstable cancers of the endometrium, stomach, and
colorectum. Cancer Res., 57: 1851-1854(1997)
Paech, K., Webb, P., Kuiper, G.G.J.M., Nilsson, S., Gustafsson, J-A., Kushner, P.J.,
Scanlan, T.S. Differential ligand activation of estrogen receptors ER a and ERp at API sites.
Science, 277: 1508-1510(1997)
Page, D. L., and Anderson, T. J. (Eds). Diagnostic histopathology of the breast. Churchill
Livingstone. Edinburgh, (1987)
Page, D. L., Dupont, W. D., Rogers, L. W., Jensen, R. A, and Schuyler, P. A. Continued
local recurrence o f carcinoma 15-25 years after a diagnosis of low grade ductal carcinoma in
situ of the breast treated only by biopsy. Cancer, 76: 1197-1200 (1995)
Paik, S. Expression o f IGF-I and IGF-II mRNA in breast tissue. Breast Cancer Res. Treat.,
22,31-38(1992)
Pang, Q., Alani, E., Kolodner, R.D., Liskay, R.M. MLH1, PMS1 and MSH2 interactions
during the initiation o f DNA mismatch repair in yeast. Science, 265: 1091-1093 (1994)
Parker, M.G. Mortyn Jones Memorial Lecture: Structure and function of the oestrogen
receptor. J. Neuroendocrinology, 5: 223-228 (1993)
Patel, U., Grundfest-Broniatowski, S., Gupta, M., Baneijee, S. Microsatellite instabilities in
five chromosomes in primary breast tumours. Oncogene, 9: 3695-3700 (1994)
Patel, U., Grundfest-Broniatowski, S., Gupta, M., and Banerjee, S. Microsatellite
instabilities at five chromosomes in primary breast tumours. Oncogene, 9: 3695-3700
(1995)
Paul, C., Skegg, D.C.G., Spears, G.F.S., Kaldor, J.M. Oral contraceptives and breast cancer:
a national study. Br. Med. J., 293: 723-726 (1986)
Paulson, T. G., Wright, F. A., Parker, B. A., Russack, V., and Wahl, G. M. Microsatellite
instability correlates with reduced survival and poor disease prognosis in breast cancer.
Cancer Res., 56: 1641-1644(1996)
Pearlman, W.H., DeHertogh, R., Laumas, K.R., Pearlman, M.R.J., Metabolism and tissue
uptake o f oestrogen in women with advanced carcinoma of the breast. J. Clin. Endocrinol.,
29: 707(1969)
259
Peifer, M. p-catenin as oncogene: the smoking gun. Science, 275: 1752-1753 (1997)
Peltomaki, P., Lothe, R. A., Aaltonen, L. A., Pylkkanen, L., Nystrom-Lahti, M. N., Seruca,
R., David, L., Holm, R., Ryberg, D., Haugen, A., Brogger, A., Borresen, A-L., and de la
Chapelle, A. Microsatellite instability is associated with tumours that characterize the
hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res., 53: 5853-5855
(1993)
Pertschuk, L.P., Feldman, J.G., Eisenberg, K.B., Carter, A.C., Thelmo, W.L., Cruz, W.P.,
Thorpe, S.M., Christensen, J., Rasmussen, B.B., Rose, C., Greene, G.L.
Immunocytochemical detection o f progesterone receptor in breast cancer with monoclonal
antibody - Relation to biochemical assay, disease free survival and clinical endocrine
response. Cancer, 62: 342-349 (1988)
Perucho, M., Peiado, M. A., Ionov, Y., Casares, S., Malkhosyan, S., and Stanbridge, E.
Defects in replication fidelity o f simple repeated sequences reveals new mutator mechanism
for oncogenesis. Cold Spring Harbor Sympos. Quant. Biol. 59: 339-348, (1994)
Perucho, M. Microsatellite instability: The mutator that mutates the other mutator. Nature
Med., 2: 630-631 (1996)
Petrangelli, E., Lubrano, C., Ortolani, F., Ravenna, L., Vacca, A., Sciacchitano, S., Frati, L.,
Gulino, A. Estrogen receptors: new perspectives in breast cancer management. J. Steroid
Biochem. Mol. Biol., 49: 327-331 (1994)
Petrangelli, E., Lubrano, C., Ravenna, L., Vacca, A., Cardillo, M.R., Salvatori, L., Sciarra,
F., Frati, L., Gulino, A. Gene methylation of oestrogen and epidermal growth factor
receptors in neoplastic and perineoplastic breast tissues. Br. J. Cancer, 72: 973-975 (1995)
Pfeffer, U., Fearotta, E., Catagnetta, L., Vidali, G. Estrogen receptor variant messenger
RNA lacking exon 4 in estrogen responsive human breast cancer cell lines. Cancer Res., 53:
741-743 (1993)
Pfeffer, U., Fecarotta, E., Vidali, G. Coexpression of multiple estrogen receptor variant
messenger RNAs in normal and neoplastic breast tissues and in MCF-7 cells. Cancer Res.,
55:2158-2165 (1995)
Pfeffer, U., Fecarotta, E., Arena, G., Forlani, A., Vidali, G. Alternative splicing of the
estrogen receptor primary transcripts normally occur in estrogen receptor positive tissues
and cell lines. J. Steroid Biochem. Molec. Biol., 56: 99-105 (1996)
Phillips, A., Chalbos, D., Rochefort, H. Estradiol increases and antiestrogens antagonise the
growth factor-induced activator protein-1 activity in MCF-7 breast cancer cells without
affecting c-fos and c-jun synthesis. J. Biol. Chem., 268: 14103-14108 (1993)
Picard, D., Kumar, V., Chambon, P., Yamamoto, K.R. Signal transduction by steroid
hormones: nuclear localisation is differentially regulated in estrogen and glucocorticoid
receptors. Cell. Regul., 1: 291-299 (1990)
Pietras, R.J., Abboleda, J., Reese, D.M., Wongvipat, N., Pegram, M.D., Ramos, L., Gorman,
C.M., Parker, M.G., Sliwowski, M.K., Slamon, D.J. HER-2 tyrosine kinase pathways targets
260
estrogen receptor and promotes hormone-independent growth ion human breast cancer cells.
Oncogene, 10: 2435-2446(1995)
Pignatelli, M., Hanby, A.N., Stamp, Q.W.H. Low expression of
a 2, a 3 subunits of VLA
integrins in malignant mammary tumours. J. Pathol., 165: 25-32 (1994)
Pike, M.C. Henderson, B.E., Krailo, M.D., Duke, A., Roy, S. Breast cancer and oral
contraceptives: reply to critics. Lancet, 1414 (1983)
Pink, J.J., Jiang, S.Y., Fritsch, M., Jordan, V.C. A unique MCF-7 human breast cancer cell
line expressing an 80-kD estrogen receptor. Proc. Am. Assoc. Cancer. Res., 35: 276 (abst
1651)(1994)
Pink, J.J., Wu, S-Q., Wolf, D.M., Billmoria, M.M., Jordan, C.V. A novel 80 kDa human
estrogen receptor containing a duplication of exons 6 and 7. Nucleic Acids Res., 24: 962-969
(1996)
Pink, J.J., Fritsch, M., Billimoria, M.M., Assikis, V.J., Jordan, V.C. Cloning and
characterization o f a 77-kDa oestrogen receptor isolated from a human breast cancer cell
line. Br. J. Cancer, 75: 17-27 (1997)
Piva, R., Gambari, R., Zorzato, F., Kumar, L., del Senno, L. Analysis of upstream sequences
o f the human estrogen receptor gene. Biochem. Biophys. Res. Commun., 183: 996-1002
(1992)
Piva, R., Bianchi, N., Aguiari, G.L., Gamvari, R., del Senno, L. Sequencing of an RNA
transcript o f the human estrogen receptor gene: evidence for a new transcriptional event. J.
Steroid Biochem. Mol. Biol., 46: 531-538 (1993)
Poller, DAN., Hutchings, C.E., Galea, M., Bell, J.A., Nicholson, R.A., Elston, C.W.,
Blarney, R.W., Ellis, I.O. p53 protein expression in human breast carcinoma: relationship to
expression o f epidermal growth factor receptor, c-erb-B2 protein overexpression, and
oestrogen receptor. Br. J. Cancer, 6 6 : 583-588 (1992)
Polyak, K. Negative regulation o f cell growth by TGFp. Biochem. Biophys. Acta, 1242: 185199(1996)
Polymeropoulos, M.H., Rath, D.S., Xiao, H., Merril, C.R. Trinucleotide repeat
polymorphism at the human transcription factor IID gene. Nucl. Acids Res., 19,4307 (1991)
Ponglikitmongkol, M., Green, S., Chambon, P. Genomic Organisation of the human
oestrogen receptor gene. E M B O J, 7: 3385-3388 (1988)
Poulson, R., Wright, N.A. Trefoil peptides: a newly recognised family of epithelial mucinassociated molecules. Am. J. Physiol., 265: 205-213 (1993)
Power, R.F., Mani, S,K., Codina, J., Conneely, O.M., O’Malley, B.W. Dopamininergic and
ligand-independent activation of steroid hormone receptors. Science, 254: 1636-1638 (1991)
Pratt, W.B. Transformation of gluticorticoid and progesterone receptors to the DNA-binding
state. J. Biol. Chem., 35: 51-68 (1987)
261
Pratt, W.B. Interaction of hsp90 with steroid receptors - organising some diverse
observations and presenting the newest concepts. Mol. Cell. Endocrinol., 74: C69-C76
(1990)
Pratt, W.B. Role o f heat-shock proteins in steroid receptor function. In Steroid Hormone
Action. IRL Press. Edited by Parker, M.G. p64-65 (1993)
Press, M.F., Piek, M.C., Chazin, V.R., Hung, G., Udove, J.J.A., Markowitcz, M., Danyluk,
J., Godolphin, W., Sliwkowski, M., Akita, R., Paterson, M.C., Salmon, D.J. Her-2tneu
expression in node negative breast cancer: direct tissue quantitation by computerised image
analysis and association o f over expression with increased risk of recurrent disease. Cancer
Res., 53: 4960-4970 (1993)
Press, M.F., Hung, G., Godolpin, W., Salmon, D.J. Sensitivity of HERJneu antibodies in
archival tissue samples: potential source of error in immuno-histochemical studies of
oncogene expression. Cancer Res., 54: 2771-2777 (1994)
Prolla, T.A., Christie, D.M., Liskay, R.M. Dual requirement in yeast DNA mismatch repair
for MLH1 and PMS1, two homologues of the bacterial mutL gene. Mol. Cell Biol., 14: 407415(1994)
Prooijen-Knegt, A.C., Raap, A.K., Van der Burg, M.J.M. Spreading and staining of juman
metaphase chromosomes on aminoalkylsilane-treated glass slides. Hitochem. J., 14: 333344(1982)
Proudfoot, N. Poly(A) signals. Cell, 64: 671-674 (1991)
Purchio, A.F., Cooper, J.A., Brunner, A.M., Lioubin, M.N., Gentry, L.E., Kovacina, K.S.,
Roth, R.A., Marquardi, H. Identification of mannose 6 -phosphate in two asparginine-linked
sugar chains o f recombinant transforming growth factor-pl precursor. J. Biol. Chem., 263:
14211-14215 (1988)
Radford, D. M., Fair, K. L., Phillips, N. J., Ritter, J. H., Steinbrueck, T., Holt, M. S., and
Donis-Keller, H. Allelotyping o f ductal carcinoma in situ of the breast: deletion of loci on
8 p, 13q, 16q, 17p, and 17q. Cancer Res., 55: 3399-3405 (1995)
Rae, D., Parker, M.G. Effects o f an exon 5 variant of the estrogen receptor in MCF-7 Breast
Cancer cells. Cancer Res., 56: 1556-1563 (1996)
Rajakariar, R. The biological nature of early invasive breast cancer. Bachelor of Science
(Medical Sciences) thesis, 1993
Rajakariar, R., Walker, R.A. Pathological and biological features of mammographically
detected invasive breast carcinomas. Br. J. Cancer, 71: 150-154 (1995)
Ramachandra, S., Machin, L., Ashley, S., Monaghan, P., Gusterson, B-A.
Immunohistochemical distribution of c-erb-B2 in in situ breast carcinoma - a detailed
morphological analysis. J. Pathol., 161: 7-14 (1990)
262
Rampino, N., Yamamoto, H., Ionov, Y., Li, Y., Sawai, H., Reed, J. C., and Perucho, M.
Somatic mutations in the BAX gene in colon cancers of the microsatellite mutator
phenotype. Science, 275: 967-969 (1997)
Rasbridge, S.A., Gillet, C.E., Sampson, S.A., Walsh, F.S., Millis, R.R. Epithelial (E-) and
placental (P-) cadherin cell adhesion molecules expression in breast carcinoma. J. Pathol.,
169: 245-250(1993)
Raycroft, L., Wu, H., Lozano, G. Transcriptional activation by wild-type but not
transforming mutants o f the p53 antioncogene. Science, 249: 1049-1051 (1990)
Raymond, W.A., Leong, A.S.Y. The relationship between growth fractions and oestrogen
receptors in human breast carcinoma, as determined by immunohistochemical staining. J.
Pathol, 158: 203-211 (1989)
Reiner, G.C.A., Katzenellenbogen, B.S. Characterisation of the estrogen and progesterone
receptors and the associated regulation of growth and progesterone stimulation by estrogen
in MDA-MB-134 human breast cancer cells. Cancer Res., 46: 1224-1131 (1986)
Risinger, J. I., Barrett, C. J., Watson, P., Lynch, H. T., and Boyd, J. Molecular genetic
evidence o f the occurrence o f breast cancer as an integral tumour in patients with the
hereditary nonpolyposis colorectal carcinoma syndrome. Cancer, 77: 1836-1843 (1996)
Rochette-Egly, C., Lutz, Y.,Saunders, M., Scheuer, I., Gaub, M.-P., Chambon, P. J. Retinoic
acid receptor-gamma-specific immunodetection and phosphorylation. Cell Biol., 115: 535545 (1991)
Rochette-Egly, C., Gaub, M.-P., Lutz, Y., Ali, S., Scheuer, I., Chambon, P. Retanoic acid
receptor beta immunodetection and phosphorylation on tyrosine residues. Mol. Endocrinol,
6:2197-2209(1993)
Rochfort, H., Capony, F., Garcia, M. Cathepsin D in breast cancer: from molecular and
cellular biology to clinical applications. Cancer Cells, 2: 383-388 (1990)
Rochlitz, C.F., Scott, G.K., Dodson, J.M., Liu, E., Dolbaum, C., Smith, H.S., Benz, C.C.
Incidence of activating ras mutations associated with primary and metastatic breast cancer.
Cancer Res., 49: 357-360 (1989)
Rodabaugh, K.J., Blanchard, G., Welch, W.R., Bell, D.A., Berkowitz, R.S., Mok, S.C.
Detailed deletion mapping o f chromosome 6 q in borderline epithelial ovarian tumours.
Cancer Res., 55, 2169-2172 (1995)
Rodenhiser, D., Chakraborty, P., Andrews, J., Ainsworth, P., Mancini, D., and Singh, S.
Heterogeneous point mutations in the BRCA1 gene occur in high frequency at the site of
homonucleotide tracts, short tandem repeats and methylatable CpG/CpNpG motifs.
Oncogene, 12: 2623-2629, (1996)
Rodgers, C.S., Hill, S.M., Hulten, M.A. Cytogenetic analysis in human breast carcinoma.
1.9 cases in the diploid range investigated using direct preparations. Cancer Genet.
Cytogenet., 13: 95-119 (1984)
263
Rodgers, C.S., Hill, S.M., Hulten, M.A. Cytogenetic analysis in a case of cancer of the male
breast. Cancer Genet. Cytogenet., 15: 113-117 (1985)
Rodriguez, R., Carson, M.A., Weigel. N.L., O’Mally, B.W., Schrader, W.T. Hormoneinduced changes in the in vitro DNA binding activity of the chicken progesterone receptor.
Mol. Endocrinol, 3: 356-362 (1989)
Roodi, N., Kao, W.Y., Verrier, C.S., Dupont, W.D., Pari, F.F. Estrogen receptor gene
analysis in estrogen receptor-positive and -negative breast cancer. Proc. Am. Assoc. Cancer
Res., 35: 278 (abst 1659) (1994)
Roodi, N., Bailey, L.R., Kao, W-Y., Verrier, C.S., Yee, C.Y., Dupont, W.D., Pari, F.F.
Estrogen receptor gene analysis in estrogen receptor-positive and receptor-negative primary
breast cancer. J. Natl. Cancer Inst., 87: 446-451 (1995)
Rosen, P.P., Menendez-Botet, C.J., Nisselbaum, J.S. Pathological review of breast lesions
analysed for estrogen receptor protein. Cancer Res., 351: 3187 (1975)
Rosen, P. P. The pathological classification of human mammary carcinoma: past, present
and future. Ann. Clin. Lab. Sci., 9: 144-156 (1979)
Rosen, P.P., Braun, D., Kinne, D.E. The clinical significance of pre-invasive breast
carcinoma. Cancer, 46: 919-925 (1980)
Rosenberg, C. L., Larson, P. S., Romo, J. D., De Las Morenas, A., and Faller, D. V.
Microsatellite alterations indicating monoclonality in atypical hyperplasias associated with
breast cancer. Hum. Pathol., 28: 214-219 (1997)
Royal College o f Pathologists Working Group. NHS Breast Screening Program: Pathology
Reporting in Breast Cancer Screening. London: Royal College of Pathologists (1996)
Rubins, R.D., Hayward, J.L. Estrogen receptors and response to endocrine therapy and
cytotoxic chemotherapy in advanced breast cancer. Cancer 46: 2922-2924 (1980)
Rush, E. B., Calvano, J. E., VanZee, K. J., Zelenetz, A. D., and Borgen, P. I. Microsatellite
instability in breast cancer. Annals Surg. Oncol., 4\ 310-315 (1997)
Russo, J., Russo, I.H. Mammary tumourigenesis. Prog. Exp. Tumour Res., 33: 175-191
(1991)
Ruttimann, C., Cotoras, M., Zaldivar, J., Vicuna, R. DNA Polymerases from the extremely
thermophillic bacterium Thermus thermophillus HB-8 . E u r.Biochem., 149: 41-46 (1985)
Sabbah, M., Radanyl, C., Baulieu, E-E. The 90kDa heat-shock protein (hsp90) modulates
the binding of the oestrogen receptor to its cognate DNA. Biochem. J., 314: 205-213 (1996)
Sager, R. Tumour suppressor genes, the puzzle and the promise. Science, 246: 1406-1412
(1989)
264
Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T., Erlich, H.A., Amheim, N.
Enzymatic amplification of P-globin genomic sequences and restriction site analysis for
diagnosis o f sickle cell anaemia. Science, 230: 1350-1354 (1985)
Saiki, R.K., Gelfand, D.H., Stoffel, S., Sharf, S.J., Higuchi, R., Horn. G.T., Mullis, K.B.,
Erlich, H.A. Primer directed enzymatic amplification of DNA with a thermostable DNA
Polymerase. Science, 239: 487-491 (1988)
Saito, S., Okui, K., Tokino, T., Oshimura, M., Nakamura, Y. Isolation and mapping of
RFLP markers on human chromosome 6 . Am. J. Hum. Genet., 50: 65-70 (1992a)
68
Saito, S., Saito, H., Koi, S., Sagae, S., Kudo, R , Saito, J., Noda, K., Nakamura, Y. Fine-scale
deletion mapping o f the distal long arm of chromosome 6 in 70 human ovarian cancers.
Cancer Res., 52, 5815-5817 (1992b)
Saito, F., Yamamoto, T., Horikoshi, M., Ikeuchi. T. Direct mapping of the human TATA
box-binding protein (TBP) gene to 6q27 by fluorescence in situ hybridisation. Jpn. J. Hum.
Genet., 39, 421-425(1994)
Sanger, F., Niklen, S., Coulson, A.R. DNA Sequencing with chain terminating inhibitors.
Proc. Natl. Acad. Sci. USA, 74: 5463-5467 (1977)
Santibanez-Koref, M.F., Birch, J.M., Hartley, A.L., Norris Jones, P.H., Craft, A.H., Eden,
T., Cowther, D., Kelsey, A.M., Harris, M. p53 germline mutations in Li Fraumeni
syndrome. Lancet, 338: 1490-1491 (1991)
Sardat, C., Vida, M., Cobrinik, C., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A.
E2F-4 and E2F-5, two members o f the E2F family, are expressed in the early phases of the
cell cycle. Proc. Natl. Acad. Sci. USA., 92: 2403-2407 (1995)
Sattin, R.W., Rubin, G.L., Webster, L.A., Huezo, C.M., Wingo, P.A., Ory, H.W., Layde,
P.M. Family history o f breast cancer, JAMA, 253: 1908-1913 (1985)
Sato, T., Tanigami, A., Yamakawa, K., Akiyama, F., Nakamura, Y. Allelotype of breast
cancer: cumulative allele losses promote tumour progression in primary breast cancer.
Cancer Res., 50: 7184-7189 (1990)
Sato, T., Akiyama, F., Sakamotot, G., Kasumi, F., Nakamura, Y. Accumulation of genetic
alterations and progression of primary breast cancer. Cancer Res., 51: 5794-5799 (1991)
Sawadogo, M., Sentenac, A. RNA polymerase-B (II) and general transcription factors. Ann.
Rev. Biochem., 59: 711-754 (1990)
Schauer, M., Chalepakis, G., Willmann, T., Beato, M. Binding of hormone accelerates the
kinetics o f the gluticoid and progesterone receptor binding to DNA. Proc. Natl., Acad. Sci.
USA. 8 6 : 1123- 1127(1989)
Scherr, C.J. G1 phase progression: cycling on cue. Cell, 79:551-555 (1994)
Scherrer, L.C., Picard, D., Massa, E., Harmon, J.M., Simons, S.S., Yamamoto, K.R., Pratt,
W.B. Evidence that the hormone-binding domain of steroid-receptors confers hormonal 265
control o f chimeric proteins by determining their hormone regulated binding to heat shock
protein-90. Biochemistry, 32: 5381-5386 (1993)
Scherer, S. J., Welter, C., Ang, K. D., and Dooley, S. Specific in vitro binding of p53 to the
promoter region o f the human mismatch repair gene. Biochem. Biophys. Res. Comm., 221:
722-728 (1996)
Schliber, U., Sierra, F. Alternative promoters in developmental gene expression. Annu. Rev.
Genet., 21:237-257(1987)
Schott, D.R., Chang, J.N., Deng, G., Kurisu, W., Kuo, W-L., Gray, J., Smith, H.S. A
candidate tumour suppressor gene in human breast cancers. Cancer Res., 54: 1393-1396
(1994)
Schottenfeld, D., Nash, A,G., Robbins, G. Ten year results of the treatment of primary
operable breast carcinoma. Cancer, 38: 1001-1007 (1976)
Schwable, J.W.R., Neuhaus, D.D., Rhodes, D. Solution structure of the DNA-binding
domain o f the oestrogen receptor. Nature, 348: 458-461 (1990)
Schwable, J.W.R., Chapman, L., Finch, J.T., Rhodes, D. The crystal structure of the
estrogen receptor DNA-binding domain bound to DNA: how receptors discriminate between
their response elements. Cell, 75: 567-578 (1993)
Scully, R., Chen, J. J., Plug, A., Xiao, Y, H., Weaver, D., Feunteun, J., Ashley, T,
Livingston, D. M. Association o f BRCA1 with Rad51 in mitotic and meiotic cells. Cell, 8 8 :
265-275 (1997)
Seemayer, T.A, Cavenee, W.E. Biology of disease: Molecular mechanisms of oncogenesis.
Lab. Invest., 60: 585-599 (1989)
Segnitz, B., Gehring, U. Subunit structure of the unliganded human oestrogen receptor.
Proc. Natl. Acad. Sci. USA, 92: 2179-2183 (1995)
Senno, L.del., Aguiari, G.L., Piva, R. Dinucleotide repeat polymorphism in the human
estrogen receptor (ESR) gene. Hum. Mol. Genet., 1(5):354 (1992)
Seto, E., Usheva, A., Zambetti, G.P., Momand, J., Horikoshi, N., Weinmann, R., Levine,
A.J., Shenk, T. Wild-type p53 binds to the TATA-binding protein and represses
transcription. P.N.A.S., 89: 12028-12032(1992)
Shapiro, S., Venet, W., Strax, P., Venet, L., Roeser, R. Selection, follow-up and analysis in
the health insurance plan study: a randomised trial with breast cancer screening. Natl.
Cancer Inst., Monogr., 67: 65-74 (1985)
Sharan, S. K., Morimatsu, M., Albrecht, U., Lim, D. S., Regel, E., Dinh, C., Sands, A.,
Eichele, G., Hasty, P., and Bradley, A N. Embryonic lethality and radiation hypersensitivity
mediated by Rad51 in mice lacking BRAC2. Nature, 386: 772-773 (1997)
Shaw, G., Kamen, R. A conserved AU sequence from the 3’ untranslated region o f GM-CSF
mRNA mediates selective mRNA degradation. Cell, 46: 659-667 (1986)
266
Shaw, J. A., Walsh, T., Chappell, S. A., Carey, N., Johnson, K., and Walker, R. A.
Microsatellite instability in early sporadic breast cancer. Br. J. Cancer, 73: 1393-1397
(1996)
Shemshedini, L., Knauthe, R., Sasson-Corsi, P., Pomon, A., Gronemeyer, H. Cell-specific
inhibitory and stimulatory effects o f Fos and Jun transcription activation by nuclear
receptors. EMBO J, 10: 3839-3849 (1991)
Shi, S-R., Key, M.E., Kalra, K. Antigen retrieval in formalin-fixed paraffin-embedded
tissues: An enhancement method for immunohistochemical staining based on microwave
oven heating of tissue sections. J. Histochem. Cytochem., 39: 741-748 (1991)
Shi, Y.B., Yaoita, Y., Brown, D.D. Genomic organisation and alternative promoter usage of
the two thyroid hormone receptor betagenes in Xenopous-Laevis. J. Biol. Chem., 267: 733738(1992)
Sidranski, D., Tokins, T., Helzisouer, K., Rauch, G., Zehnbauer, B., Shelton, B.,
Pretiglacomo, L., Vogelstein, B., Davidson, N. Inherited p53 gene mutations in breast
cancer. Cancer Res., 52: 2984-2986 (1992)
Silvestrini, R., Daidone, M.G., Costa, A. Cell kinetics of solid tumours with time and its
clinical implication. Tumouri., 75: 367-370 (1989)
Simms, L. A., Zou, T-T., Young, J., Shi, Y-Q., Lei, J., Appel, R., Rhyu, M-G., Sugimara,
H., Cheneivix-Trench, G., Souza, R. F., Meltzer, S. J., and Legget, B. A. Apparent
protection from instability o f repeat sequences in cancer-related genes in replication error
gastrointestinal cancers. Oncogene, 14: 2613-2618 (1997)
Singer, C., Rasmussen, A., Smith, H.S., Lippmann, M.E., Lynch, H.T., Cullen, K.J.
Malignant breast epithelium selects for insulin-like growth factor II expression in breast
stroma: evidence o f paracrine function. Cancer Res., 55, 2448-2454 (1995)
Skipper, J.K., Young, L.J., Bergeron, J.M., Tetzlaff, M.T., Osborn, C.T., Crews, D.
Identification o f an isoform o f the estrogen-receptor messenger RNA lacking exon 4 and
present in the brain. Proc. Natl. Acad. Sci. USA, 90: 7172-7175 (1993)
Slamon, D.J., Clark, G.M., Wong, S.G., Levin, W.J., Ullrich, A., McGuire, W.L. Human
breast cancer: correlation o f relapse and survival with amplification of the HER-2/«ew
oncogene. Science, 235: 177-182 (1987)
Slamon, D.J., Godolphin, W., Jones, L.A., Holt, J.A., Wong, S.G., Keith, D.E., Levin, W.J.,
Stuart, S.G., Udove, J., Ullrich, A., Press, M.F. Studies of the HER-2/«ew proto-oncogene in
human breast and ovarian cancer. Science, 244: 707-712 (1989)
Sluyser, M., Mester, J. Oncogenes homologous to steroid receptors? Nature, 315: 546
(1985)
Sluyser, M. Mutations in the estrogen receptor gene. Human Mutation, 6 : 97-193 (1995)
Smale, S.T., Baltimore, D. The ‘initiator’ as a transcriptional control element. Cell, 57: 103113(1989)
267
Smale, S.T., Schmidt, M.C. Berk, A.J. Baltimore, D. Transcriptional activation by Spl as
directed through TATA or initiator: Specific requirement for mammalian transcription
factor IID. Proc. Natl. Acad. Sci. USA, 87: 4509-4513 (1990)
Smith, C.L., Conneely, O.M., O’Mally, B.W. Modulation of the ligand-independent
activation o f the human oestrogen receptor by hormone and antihormone. Proc. Natl. Acad
Sci. USA, 90: 6120-6124(1993)
Smith, C.L., Conneely, O.M., O’Mally, B.W. Oestrogen receptor activation in the absence
o f ligand. Biochemical Society Transactions, 23: 935-939 (1995)
Soule, H. D., Vazquez, J., Long, A., Albert, S., and Brennan, M. A human cell line from
pleural effusion derived from a breast carcinoma. J. Natl. Cancer. Inst., 51: 1409-1416
(1973)
Sourvinos, G., Kiaris, H., Tsikkinis, A., Vassilaros, S., and Spandidos, D. A. Microsatellite
instability and loss o f heterozygosity in primary breast tumours. Tumour Biol, 18: 157-166,
(1997)
Southern, E.M. Detection o f specific sequences among DNA fragments separated by gel
electrophoresis. J. Mol. Biol., 98: 503 (1975)
Souza, R. F., Appel, R., Yin, J., Wang, S., Smolinski, K. N., Abraham, J. M., Zou, T-T.,
Shi, Y-Q., Lei, J., Cottrell, J., Cymes, K., Biden, K., Simms, L., Legget, B, N., Lynch, P.
M., Frazier, M., Powell, S. M., Harpaz, N., Sugimara, H., Young, J., and Meltzer, S. J.
Microsatellite instability in the insulin-like growth factor II receptor gene in gastrointestinal
tumours. Nat. Genet., 14: 255-257 (1996)
Souza, R. F., Yin, J., Smolinski, K. N., Zou, T-T., Wang, S., Shi, Y-Q., Rhyu, M-G.,
Cottrell, J., Abraham, J. M., Biden, K., Simms, L., Legget, B, N., Ikeda, M., Tanino, M.,
Hanafusa, H., Tanaka, N., and Shimizu, K. ., Bova, G. S., Frank, T., Powell, S. M.,
Sugimara, H., Young, J., Harpaz, N., Shimizu, K., Matsubara, N.,and Meltzer, S. J.
Frequent mutation o f the E2F-4 cell cycle gene in primary human gastrointestinal tumours.
Cancer Res., 57: 2350-2353 (1997)
Strahle, U., Schmidt, A., Kelsey, G., Stewart, A.F., Cole, T.J. Scmid, W., Schutz, G. At least
3 promoters direct expression o f the mouse glucocorticoid receptor gene. Proc. Natl. Acad.
Sci. USA., 89: 6731-6735 (1992)
Stratton, M. R., Collins, N., Lakhhani, R., Sloane, J.P. Loss of heterozygosity in ductal
carcinoma in situ of the breast. J. Pathol., 175, 195-201 (1995)
Strax, P. Control of breast cancer through mass screening: From research to action. Cancer,
63: 1881-1887(1989)
Streisinger, G., Okada, Y., Emrich, J., Newton, J., Tsugita, A., Terzaghi, E., and Inouye, M.
Frameshift mutations and the genetic code. Cold Spring Harbor Symp. Quant. Biol., 31: 7784 (1966)
Stephen, D., Jones, C., Schofield, J.P. A rapid method for isolating high quality plasmid
DNA suitable for DNA sequencing. Nucleic Acids Res., 18: 7463-7464 (1990)
268
Sumida, C., Lecerf, F., Pasqualini, J.R. Control of progesterone receptors in foetal uterine
cells in culture: effects of estradiol, progestins, antiestrogens, and growth factors.
Endocrinology, 1 2 2 : 3-11 (1988)
Sumida, C., Pasqualini, J.R. Antiestrogen antagonises the stimulatory effect of epidermal
growth factor on the induction of progesterone receptor in foetal uterine cell in culture.
Endocrinology, 124: 591-597 (1989)
Sumida, C., Pasqualini, J.R. Stimulation of progesterone receptors by phorbol ester and
cyclic AMP in foetal uterine cells in culture. Mol. Cell Endocrinol, 69: 207-215 (1990)
Swift, M., Morrell, D., Massey, R.B., Chase, C. Influence of breast cancer in 161 families
affected by ataxia telangiectasia. N. Eng. J. Med., 325- 326: 1831 (1991)
Tabar, L., Fagerberg, C.J.G, Gad, A., Baldetorp, L., Holmberg, L.H., Grontoft, O. Reduction
in mortality from breast cancer after mass screening with mammography. Lancet, I: 829-832
(1985)
Takahashi, H., Chiu, H-C., Bandera, C. A., Behbakht, K., Liu, P. C, Couch, F. J., Weber, B.
L., liVolsi, V. A., Furusato, M., Rebae, B., Cardonick, A., Benjamin, I., Morgan, M. A.,
King, S. A., Mikuta, J. J., Rubin, S. C., and Boyd, J. Mutations of the BRCA2 gene in
ovarian carcinomas. Cancer Res., 56: 2738-2741 (1996)
Takimoto, G.S., Tasset, D.M., Eppert, E.C., Horwitz, K.B. Hormone induced progesterone
receptor phosphorylation consists of sequential DNA-independent and DNA dependent
stages. Analysis with zinc finger mutants and the progesterone antagonist ZK-98299. Proc.
Natl. Acad. Sci, USA, 89: 3050-3053 (1992)
Takita, K., Sato, T., Miyagi, M., Watatani, M., Akiyama, F., Sakamoto, G., Kasumi, F.,
Abe, R., Nakamura, Y. Correlation of loss of alleles on the short arm of chromosome 11 and
chromosome 17 with metastasis of 10 breast cancer to lymph nodes. Cancer Res., 52: 39143917(1992)
Tandon, A.C., Clarke, G.M., Chamness, G.C., Chirgwin, J.M., McGuire, W.L. Cathepsin D
and prognosis in breast cancer. New Eng. J. Med., 322: 297-302 (1990)
T ’Ang, A., Varley, J.M., Chakraborty, S., Murphree, A.L., Fung, Y-KT. Structural
rearrangement o f the retinoblastoma gene in human breast carcinoma. Science, 242: 263-266
(1988)
Tasset, D., Tora, I., Fromental, C., Scheer, E., Chambon, P. Distinct classes of
transcriptional activating domains function by different mechanisms. Cell, 62: 1177-1187
(1990)
Taute, T., Rentz, M. Simple DNA sequence of Drosophila virillis; Isolated by screening
with RNA. J. Mol. Biol., I l l : 229-235 (1984)
Terenius, L. Selective retention of estrogen isomers in estrogen dependent breast tumours of
rats demonstrated by in vitro methods. Cancer Res., 28: 328 (1968)
269
Theillet, C., Liderau, R., Escot, C., Hutzell, P., Brunet, M., Gest, J., Schlom, J., Callahan, R.
Loss o f c-Ha-ras-1 allele and aggressive human primary breast carcinomas. Cancer Res., 46:
4776-4781 (1986)
Thiagalingam, S., Lengauer, C., Leach, C., Schutte, Hahn, S. A., Overhauser, j., Willson, J.
K. V., Markowitz, S., Hamilton, S. R., Kem, S. E., Kinzler, K. W., and Vogelstein, B.
Evaluation of candidate tumour-suppressor genes on chromosome 18 in colorectal cancers.
Nature Genet., 13: 343-346 (1996)
Thibodeau, S.N., Bren, G., Schaid, D. Microsatellite instability in cancer of the proximal
colon. Science, 260: 816-819 (1993)
Thibodeau, S. N., French, A. J., Roche, P. C., Cunningham, J. M., Tester, D. J., Lindor, N.
M., Moslein, G., Baker, S. M., Liskay, R. M., Burgart, L. J., Honchel, R., and Hailing, K. C.
Altered expression o f hMSH2 and hMLHl in tumours with microsatellite instability and
genetic alterations in mismatch repair genes. Cancer Res., 56: 4836-4849 (1996)
Thomas, G.A., Raffel, C. Loss o f heterozygosity on 6q, 16q, and 17p in human central
nervous system primitive neuroectodermal tumours. Cancer Res., 51: 639-643 (1991)
Thompson, A.M., Morris, R.G., Wallace, M., Wyllie, A.H., Steel, C.M., Carter, D.C. Allele
loss from 5q21 (APC/MCC) and 18q21 (DCC) and DCC mRNA expression in breast
cancer. Br. J. Cancer, 68: 64-68 (1993)
Thor, A.D., Moore, D.H. II., Edgerton, S.M., Kawasaki, E.S., Reihaus, E., Lynch, H.T.,
Marcus, J.N., Schwartz, L., Chen, L.C. Accumulation of p53 tumour suppressor gene
protein: an independent marker of prognosis in breast cancer. J. Natl. Cancer Inst., 84: 845855 (1992)
Thorpe, S.M., Rochefort, H., Garcia, M., Freiss, G., Christensen, J., Khalaf, S., Paolucci, F.,
Parr, B., Rasmussen, B.B., Rose, C. Association between high concentrations of 52kDa
cathepsin D and poor prognosis in primary breast cancer. Cancer Res., 49: 6008-6014
(1989)
Toft, D., Gorski, J. A receptor molecule for estrogens; isolation from the rat uterus and
preliminary characterization. Proc. Natl. Acad. Sci. USA 55: 1574 (1966)
Tokunga, M., Land, C.E., Yamamoto, T. Breast cancer among atomic bomb survivors. In
Boice, J.D. Jr., Fraumeni, J.R. (eds): Radiation Carcinogenesis: Epidemiology and
Biological Significance, pp 45-55. New York, Raven Press, 1984.
Tora, L., Gronmeyer, H., Turcotte, B., Gaub, M.P., Chambon, P. The N-terminal region of
the chicken progesterone receptor signifies target gene activation. Nature, 333: 185-188
(1988)
Tora, L., White, J., Brou, C., Tasset, D., Webster, N., Scherrer, E., Chambon, P. The human
estrogen receptor has two independent nonacidic transcriptional activation functions. Cell,
59:477-487(1989)
270
Toutieu, I., Mathieu, M., Rochfort, H. Stable transfection of the oestrogen receptor cDNA
int HeLa cells induces oestrogen responsiveness of endogenous cathepsin D gene but not of
cell growth. Biochem. Biophys. Res. Commin., 169: 109-115 (1989)
Toutieu, I., Vignon, F., Cavialles, V., Rochefort, H. Hormonal regulation of cathepsin D
following transfection of the oestrogen or progesterone receptor into three sex steroid
hormone resistant cell lines. J. Steroid Biochem. Mol. Biol, 40: 231-237 (1991)
Toyama, T., Iwase, H., Yamashita, H., Iwata, H., Yamashita, T., Ito, K., Hara, Y., Suchi,
M., Kato, T., Nakamura, T., and Kobayashi, S. Microsatellite instability in sporadic human
breast cancers. Int. J. Cancer, 68: 447-451 (1996a)
Toyama, T., Iwase, H., Iwata, H., Hara, Y., Omoto, Y., Suchi, M., Kato, T., Nakamura, T.,
and Kobayashi, S. Microsatellite instability in in situ and invasive sporadic breast cancers of
Japanese women. Cancer Letts., 108: 205-209 (1996b)
Trent, J. Cytogenetic and molecular biologic alterations in human breast cancer: a review.
Breast Can. Res. Treat., 5: 221-229 (1985)
Trent, J.M., Stanbridge, E.J., McBride, H.L., Meese, E.U., Casey, G., Araujo, D.E.,
Witkowski, C.M., Nagle, R.B. Tumourigenicity in human melanoma cell lines controlled by
introduction of human chromosome 6. Science, 247, 568-571 (1990)
Trichopoulos, D., MacMahon, B., and Cole, P. Menopause and breast cancer risk. J. Natl.
Cancer Inst., 48: 605-613 (1972)
Truant, R., Xiao, H., Ingles, C.J., Greenblatt, J. Direct interaction between the
transcriptional activation domain o f human p53 and the TATA box-binding protein. J. Biol.
Chem., 268: 2284-22887 (1993)
Tso, J.Y., Van den Berg, D.J., Korn, L.J. Structure of the gene for Xenepus transcription
factor TFIIIA. Nucleic Acids Res., 14:2187-2200(1986)
Turcotte, B., Meyer, M.E., Bellard, M., Dretzen, G., Grommeyer, H., Chambon, P. Control
o f transcription of the chicken progesterone receptor gene. J. Biol. Chem., 266: 2582-2589
(1990)
Uchida, T., Wada. C., Wang, C., Ishida, H., Egawa, S., Yokoyama, E., Ohtani, H., Koshiba,
K. Microsatellite instability in prostate cancer. Oncogene, 10: 1019-1022 (1995)
Unesono, K., Evans, R.M. Determinants of target gene specificity for steroid/thyroid
hormone receptors. Cell, 57: 1139-1146 (1988)
Unesono, K., Giguere, V., Glass, C.K., Rosenfeld, M.G., Evans, R.M. Retanoic acid and
thyroid hormone induce gene expression through a common responsive element. Nature,
336: 262-265 (1989)
Unger, T., Nau, M.M., Segal, S., Minna, J.D. A transdominant regulator of transcription
whose function is ablated by mutations occurring in human cancer. EMBO J., 11: 13831390(1992)
271
UK National case-control study group. Oral contraceptive use and breast cancer risk in
young women. Lancet, I: 973-982 (1989)
Valagussa, P., Bondadonna, G., Veronsei, V. Patterns of relapse and survival following
radical mastectomy. Cancer, 41: 1170-1178 (1978)
van der Vijer, M.J., Peterse, J.L., Mooi, W.J., Wiserman, P., Lomans, J., Dalesio, O., Nuse,
R. Afew-protein over-expression in breast cancer: association with comedo-type ductal
carcinoma in situ and limited prognostic value in Stage II breast cancer. New Eng. J. Med.,
319: 1239-1245 (1988)
Van Dyke, M.W., Sawadogo, M., Roeder, R.G. Stability of transcription complexes on class
II genes. Mol. Cell. Biol., 9: 342-344 (1989)
van Laar, J.H., Berrevoets, C.A., Trapman, J., Zegers, N.D., Brinkman, A.O. J. Hormonedependant androgen receptor phosphorylation is accompanied by receptor transformation in
human lymph node carcinoma of the prostate cells. Biol. Chem., 266: 3734-3738 (1991)
Van-Prooijen-Knegt, A.C., Raap, A.K., Van-der-Burg, M.J., Vrolijk, J., Van-der-Ploeg, M.
Spreading and staining o f human metaphase chromosomes on aminoalkylsilane-treated
slides. Histochem. J., 14:333-344 (1982)
Varley, J.M., Swallow, J.E., Brammer, W.J., Whittaker, J.L., Walker, R.A. Alterations to
either c-erbB2 (neu) or c-myc proto-oncogenes in breast carcinomas correlate with poor
short-term prognosis. Oncogene, 1:423-430 (1987)
Varley, J.M., Armour, J., Swallow, J.E., Jefferys, A.J., Ponder, B.A.J., Tang, A., Fung, YK.T., Brammer, W.J., Walker, R.A. The retinoblastoma gene is frequently altered leading to
loss o f expression in primary breast tumours. Oncogene, 4: 725-730 (1989)
Varley, J.M., Brammer, W.J., Lane, D.P., Swallow, J.E., Dolan, C., Walker, R.A. Loss of
chromosome 17 p l3 sequences and mutation of p53 in human breast carcinomas. Oncogene,
6:413-421 (1991)
Venter, D.J., Tuzi, N.L., Kumar, S., Gullick, W.J. Over-expression of the c-erbB2
oncoprotein in human breast carcinomas: immunohistological assessment correlates with
gene amplification. Lancet ii: 69-72 (1987)
Verbeek, A.L.M., Holland, R., Sturmans, F., Hendricks, J.H.C.L., Mravunacm M., Day,
N.E. Reduction o f breast cancer mortality through mass screening with modem
mammography. First results of the Nijmegen project, 1975-1981. Lancet, II: 122-1224
(1984)
Veronese, S.M., Gambacorta, M. Detection of Ki67 proliferation rate in breast cancer.
Correlation with clinical and pathologic features. Am. J. Clin. Pathol, 6: 30-34 (1991)
Vleminckx, K., Vakaet, L., Mareel, M.M., Fiers, W., Van Roy, F. Genetic manipulation of
E-cadherin expression by epithelial tumour cells reveals an invasion suppressor role. Cell,
66: 107-119(1991)
272
Volgestein, B., Fearon, E.R., Hamilton, S.R. Kern, S.E., Presinger, A.C., Leppert, M.,
Nakamura, Y., White, R., Smits,
Bos, J.L. Genetic alterations during colorectal
tumour development. New Engl. J. Med., 319: 525-532 (1988)
Walker, G., Palmer, J.M., Walters, M.K., Nancarrow, D.J., Parsons, P.G., Hayward, N.K.
Simple tandem repeat allelic deletions confirm the preferential loss of distal chromosome 6q
in melanoma. Int. J. Cancer, 58, 203-206 (1994)
Walker, R.A., Camplejohn, R.S. Comparison of monoclonal antibody Ki67 reactivity with
grade and DNA flow cytometry o f breast carcinomas. Br. J. Cancer. 57: 281-283 (1988)
Walker, R.A., Gullick, W.J., Varley, J. An evaluation of immunoreactivity for c-erbB-2
protein as a marker o f short term prognosis in breast cancer. Br. J. Cancer, 60: 426-429
(1989a)
Walker, R.A., Senior, P.V., Jones, J.L., Critchley, D.R., Varley, J.M. An immunohistical
and in-situ hybridisation study o f c-myc and c-erbB2 expression in primary breast tumours.
J. Pathol., 158: 97-105 (1989b)
Walker, R.A., Dearing, S.J., Lane, D.P., Varley, J.M. Expression of p53 protein in
infiltrating and in-situ breast carcinoma. J. Pathol., 165: 203-211 (1991)
Walker, R.A., Dhadly, P.P.S., Cowl, J., Jones, J.L. Oestrogen receptor, epidermal growth
factor receptor and oncoprotein expression in non-involved tissue of cancerous breasts. The
Breast, 2: 97-91 (1992)
Walker, R.A., Denley, H., Dookeran, K.A. Cathepsin D in breast carcinomas - the role of
the stromal cell competent. Oncology Rep, 1: 227-231 (1994)
Walker, R.A., Rajakariar, R., Dookeran, K.A. Comparison of pS2 and cathepsin D in
primary breast carcinomas. The Breast, 4: 137-142 (1995) 1995
Walker, R. A. Breast cancer: Oncogenes and tumour suppressoe genes. In: The biology of
female cancers, pp. 61-74. (Eds). Miller, W. R, Langdon, S., and Berchuck, A. CRC Press.
Florida. (1997)
Walter, P., Green, S., Greene, G., Krust, A., Bomert, J.M.,. Jeltsch, J.M., Staub, A., Jensen,
E., Scarce, G., Waterfield, M., Chambon, P. Cloning of the human oestrogen receptor
cDNA. Proc. Natl. Acad. Sci. USA, 82: 7889-7893 (1985)
Wang, Y., Miksicek, R.J. Identification of a dominant negative form of the human estrogen
receptor. Mol. Endocrinol., 5: 1707-1715 (1991)
Wang, H., Peters, G.A., Zeng, X., Tang, M., Ip, W., Khan, S.A. Yeast two-hybrid system
demonstrates that the estrogen receptor dimerisation is ligand-dependent in Vivo. J. Biol.
Chem., 270: 23322-23329 (1995)
Wang, M., Dotzlaw, H., Fuqua, S.A.W., Murphy, L.C. A point mutation in the human
estrogen receptor gene is associated with the expression of an abnormal estrogen receptor
mRNA containing a 69 novel nucleotide insertion. Breast Cancer Res. Treat., 44: 145-151
(1997)
273
Warren, W., Eeles, R.A., Ponder, B.A.J., Easton, D.F., Averill, D., Ponder, M.A., Anderson,
K., Evans, A.M., De Mars, R., Love, R. No evidence for germline mutations in exon 5-9 of
the p53 gene in 25 breast cancer families. Oncogene, 7: 1043-1056 (1992)
Washburn, T., Hocutt, A., Brautigan, D.L., Korach, K.S. Uterine oestrogen receptor in vivo:
Phosphorylation o f nuclear specific forms on serine residues. M ol Endocrinol., 5: 235-242
(1991)
Webb, P., Lopez, G.N., Uht, R.M., Kushner, P.J. Tamoxifen activation of the estrogen
receptor/AP-1 pathway - potential origin for the cell specific estrogen-like effects of
antiestrogens. Mol. Endocrinol, 9: 443-456 (1995)
Weber, J.L., and May, P.E. Abundant class of human DNA polymorphism which can be
typed using the polymerase chain reaction. Am. J. Hum. Genet., 44, 388-396 (1989)
Webster, N.J.G., Green, S., Jin, R.R., Chambon, P. The Hormone binding domains of the
estrogen and glucocorticoid receptors contain an inducible transcription activation function.
Cell, 54: 199-207 (1988)
Webster, N.J.G., Green, S., Tasset, D., Ponglikmongkol, M., Chambon, P. The
transcriptional activation function located in the hormone binding domain of the human
estrogen receptor is not encoded in a single exon. EMBOJ., 8: 1441-1446 (1989)
Weidner, N., Folkman, J., Pozza, F., Bevilacqua, P., Allred, E., Moore, D.H., Meli, S.,
Gasparini, G. Tumour angiogenesis: A new significant prognostic indicator in early-stege
breast carcinoma. J. Natl. Cancer Inst., 84: 1875-1887 (1992)
Weigel, N., Carter, T.H., Schrader, W.T., O’Malley, B.W. Chicken progesterone receptor is
phosphorylated by a DNA dependent protein kinase during in vitro transcription assays.
Mol. Endocrinol., 6: 8-14 (1992)
Weigel R.J., Crooks, D.L., Iglehart J.D., deConnick E.C. Quantitative analysis of the
transcriptional start sites o f the estrogen receptor in breast carcinoma. Cell Growth. Diff., 6:
707-711 (1995)
Weiler, I.L., Lew, D., Shapiro, D.J. The Xenopous leavis estrogen receptor: sequence
homology with the human and avian receptors and identification of multiple estrogen
receptor messenger ribonucleic acids. Mol. Endocrin., 1: 355-362 (1987)
Weinberg, R.A. Oncogenesis, anti-oncogenes, and the molecular basis of multistep
carcinogenesis. Cancer Res., 49: 3713-3721 (1989)
Weinberg, R.A. Tumour suppressor genes. Science, 254: 1138-1146 (1991)
Westly, B., Rochfort, H. A secreted glycoprotein induced by estrogen in human breast
cancer cell lines. Cell, 20: 352-362 (1980)
White, J.H., Metzger, D., Chambon, P. Expression and function of the human oestrogen
receptor in yeast. Cold Spring Harb. Symp. Quant. Biol., 53 (II): 819-828 (1988)
274
WHO Histological Typing of Breast Tumours, 2nd ed, p i 9. International histological
classification of tumours no. 2. Geneva, World Heath Organisation, 1981.
WHO collaberative study of neoplasia and steroid contraceptives. Breast cancer and
combined oral contraceptives: results from a multinational study. Br. J. Cancer, 61: 110119(1990)
Williams, D., Gorski, J. Kinetic and equilibrium analysis of estradiol in uterus: A model of
binding-site distribution in uterus cells. Proc. Nat. Acad. Science USA, 69: 3464 (1972)
Winquist, R., Hampton, G.M., Mannermoa, A., Blanco, G., Alavaikko, M., Kiviniemi, H.,
Taskinen, P.J., Wright, F.A., Newsham, I., Cavenee, W.K. Loss of Heterozygosity for
chromosome 11 in primary human breast tumours is associated with poor survival after
metastasis. Cancer Res., 55: 2660-2664 (1995)
Winstanley, J., Cooke, T., Murray, G.D., Platt-Higgins, A., George, W.D., Holt, S., Mystov,
M., Spedding, A., Barraclough, B.R., Rudland, P.S. The long -term prognostic significance
of c-erb-&2 in primary breast cancer. Br. J. Cancer, 63: 447-450 (1991)
Winter, E., Yamamoto, F., Almoguera, C., Perucho, M. A method to detect and characterise
point mutations in transcribed genes. Proc. Natl Acad. Sci. USA, 82: 7575-7579 (1985)
Wolf, D.M., Jordan, V.C. The estrogen receptor from tamoxifen stimulated MCF-7 tumour
variant contains a point mutation in the ligand binding domain. Br. Cancer Res. Treat., 31:
129-138 (1994a)
Wolf, D.M., Jordan, V.C. Characterization of tamoxifen stimulated MCF-7 tumour variants
grown in athymic mice. Br. Cancer Res. Treat., 31: 117-128 (1994b).
Wooster, R., Neuhausen, S. L., Mangion, J., Quirk., Y., Ford, D., Collins, N., Nguyen, K.,
Seal, S., Tran, T., Averill, D., Fields, P., Marshall, G., Narod, S., Lenoir, G. M., Lynch, H.,
Feunteun, J., Devilee, P., Comelisse, C. J., Menko, F. H., Daly, P. A., Ormiston, W.,
McManus, R., Pye, C., Lewis, C. M., Cannon-Albright, L. A., Peto, J., Ponder, B. A. J.,
Skolnick, M. H., Easton, D. F., Goldgar, D. E., and Stratton, M. R. Localization of a breast
cancer susceptibility gene, BRCA2, to chromosome 13ql2-13. Science, 265: 2088-2090
(1994a)
Wooster, R., Cleton-Jansen, A-M., Collins, N., Mangion, J., Comelisse, R.S., Cooper, .S.,
Gusterson, B.A., Ponder, B.A.J., von Deimling, A., Wiestler, O.D., Comelisse, C.J.,
Devilee, P., Stratton, M.R. Instability of short tandem repeats (microsatellites) in human
cancers. Nature Genet., 6: 1641-1644 (1994b)
Wooster, R., Bignell, G. Lancaster, J., Swift, S., Seal,S., Mangion, J., Collins, N., Gregory,
S., Gumbs, C., Micklem, G., Barfoot, R., Hamoudi, R., Patel, S., Rice, C., Biggs, P.,
Hashim, Y., Smith, A., Conner, F., Arason, A., Gudmundsson, J., Ficenec, D., Kelsell, D.,
Ford, D., Tonin, P., Bishop, D. T., Spurr, N. K., Ponder, B. A. J., Eeles, R., Peto, J.,
Devilee, P., Comelisse, C., Lynch, H., Narod, S., Lenoir, G., Egilsson, V., Barkadottir, R.
B., Easton, D. F., Bentley, D. R., Futreal, P. A., Ashworth, A., and Stratton, M. R.
Identification of the breast cancer susceptibility gene, BRCA2. Nature, 378: 789-792,
(1995).
275
Wreschner, D.H., Rechavi, G. Differential mRNA stability to reticulosyte ribonuclease
correlates with 3’ non coding (U)nA sequences. Eur. J. Biochem., 172: 333-340 (1988)
Xiong, Y., Menninger, J., Beach, D., Ward, D.C. Molecular cloning and chromosomal
mapping o f CCND genes encoding human D-type cyclins. Genomics, 13: 575-584 (1992)
Yamada, H., Wake, N., Fujimoto, S., Barrett, J.C., Oshimura, M. Multiple chromosomes
carrying tumour suppressor activity for a uterine endometrial carcinoma cell line identified
by microcell-mediated chromosome transfer. Oncogene, 5,1141-1147 (1990)
Yap, E.P.H., O'D McGee, J. Nonisotopic SSCP detection in PCR products by ethidium
bromide staining. Trends Genet., 8(2):49 (1992)
Yee, C.J., Roodi, N., Verrier, C.S, Pari, F.F. Microsatellite instability and loss of
heterozygosity in breast cancer. Cancer Res., 54: 1641-1644 (1994)
Yoshitaka, T., Matsubara, N., Ikeda, M., Tanino, M., Hanafusa, H., Tanaka, N., and
Shimizu, K. Mutations o f the E2F-4 trinucleotide repeat in colorectal cancer with
microsatellite instability. Biochem. Biophys. Res. Comm., 227: 553-557 (1996)
Young, P.C.M., Ehrlich, C.E., Einhom, L.H. Relationship between steroid receptors and
response to endocrine therapy and cytotoxic chemotherapy in metastatic breast cancer.
Cancer 46: 2961 (1980)
Zalent, A., Mendelsonhn, C., Kastner, P., Krust, A., Gamier, J.M., Ruffenach, F., Leroy, P.,
Chambon, P. Differentially expressed isoforms of the mouse retanoic acid receptor beta
generated by usage o f two promoters and alternative splicing. EMBOJ., 10: 71-81 (1991)
Zeillinger, R., Kury, F., Czerwenka, K., Kubista, E., Sliutz, G., Knogler, W., Huber, J.,
Zielinski, C., Reiner, G., Jakesz, R., Staffen, A., Reiner, A., Wrba, F., Spona, J. HER-2
amplification, steroid receptors and epidermal growth factor receptor in primary breast
cancer. Oncogene, 4: 109-114 (1989)
Zhang, Q-X, Borg, A., Fuqua, S.A.W. An exon 5 deletion variant of the estrogen receptor in
human breast cancer. Cancer Res., 53: 5882-5884 (1993)
Zhang, Q.-X., Hilsenbeck, S.G., Fuqua, S.A.W., Borg, A. Multiple splicing variants of the
oestrogen receptor are present in individual human breast tumours. J. Steroid Biochem.
Molec. Biol., 59: 251-260 (1996)
Zhang, R., Wiley, J., Howard, S.P., Meisner, L.F., Gould, M.N. Rare clonal karyotype
variants in primary cultures o f human breast carcinoma cells. Cancer Res., 49: 444-449
(1989)
Zhen, L., Swank, R.T. A simple and high yield method for recovering DNA from agarose
gels. BioTechniques, 14: 894-898 (1993)
Zhoa, Y., Escot, C., Maudelonde, T., Puech, C., Rouanet, P., Rochefort, H. Correlation
between mannose-6-phosphate/IGFII receptor and cathepsin D RNA levels by in situ
hybridization in benign and malignant mammary tumors. Cancer Res., 53, 2901-2905
(1993)
276
Zhou, D., Battifora, H., Yokata, J., Yamamoto, T., Cline, M.J. Association of multiple
copies o f the c-e/7>-B2 oncogene with the spread of breast cancer. Cancer Res., 47: 61236125 (1987)
Zhou, D., Ahuja, H., Cline, M.J.: Proto-oncogene abnormalities in human breast cancer: cerb-Y$2 amplification does not correlate with recurrence of disease. Oncogene, 4: 105-108
(1989)
Zhu, X., Daffada, A.A.I., Chan, C.M.W., Dowsett, M. Identification of an exon 3 deletion
splice variant androgen receptor mRNA in human breast cancer. Int. J. Cancer, 72: 574-580
(1997)
Zhuang, Z., Merino, M. J., Chuqui, R., Liotta, L. A., and Emmert-Buck. Identical allelic loss
on chromosome 11 q l 3 in microdissected in situ and invasive human breast cancer. Cancer
Res., 55: 467-471 (1995)
Zutter, M.M., Mazoujian, G., Santoro, S.A. Decreased expression of integrin adhesive
protein receptors in adenocarcinoma of the breast. Am J. Pathol., 137: 863-870 (1990)
Zutter, M.M., Krigman, H.R., Santoro, S.A. Altered integrin expression in adenocarcinoma
of the breast. Analysis by in situ hybridisation. Am. J. Pathol., 142: 1439-1448 (1993)
277
(
\
J
Br.tis.1 Journal o f C s r t c f {\99Tt 75(9). 1324-1329
O 1 997 Cartc«r Resaarcn Campaign
L o s s o f h e t e r o z y g o s it y at c h r o m o s o m e 6q in preinvasive
a n d e a r l y in v a siv e b rea st c a r c i n o m a s
SA C h a p p ell, T W alsh, RA Walker and JA Shaw
B ra a s t C a n c e r R e s e a r c h Unit. Department of Patnotogy. University of Leicester. Clinical S c ie n c e s Building. Glenfield General Hospital. GroPy Road.
Letcester LS3 9QP. UK
\
Summary We have used polymerase chain reaction (PCR) analysis to study the incidence of allelic imbalance at four polymorphic
microsateilite markers on chromosome 6q25.l-27, three dinucleotide repeats and one tnnucleotide repeat, for microdissected tumour foci
from a group of 75 early' breast carcinomas. The tumours comprised 16 preinvasive cases of ductal carcinoma in situ (DCIS) and 59
mammograohically detected early invasive carcinomas. Loss of heterozygosity (LOH) was detected at all four loci and in all types and grade
of cisea se. The frequency of LOH ranged from 23% to 50% depending on the marker studied. The highest frequency of LOH was observed
at the D6S186 locus for the cases of DOS and at the oestrogen receptor locus for the invasive carcinomas. These data suggest that the
inactivation of tumour-suppressor genes within this region on chromosome 6q is important for the development of these early lesions.
Keywords: breast carcinoma: mammograpny; loss of heterozygosity; tumour-sucpressor genes
^
J
^
*
A ccording to the multistep model of carcinogenesis, tumours may
d evelop and progress as a result of alterations in oncogene and
tum our-suppressor gene loci. In colon cancer, a benign to malig­
nant progression with recognizable molecular changes has been
described <Fearon and Volgelstein. I990). However, the situation
for b reast can cer is less clear, since there is no clear understanding
o f the natural history o f the disease.
C ytogenetic analyses o f primary breast tumours have identified
frequent alterations to a number of chromosomes, notably dele­
tions. suggesting the potential localization o f tumour-suppressor
genes (fo r a review, see Devilee and Comelisse. 1994). These
studies dem onstrated that deletion of chromosome 6q was one o f
the m ost frequent chromosomal changes (Dutrillaux et al. 1990:
M ars and Saunders. 1990). A subsequent study, using Southern
analysis o f restriction fragment length polymorphisms to compare
co n stitu tio n al and tum our DNAs. identified chromosome 6q as the
second m ost frequent site for allelic loss (loss o f heterozygosity)
after ITp in breast cancer (Deville et al. 1991).
O th er ev id en ce for the presence of putative tumour-suppressor
genes on chrom osom e 6q comes from chromosome-mediated
tran sfer experim ents o f normal chromosome 6 into melanoma cell
lines iT ren t et al. 1990). uterine endometrial cell lines (Yamada
et al. 1990) and the breast cancer cell lines. MDA-MD23I and
M C F-7 i N eg n m et al. 1994). all resulting in the suppression o f
tu m o n g en esis.
The advent o f polym erase chain reaction (PCR) analysis o f
m icrosateilite polym orphisms (Weber and May. 1989) has
c o n firm e d the cytogenetic evidence for chromosomal deletion at
6q and has enabled construction of a more detailed deletion map.
A iieiic loss at 6q24-27 has been observed in different tumour
ty p e ', nciuding breast carcinoma lOrphanos et al. 1995). ovarian
J ’ Jury
’ 9 96
a 9. 593 2 ' C c : c c 9 f '9 9 6
A c 393193 ’ i \C'*9rno9f 1996
Cc~9sccrc9rc9 '3: JA Shaw
-
1324
carc in o m a (Saito et al. 1992: Rodabaugh et al. 1995). hepatic
carc in o m a (De Souza et al. 1995). small-cell lung carcinoma
(M erlo e: al. 1994). renal cell carcinoma (Morita et al, 1991).
m alig n an t melanoma (Millikin et al. 1991: Walker e: al. 1994) and
non-H odgkin s lymphoma (Menasce et al. 1994). The reported
freq u en cies o f allelic loss range from
to 6 0 ^ depending on
the tu m o u r types and markers studied. This shared region of allelic
loss m ay harbour putative tumour-suppressor genes that are
p leiotropic for these tumour types and reflect a common mecha­
nism o f tumorigenesis.
R ecent detailed analyses of microsateilite markers on chromo­
som e 6q in breast cancers have highlighted two key regions
show ing high levels of LOH at 6q 13 and 6q26-27. indicating the
p resen ce o f at least two tumour-suppressor genes (Devilee et al.
1991: O rphanos et al. 1995). Since these studies were concerned
w ith sy m ptom atic, well-established breast carcinomas, it is not
c lea r w h ether allelic loss on chromosome 6q is an early event in
the developm ent of breast cancers. Small, mammographicallv
d e tec ted breast cancers form a useful group for study of the
in volvem ent o f tumour-suppressor genes in tumour development
and eariier stages of progression. In this report, we examined LOH
at the m ore distal region, chromosome 6q25-q27. using four poly­
m orphic microsateilite markers, in a group of 75 ‘early' lesions
com prising 59 mammographicaily detected invasive carcinomas
and 16 preinvasive lesions of ductal carcinoma in situ (DCIS).
The m arkers span a chromosomal region of approximately
15 M b. Two o f the markers (D65136 and D6SI93' were analysed
prev iousiy in well-established carcinomas (Orphanos et al. 1995).
The tw o other markers studies comprise repeats at or within
codin g -equences that might be candidates for early' mutations in
breast cancer: a (TA), repeat positioned I kb upstream of the
o e stro g e n receptor gene <ESR Del Senno et al. 1992) and a
i CAC- . repeat within the human TATA box-binding protein (TBP)
i Poly m eropoulos et al. 1991). We nave analysed the frequency of
LOH .r. the two groups of cases and correlated these data with
o estro g en receptor t E R ) and progesterone receptor i PR >status and
o th er clmicopathological findings.
30ac
LOH at chromosome 6q in breast carcinoma
Table 1 Ciinicopathological features of 59 mammograpbically detected early
invasive breast carcinomas
Type
TuO
Lee. tub
ICC ILC
ILC
ICC
ICC
ICC
Total
Grade
Number of
cases
Tumour si2e
(mm)
Number of
cases
1
1
II
II
1
II
III
6
1
1
1
17(1)
29(2)
4
< 10
10
11
12
13
14
15
10
14
4
5
3
2
21
59
TJ
Tt
T3
1325
T4
3
59
Tuo. tubular carcinoma: Lob/tub. lobular and tubular carcinoma: IDC/ILC.
infiltrating ductal with infiltrating lobular carcinoma; ILC, infiltrating lobular
carcinoma: IOC. infiltrating ductal carcinoma. Numbers in brackets, nodepcsitive cases.
T1
-2
s
MATERIALS A N D METHODS
P a tie n ts
A to ta l o f 59 invasive breast carcinomas that were impalpable and
d e te c te d by m am m ography were studied. All were from the preva­
len t ro u n d o f screening and were detected by the Leicestershire
B re a st Screening Service. Cases of 15 mm or less in m aximum
d ia m e te r w ere exam ined. All had either axillary node sam pling or
ax illa ry dissection. None of the tumours were from women with
e ith e r a strong family history of breast cancer or any known inher­
ited predisposition to the development of tumours. Some 56 cases
w ere n o d e negative.
A to tal o f 16 cases o f pure ductal carcinoma in situ (DCIS) were
stu d ied . T hese com prised three low. three intermediate and ten
h ig h n u c le a r grade cases. Ten of these were mammographicallv
d e te c te d and six w ere clinically presenting.
T is s u e s
N
»
»s
Tt
*2
Tl
T2
M
N
ic-
A ll tissu es w ere fixed in 4% formaldehyde in saline for 18-36 h.
A fte r slicing, selected blocks were processed through graded
aic o h o is and xylene to paraffin wax. Following review o f
haem a to x y lin and eosin-stained sections, representative blocks
w ere chosen for further study. Tissue from histologically normal
m pn node served as the source of normal DNA.
Tl
1
1
N
H is to lo g y
Ail carcinom as were reported according to the NHS Breast
S creen in g Program m e National Coordinating Group for Breast
S creen in g Pathology Guidelines (1995). Infiltrating ductal carci­
n o m as w ere graded using the modified Bloom and Richardson
sy stem Elston and Ellis. 1991). Cases of DCIS were graded as
l o w . interm ediate o r high nuclear grade. All histology was under­
taken b y RA Walker. The ciinicopathological features o f the
in v a siv e carcinom as are shown in Table I.
O e s tr o g e n recep tor and progesterone receptor
im m u n o h isto c h e m istr y
A', tc ir.-b io tin com plex peroxidase immunohistochemistry was
c a n te d out for the 59 eariv invasive carcinomas as described
2 C a re e r Research Campaign 1997
Figure 1 L o ss of heterozygositv n eanv creast cancer. Genom ic DNA
s a m c ie s from paireO normai ,y rrc r •rcce N) and m icrodissectec tumour (T)
s a m p le s w e re c om pared by
amplification. electrophoresis on 6*9
s e c u e n c in g g e ls and autoradicg'acny At LOH at D6S193 m c a s e 37
( T t- T 3 ) solid co m p o n en t of innitrstmg ouctai carcinom a. (B) LOH at C 6S186
■n c a s e 2 v (T t an d 72) soiid com ccner: or infiltrating ductal carcinoma.
(C l LCH a t Q 6S193 in c a s e *02:
a r c r 2) sotid com ponent of infiltrating
d u c ta l carcin o m a. \D) LOH at C53*36 n DCIS c ase D5: (T1 and T2)
individual d u c ts. (El LCH at SR7A n case 57: (T1) a re a of DCIS. (T2) tubular
c o m p o n e n t. (T3) tubular com ponent
Bntisn Journal cf Cancer (1997) 75(9). 1324-1329
1326
SA C happell e t al
T a b le 2 P a rte m of lo s s of heterozygosity observed using four microsateilite m a rk e rs from 5 9 early invasive b re a st carcinomas
C ase no.
3
7
13
IS
17
19
21
23
29
31
37
41
49
57
59
70
76
78
80
98
102
106
108
122
Type
ILC
IDC
Tub
IDC
Tub
IDC
IDC
IDC
Lob/ tub
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
IDC
Tub
IDC
1DC
IDC
IDC
IDC
Grade
Loss of heterozygosity at markers
H scores
ESR (q25.1)
D6S186 (q26)
D6S193 (q27)
TBP (q27)
•
O
Nl
O
Nl
O
Nl
Nl
O
.N l
Nl
Nl
O
O
•
•
•
•
•
•
MSI
Nl
Nl
Nl
•
•
Nl
Nl
O
0
•
Nl
•
Nl
Nl
•
Nl
•
•
Nl
O
Nl
0
Nl
NA
Nl
II
I
I
I
I
II
II
I
I
I
II
II
II
I
I
I
I
I
I
II
II
II
II
II
•
O
•
o
o
•
•
Nl
Nl
•
•
•
•
Nl
Nl
Nl
o
Nl
•
•
Nl
Q
•
NA
NA
Q
•
•
•
•
•
•
ER
175
182
142
181
195
162
151
175
196
202
22
192
215
173
196
231
219
202.5
162
187.5
142
195
0
160
•
o
•
•
O
o
Nl
NT
Nl
Nl
Nl
NT
Nl
NT
NT
NT
Nl
Nl
PR
106
49
94
97
3
168
92
31
0
115
3
170
46
192
19.2
0
0
85
83
0
0
33.5
0
0
• c s s of h e te ro z y g o sity : O , heterozygosity: MSI. microsateilite instability; Nl, not inform ative: NA, no amplification: NT. not :ested. Tub, tubular carcinoma;
L oo /tu o . ic b u ia r a n d tubular carcinom a: ILC. .-nitrating lobular carcinom a: IOC. infiltrating du ctal carcinom a.
T a b le 3 °a t* e rr: of lo s s of heterozygosity observed using three microsateilite
m a rx e rs from 1 6 p rein v asiv e lesions of DCIS
d a se no.
0 2 (M )
D 3 iM )
D4 (Ml
D5 (M l
D8 tM )
D 12 C )
0 1 3 C)
0 1 4 iC i
C h ro m o so m e 6q m a rk e rs
G ra d e
High
Low
Low
High
Low
High
High
High
ESR
(q25.1)
D6S186
(q26)
D6S193
(q27)
TBP
(q27)
NA
•
Nl
•
Nl
Nl
•
Nl
•
Nl
•
•
Nl
•
Nl
Nl
Nl
0
•
•
•
•
0
•
NT
NT
NT
NT
NT
NT
NT
NT
• o s s of -e te ro z y g o s ity : 2 . heterozygosity: Nl. not informative: NA. no
am plificatio n : NT. n o t te ste d : (M), m ammograc-icaily detected: (C). clinically
p re s e n tin g .
p rev io u sly ‘ Ra.iakariar and Walker. 1995) with minor modifica­
tions. F o r antigen retrieval pretreatmer.t. sections were exposed to
tw o c y c le s o f m icrow aving for oestrogen receptor [mouse mono­
clo n al ID 5 D a k o ] and progesterone receptor [mouse monoclonal
N C L -P g R N ovaC astra)|.
pair. D N A was extracted from 10-um paraffin-embedded sections
as d escrib ed previously (Shaw et al. 1996).
M icrodissection o f tumour foci from invasive carcinomas and
areas o f DCIS was carried out using a method based on that
describ ed by Koreth et al (1995 >. In brief, serial 10-um paraffin
sections w ere deparaffinized in xylene ( 2 x 5 min) and dehydrated
in
ethanol (2 x 2 min) and 9 5 ^ ethanol ( 1 x 2 min). and rehydrated in w ater before staining. Tissues were stained in
eosin
solution for 20 s. washed in water and allowed to air dry. A serial
referen ce slide for each tumour was stained with haematoxylin and
eosin. dehydrated and coverslipped. Tumour foci of interest from
the invasive carcinomas included tubular, solid, invasive lobular
co m p o n en ts and preinvasive areas of ductal carcinoma in situ
(D C IS) w ithin infiltrating ductal carcinomas. These ureas were
visualized using the haematoxylin and eosin reference and
m icrodissected from corresponding eosin sections using a x 40
m agnification microdissection microscope (American Optical
C o rp o ratio n ) using sterile, 20-ul drawn-out glass Pasteur pipettes.
T u m o u r foci (approximately 100 cells) were placed into 25 ,ul of
d ig estio n buffer [100 m.vt Tris-HCl ipH 7.6). 1 m.vt EDTA (pH 8).
200 jig m i-1 proteinase K] and incubated at 553C for 3 h. then at
9 -* C for 10 min. Volumes (2 pi) of this mixture were used in the
PCR analysis.
99%
0.5%
PCR a n a ly s is at 6q 2 5 .1 -2 7
DNA e x tr a c tio n and m icrodissection from paraffin
e m b e d d e d s e c t io n s
F o rm alin -fix ed , paraffin-embedded tissue from breast tumour
sa m c ie s and non-involved lymph nodes s e r v e d as the source o f
tu m o u r a n d norm al DNA respective!). For each tum our-normal
Britisr. Journal of C areer (1997) 75(9). ’324-1329
A :otai o f 75 ’early' breast carcinomas were studied for LOH at
four polym orphic markers from chromosome 6q25.l-27: the
oestro g en receptor (ESR) at 6q25.l (Del Senno et al. 1992),
D 651S6 ' 6q26) and D6S193 (6q2") (Saitoetal. 1992: Orphanos et
ai. 1995 ■and the TATA box-binding protein (TBP) gene at 6q27
© Cancer Research Campaign 1997
LOH at chromosome 6q in breast carcinoma 1327
(P olym eropoulos et al. 1991: Saito et al. 1994). PCR reaction
c o n d itio n s were as follows: 45 m.M Tris-HCl. pH 8.8. 11 m.M
am m o n iu m sulphate. 4.5 m.M magnesium chloride, 200 p.M dTTP.
d C T P and dGTP. 25 \isi dATP (Pharmacia. UK), 0.3 ^1 [a "SJdeoxyadenosine-5 -trip h o sp h ate (600 Ci mmol-1, 10 mCi ml“‘;
IC N P harm aceuticals. UK). 113 ug m l-1 bovine serum albumin
(B e e h n n g h e r M annheim ). 6.7 m.M (3-mercaptoethanol. 4.4 j im
H D T A . pH 8.0. 10 pm ol o f forward and reverse primers. 2 p.1 o f
m ic ro d isse c te d DNA and 1 unit Tciq DNA polymerase (Gibco
B R L . U K ) in a total volume o f 25 |il. Hot-start PCR was carried
o u t u sin g the follow ing cycles: 5 min denaruration at 94°C
fo llo w e d by 30 (40 for microdissections) cycles of I min denaturatio n a t 9 4 aC . I min annealing and I min extension at 72°C with a
fin al e x ten sio n o f 7 min at 72°C on a DNA Thermal Cycler
(P e rk in E lm er C etus. UK). Analysis o f PCR products was as
d e s c rib e d previously (Shaw et al. 1996).
D e t e c t io n o f LOH
A u to ra d io g ra p h s w ere scored independently by two individuals
(S C a n d JS ) and the results compared. All examples of LOH were
c o n firm e d by microdissection analysis to prepare multiple tum our
fo ci a n d then by repeating the PCR analysis where possible.
RESULTS
A to tal o f 59 early invasive breast tumours and 16 preinvasive
le s io n s o f DCIS were screened for LOH with four polym orphic
m ic ro sa te ilite markers, mapping to chromosome 6q25.1-q27.
L O H w as considered to be present when the constitutive tissue
D N A w as heterozygous (informative) for the locus under investi­
g a tio n . and w here there was complete or > 50% loss o f one allele
in th e corresponding tumour DNA as estimated by visual inspec­
tion. T h e com plex heterogeneity of the disease and the presence o f
n o n -ru m o u r cells can mask LOH. therefore all analyses w ere
c o n firm e d using DNA prepared by microdissection from different
h is to lo g ic a l tum our foci within the same tumour section (Figure
I ‘. T h e use o f microdissected tumour material produced alm ost
c o m p ie te allelic loss (Figure 1). such that densitometric analysis o f
th e d a ta w as not considered necessary.
F o r exam ple. Figure 1 B. C and D shows two invasive carci­
n o m a s and one case o f DCIS that all exhibit complete LOH for
tw o sep arate microdissected foci at the markers studied. The
tu m o u rs analysed in Figure 1A and E show some evidence o f
h e te ro g e n e ity with variation between different microdissected
feet. F o r exam ple Figure IE is an infiltrating ductal carcinom a
g ra c e I that exhibits LOH at ERTA. Analysis o f three distinct
rrd c ro c isse c te d foci shows an area of in situ carcinoma with LOH
• 71 . a tubular com ponent with heterogeneity o f LOH (T2) and a
seco n d tubular component with complete LOH (T3). This discrep­
ancy could be attributable to the presence o f contaminating non­
neoplastic stromal cells in the tubular component, even when it
w as dissected away from normal tissue. Although microdissection
analysis revealed occasional heterogeneity of distinct structural
com ponents, e.g. in situ, solid or tubular lesions within a tumour
section, with some foci showing clear LOH and others showing no
evid en ce o f LOH. no clear correlation was seen between specific
structural com ponents and LOH at any particular locus.
Table 2 and 3 summarize the observed patterns of LOH at
6 q 2 5 .1—q27 for the invasive and preinvasive study groups respec­
tively. LOH was seen for all ty pes and grades of disease studied.
A ltogether. 24 o f 59 invasive carcinomas (48%) showed evidence
o f LO H . O f these. 17 exhibited LOH only at a single locus (Table
2) and one tum our (case 31) also showed clear microsateilite insta­
bility at ESR. The situation was similar for the cases of DCIS with
eig h t o f 16 cases (50%) showing evidence of LOH and five o f
these only exhibiting LOH at a single locus (Table 3). The
frequency o f LOH at individual markers ranged from 23% to
40.6% for the early invasive cases and from 33.3% to 50% for the
D C IS group (Table 4). The highest frequency of LOH was
o b serv ed at the ESR locus for the invasive carcinomas and at the
D 6S 186 locus for the cases of DCIS. LOH was observed in both
high- and low-grade DCIS. The cases of DCIS were not studied
fo r LO H at the TBP marker owing to the paucity of available
m aterial for study.
In addition, the 59 early invasive carcinomas were studied for
o estrogen receptor and progesterone receptor status by immuno­
histochem istry (Table 2). In all. 53 (90%) were oestrogen receptor
p o sitiv e and 27 (46%) were progesterone receptor positive.
T hirteen o f the early invasive carcinomas showed LOH at the ESR
locus. O f these. 12 were ER positive (92%) and four were PR posi­
tive (31% ) by immunohistochemistry (Table 2). Therefore, LOH
at ESR is not necessarily reflected in negative values for ER
an d /o r PR. There was no significant relationship between LOH at
E SR and either ER or PR status.
DISCUSSION
U sing microdissection of distinct structural components from
w ithin a tumour tissue section and PCR amplification o f
m icrosateilite repeats, we have demonstrated loss o f heterozygosity
i LOH) at chromosome 6q25.1—2“ in foci of both DCIS and ‘early’
invasive carcinomas. Comparing the proportion of in situ lesions
w ith the proportion of invasive lesions exhibiting LOH at each
iocus revealed similar frequencies. Moreover, there was a general
spread o f LOH detected for ail types and grades of disease studied.
T hese data for the ‘early’ carcinomas suggest that the majority
o f allele losses previously reported at these loci in symptomatic
T a e ;e a 3 -m m a r y of ch ro m o so m e Sc _OH d a ta in earty invasive tum ours a n a ore E v a s iv e c a s e s of OCIS
Locus
53s
233*35
z“3=
C h ro m o s o m a l location
6a25 I
6a26
6a27
6a27
Z Ca ~C5 r R e se a rc h Campaign '997
No. of c a s e s te s te d
N o. of in fo rm a tiv e c a s e s (®»)
In v asiv e
D CIS
57
5a
57
23
15
16
16
—
In v a s iv e
32
23
35
13
56)
-42.6)
61.4)
i46)
No. of c a s e s with LOH (%)
OCIS
Invasive
OCIS
9 icC>
13(40.6)
7 (30.4)
11 (31.4)
3 (2 3 )
3 (33.3)
4 (SO)
5 (4 1 .7 )
—
8 ' 501
12 i75i
—
British Journal of Cancer (1997) 75(91 1324—1329
1323
SA Chappell e t al
in v asiv e cancers (O rphanos et al. 1995; Iwase et al. 1995) can be
found in preinvasive carcinomas. This suggestion is supported by
e v id en c e from a small number o f the infiltrating ductal carcinomas
(e.g. F ig u re 1) in which it was possible to analyse an invasive and in
situ co m p o n en t from the same tumour section. In each case. LOH
w as d etec ted in both lesions. In combination, these data suggest
that loss o f alleles on chromosome 6q is an early event in the
p ro g ressio n o f malignancy in the breast.
A lth o u g h the highest frequency of LOH was observed at the
E SR lo cu s (40.6% ) for the invasive carcinomas and at the D6S186
locus (5 0 % ) for the cases of DCIS. these differences may merely
reflect th e different groups o f cases studied and the difference in
sam p le size betw een the two groups. In addition, the slightly
h ig h er freq u en cy o f LOH noted for the cases of DCIS may reflect
the fact th at m ost were o f high nuclear grade, and therefore a more
a g g ressiv e disease type. It is interesting to note that LOH was
found fo r all three markers studied in both high- and low-grade
D C IS . su g g estin g the early involvement of loci on 6q in the devel­
o p m en t o f these lesions. In a study of chromosome I. differences
w ere fo u n d betw een chromosomal regions for the different
su b ty p es o f D CIS with no alteration at two regions in low-grade
D C IS (M u n n e ta l . 1995).
T h e p rev alen t detection of LOH at a single locus rather than
m u ltip le loci in both the 'early' invasive carcinomas and cases
o f D C IS argues against random losses resulting from general
c h ro m o so m a l instability and gross chromosomal alterations. Some
invasiv e carcinom as and cases of DCIS showed LOH at more than
one lo cu s. Since the markers studied map from 6q25-27. a
d istan ce o f several megabases, it is not possible to say whether
a c o n tig u o u s region harbouring the relevant loci has been lost,
§r w h e th e r there are distinct areas of LOH within this region o f
c h ro m o s o m e 6q.
K now ied g e o f the ER status of a carcinoma is o f value in aiding
p red ictio n o f horm one responsiveness and can provide some prog­
nostic in fo rm atio n . Tumours lacking ER and PR generally grow
faster th an those containing both ER and PR (M cGuire and Clark.
1989). O v erall. 46.1% o f informative cases for the invasive carci­
nom as ex h ib ited LOH at the oestrogen receptor locus (ESR). and
33% o f th e inform ative cases of DCIS showed LOH at ESR. These
fre q u e n c ie s are higher than the 19% LOH at ESR observed by
Iw ase e t ai «1995) and may reflect different groups o f cases, or
m ig h t be d ue to the more informative analysis o f material prepared
by m ic ro c isse c tio n in this study. The high incidence o f LOH at the
ESR lo cu s in the invasive carcinomas was not reflected by loss o f
ER as d e te c te d by immunohistochemistry. Indeed, the majority o f
the group o f early invasive lesions were ER positive. This would
be e x p ected , since the group studied was predominantly well or
m oderate!;, differentiated. Evolving tumour cells may later acquire
new p ro liferativ e pathways as a consequence o f multiple genetic
alteratio n s, enabling the tumour cells to bypass oestrogend ep en d en t proliferation (Liu et al. 1988).
O th e r 'tu d ie s have found no relationship between LOH on chro­
m o so m e ~c and ER status (Devilee et al. 1991; Iwase et al. 1995).
^uggesr.r.g that allele loss may not play an important role in the
lacx
E ? function in breast cancer tissues. However, our results
have id e n tified nine o f 13 tumours exhibiting LOH at ESR that
w ere ER po sitiv e and PR negative. This might indicate inactiva­
tion .'f the rem aining ESR alleie leading to production o f an inac­
tive ?u: u c e c ta b le ER protein, and hence loss o f PR. These cases
are candic^teN for screening for either mutations or spliced vari­
a n t' -'f the oestrogen reeep.tor gene. The identification o f spliced
?i
B rr s - „ 'c .
—at of Cancer (1997) 75(9).
1324-1329
v ariants w ould seem most likely, since ER-positive/PR-negative
phenoty pe breast tumours were shown by Fuqua et al (1993) to
co n tain a variant ER (missing exon 7) that was unable to function
as a transcriptional inducer of PR expression.
O nly three invasive carcinomas showed evidence of LOH at the
T B P locus (23% ). at 6q27. This frequency of LOH is only margin­
ally raised above expected levels for random background loss.
T h ese data suggest that inactivation of this region of the chromo­
som e is o f lesser importance than of that harbouring the ESR,
D 6 S 186 and D 6SI93 loci in these early lesions.
T h e m annose 6-phosphate/insulin-like growth factor 2 receptor
(M 6P /IG F r) functions in the intracellular trafficking o f lysosomal
en zy m es, the degradation of IGF2. a mitogen often overproduced
in tum ours (Komfeld. 1992), and the activation of the potent
grow th inhibitor, transforming growth factor (3 (Dennis and Rifkin,
1991). Som e 70% of human hepatocellular carcinomas show LOH
at this locus, which maps to chromosome 6q26-27 (Laureys et al.
1988). and 25% o f these show point mutations in the remaining
allele (De Souza et al. 1995). Clearly. M6P/IGFr might be inacti­
vated in breast cancers also. Recently. Hankins et al (1996)
reported point mutations in two comedo-tvpe (high-grade) DCIS
cases, suggesting that this is a candidate tumour-suppressor gene
in som e breast cancers. Our preliminary analyses of this locus have
show n no evidence of LOH in the well to moderately differenti­
ated invasive carcinomas suggesting that inactivation of
M 6 P /IG F r is not common in these tumours (manuscript in prepara­
tion!. In combination, these data suggest that inactivation of
M 6 P /IG F r may occur only within certain more aggressive
subgroups ipoorly differentiated cases) of breast cancers. The
frequent LOH that we have detected at 6q25.l-q27 might, there­
fore. be caused by inactivation of other tumour-suppressor genes
on chrom osom e 6q as well as M6P/IGFr.
In sum m ary, we have detected frequent LOH at three polymor­
phic loci from chromosome 6q25.l-27 in cases o f both high- and
low -grade DCIS and all types and grades of e a r l y invasive carci­
nom as. In combination, these data confirm distal chromosome 6q
as a m ajor site for genetic change in the early stages of develop­
m ent o f som e sporadic breast cancers, and form the starting point
to identify the corresponding genes.
REFEREN CES
Dei Senno L. Aguian GL and Piva R 1 1992i Dmucieotide repeat polymorphism in
the human estrogen receptor i ESR) gene. Hum Mol Genet 1; 354
D ennis PA and Rifkin DB 1 1991) Cellular actnanon of latent transforming growth
factor 3 requires binding to the cation-incerendent mannose n-phosphate/
nsuiin like growth factor type II receptor ?mc Suit Acad Set USA 88:
550 - 5*4
De Souza A. Hankins GR. Washington NIK. Fine RL. Orton TC and Jirttle RL
! 9 9 5 1 Frequent loss of heterozygosity on -o at the mannose ''-ohosphate/
insulin like growth factor (I receptor locus .n human nepatoceilurar tumours.
O n o ven e 10: 1725-1729
De- :lee ? and Comelisse CJ (1994) Somatic genetic changes tn human brexst
cancer. 3t<K'him Biopitxs Acta 1198: I ! ? - .39
De* :;ee ?. Van Vi let M. Van Sloun P. Kuiperx Dnkshoom M. Hermans J. Pearson PL
and Com eiisse Li <I 9 9 |) Alteloty pe of human brexst carcinoma: a second
maior 'ite for loss of heterozygosity is on chromosome 6q. Oncoeene 6:
■ ■ 0 5 -" : I
D utniiaux 3. Oerbault-Seureau M and Zafrar.i 3 • 1990) Characterization of
c.nrnmosomal abnormalities in human brea>; cancer. Cancer Genet Cyto^enet
49- 205-217
Elston C'.V -nd Ellis TO 1 1991) Pathological prognostic factors in brexst cancer,
i. The •• aiue of histological grade in breast cancer experience from a large
.iudy with long term follow-up. Histopatnviniy 19: 403—110
<0CancerResearchCampaign 1997
LOH at chromosome 6q in breast carcinoma
F earo n ER and Vblgelstein BA ( I990) A genetic model for colorectal
aim oungenesis. Celt 61: 759-767
F uqu a SAW. Chamness GC and McGuire WL (1993) Estrogen receptor mutations in
breast cancer. J Cell Biochem 51: 135-139
H ankins G R. De Souza AT. Bentley RC. Patel MR. Marks JR. Iglehart JD and Jirtle
RL (1996) M6P/IGF2 receptor a candidate breast tumour suppressor gene.
Oncogene 12: 2003-2009
Iw ase H. G reen man JM. Barnes DM. Bobrow L. Hodgson S and Mathew CG (1995)
Loss o f heterozygosity o f the oestrogen receptor gene in breast cancer. Br J
C anc e r 71: 448-150
KofCth J. Bethwaite PB and O 'D McGee J (1992) Mutation at 1lq23 in human nonfam ilial breast cancer. A microdissection microsateilite analysis. J Pathol 176:
U -IS
K om feld S 1 1992) Structure and function of the mannose 6-phosphate/insulin-like
grow th factor IT receptor. Annu Rev Biochem 61: 307-330
L aureys G. Barton DE. Ullrich A and Franke U (1988) Chromosomal mapping o f
the gene for type II insulin-like growth factor receptor/cation-independent
m annose 6-phosphate receptor in man and mouse. Genomics 3: 120-129
Liu E. D ollbaum C and Scon G (1988) Molecular lesions involved in the
progression o f human breast cancer. Oncogene 3: 323-327
M cG u u e W L and Clark GM (1989) Prognostic factors for recurrence and survival in
a xillary node-negative breast cancer. J Steroid Biochem 34: 145-148
M ars W M and Saunders G F (1990) Chromosomal abnormalities in human breast
cancer. Cancer Mecast Rev 9: 35-43
M enasce LP. Orphanous V. Sanubanez-Koref M. Boyle JM and Hamson CJ (1994)
C om m on region o f deletion on the long arm of chromosome 6 in nonh o d g x tn 's lymphoma and acute lymphoblastotc leukaemia. Genes Chrom
C a /i er 10: 286-288
M erio A . G abnelson E. Mabry M. Voilmer R Baylin SB and Sidransky D (1994)
H om ozygous deletions on chromosome 9p and loss of heterozygosity on 9q,
6p. and 6q in primary human small cell lung cancer. Cancer Res 54:
2 3 2 2 -2 3 2 6
M illikin D. M eeses E. Vogelstein B, Wi tow ski C and Trent J (1991) Loss of
heterozygosity for loci on the long arm of chromosome 6 in human malignant
T.eianom a. Cancer Res 51: 5449-5453
M e n u R. Saito S. Ishikawa J. Ogawa 0 . Yoshida O. Yamakawa K and Nakamura Y
•! 9 9 1 1 Common regions of deletion on chromosomes 5q. 6q. and lOq in renal
ceil carcinom a. Cancer Res 51: 5817-5820
M unn K E. W alker RA and Varley JM (1995) Frequent alterations of
chrom osom e 1 in ductal carcinoma in situ of the breast. Oncogene 10:
1653-1657
£ C a r e s ' R esearch Campaign 1997
1329
National Coordinating Group for Breast Screening Pathology (1995) Pathology
Reporting in Breast Cancer Screening. 2nd edn. NHSBSP: Sheffield
N egrini M. Sabbioni S. Possati L. Corallini A. Barbanti-Brodano G and Croce CM
(1994) Suppression of tumourogenicity of breast cancer cells by microcellmediated chromosome transfer studies on chromosome 6 and 11. Cancer Res
54: 1331-1336
O rphanps V. McGowen G. Hey Y. Boyle JM and Santibanez-Koref M (1995)
Proximal 6q. a region showing allele loss in primary breast cancer. Br J Cancer
71: 290-293
Polym eropoulos MH. Rath DS. Xiao H and Merril CR (1991) Trinucleotide repeat
polymorphism at the human transenption factor IID gene. Nucleic Acids Res
19: 4307
R ajakariar R and Walker RA (1995) Pathological and biological features of
mammographically detected invasive breast carcinomas. Br J Cancer 71:
150-154
Rodabaugh KJ. Blanchard G. Welch W R Bell DA. Berkowitz RS and Mok SC
(1995) Detailed deletion mapping of chromosome 6q in borderline epithelial
ovarian tumours. Cancer Res 55: 2169-2172
Saito S. Saito H. Koi S. Sagae S. Kudo R Saito J. Noda K and Nakamura Y (1992)
Ftne-scaie deletion mapping of the distal long arm of chromosome 6 in 70
human ovarian cancers. Cancer Res 52: 5815-5817
Saito S. Yamamoto T. Horikoshi M and Ikeuchi T (1994) Direct mapping of the
human TATA box-binding protein (TBP) gene to 6q27 by fluorescence in situ
hybridisation. Jpn J Hum Genet 39: 421-425
Shaw JA. Walsh T. Chappell SA. Carey N. Johnson K and Walker RA (19%)
Microsateilite instability in early sporadic breast cancer. Br J Cancer 73:
1393-1397
Trent JM . Stanbridge EJ. McBride KL. Meese EU. Casey G. Araujo D E Wnkowski
CM and Nagle RB (1990) Tumourieenicity in human melanoma cell lines
controlled by introduction of human chromosome 6. Science 247: 568-571
W alker G. Palmer JM. Walters MK. Nancaxrow DJ. Parsons PG and Hayward NK
(1994) Simple tandem repeat alletic deietions confirm the preferential loss of
distal chromosome 6q in melanoma. Int J Cancer 58: 203-206
W'eber JL and May PE (1989) Abundant class of human DNA polymorphism which
can be typed using the polymerase chain reaction. Am J Hum Genet 44:
338-396
Yamada H. Wake N. Fujimoto S. Barren JC and Oshimura M ( I990» Multiple
chromosomes carrying tumour suppressor activity for a uterine endometrial
carcinoma cell line identified by rrucroceil-mediated chromosome transfer.
Oncogene S\ 1141-1147
British Journal of Cancer (1997) 75(9). 1324-1329
Bntisn Journal o l Cancer (1997) 00(0). 000-000
O 1997 C ancer R esearch Campaign
L o s s o f h e te r o z y g o s ity a t t h e m a n n o s e 6 -p h o sp h a te
in su lin -lik e growth fa c to r 2 r e c e p t o r g e n e c o r r e la t e s
w i t h poor differentiation in ea rly b r e a st c a r c in o m a s
SA C hap pell1, T Walsh2, RA Walker2 and JA Shaw 2
'D epartm ent of Biochemistry, University of Leicester. Adrian Building, University Road, Leicester, LE1 7RH, US; *Breast Cancer Research Unit. Department of
Pathology. University of Leicester. Clinical Sciences Building. Glenfield G eneral Hospital. Groby Road. Leicester LE3 9QP. UK
Summary Chromosome 6q has been shown to be one of the most frequent sites for allelic loss in human breast cancer. The mannose 6phosphate/insulin-like growth factor 2 receptor ( IGF2R) gene, which maps to chromosome 6q26-27, functions in the activation of TGF-{51. a
potent growth inhibitor for most cell types, the degradation of the mitogen IGF2 and the intracellular trafficking of lysosomal enzymes. Loss of
heterozygosity (LOH) at the IGF2R locus with mutations in the remaining allele have been reported in liver cancers and recently in two highgrade c a se s of ductal carcinoma in situ of the breast We have sought to confirm that allelic loss of IGF2R is an early event in the aetiology of
breast cancer by screening a group of ‘early1lesions for LOH at a polymorphic microsateilite marker within the IGF2R gene using polymerase
chain reaction (PCR). Several microdissected tumour foci were analysed for each of 40 mammographically detected invasive carcinomas and
22 ca se s of pure ductal carcinoma in situ (DCIS). None of 25 (62.5%) informative early invasive carcinomas showed any evidence of LOH.
This group comprised predominantly of well- to moderately differentiated ca se s (95%). However, 4 out of 18 informative DCIS ca ses (22%)
showed clear evidence of LOH. Three of these were poorly differentated (high-grade) lesions. These data suggest that loss of heterozygosity
at the IGF2R gene is associated with poor differentiation at this early stage of breast cancer development and progression.
Keywords: loss of heterozygosity; microdissection; mannose 6-phosphate/insulin-like growth factor 2 receptor; breast carcinoma
E xperim ental evidence suggest that paracrine interactions between
strom al and epithelial cells influence the growth and malignant
behaviour of breast cancers (Singer et al. 1995). The use o f both in
situ hybridization (Paik, 1992) and immunohistochemistry (Ellis
et al. 1994) have demonstrated that insulin-like growth factor 2
(IG F2) is expressed by fibroblasts in both benign and malignant
breast lesions. IGF2 is a potent mitogen for a number of breast
cancer epithelial cell lines in vitro and it is thought to exert its
m itogenic effect primarily through the high affinity insulin-like
grow th factor 1 receptor (IGF1R). In contrast, binding of IGF2 to
the m annose 6-phosphate/insulin-like growth factor 2 receptor
(IG F2R ) results in internalization and subsequent degradation o f
the ligand, making it unavailable to activate IGF1R. In addition,
the activation of TGF-fJl, a potent growth inhibitor of epithelial
cells, is dependent on binding of the TGF-01 latent complex to
IG F2R (D ennis and Rifkin. 1991; Komfeld. 1992). Thus
effectively operates as a growth-suppressor gene by antagonizing
the grow th stimulatory effect of IGF2 and activating the growthinhibitory effect o f T G F-pl.
mRNA has been detected in breast cancer cell lines and
tissue (De Leon et al, 1988; Cullen et al. 1990). In situ hybridiza­
tion analyses of breast tumour biopsies identified a higher level of
expression in carcinomas than in the corresponding benign
epithelium or stroma (Zhoa et al. 1993) that would not support a
suppressor role for IGF2R. Comparisons of
RNA levels
1GF2R
IGF2R
IGF2R
betw een tumour and non-tumour breast tissue using northern
analysis have demonstrated expression in all tissue tested with no
significant differences in the level of expression between tumour
and non-tumour tissue (Hubert et al, 1994). Analysis of the
gene copy number in this same tumour group, showed no amplifi­
cation of the gene whatever the clinical presentation o f the tumour
and irrespective of a concomitant amplification of c - e r b B 2 or i n t - 2
genes in several tumours. These data might reflect groups of
tum ours that do not involve
inactivation in their aetiology.
T he
gene has been mapped to chromosome 6q26-27
(Laureys et al, 1988). Allelic loss at this region has been observed
previously in several tumour types, including ovarian carcinomas
(Rodabaugh et al, 1995), malignant melanomas (M illikin et al,
1991), renal cell carcinomas (Morita et al, 1991), small-cell lung
cancers (Merio et al, 1994), T-cell acute lymphocytic leukaemias
(M enasce et al, 1994) and breast carcinomas (Devilee et al. 1991;
Orphanos et al, 1995). This shared region of allelic loss may
reflect the involvement of putative tumour-suppressor genes that
are pleiotrophic for these tumours. Detailed studies of chrom o­
som e 6q in breast cancer have highlighted two regions (6 ql3 and
6q26-27) that show high levels of loss o f heterozygosity (LOH)
and indicate the presence of at least two tumour-suppressor genes
(Devilee et al. 1991; Orphanos et al, 1995).
De Souza et al (1995
first demonstrated frequent LOH at
the
locus in human hepatocellular tumours and identified
point mutations in the remaining allele in 25% o f these cases,
strongly suggesting that the
gene functions as a tumoursuppressor gene in human liver carcinogenesis. Recently, the same
group also reported LOH for 12 out of 40 breast tumours studied
(Hankins et al. 1996). No clinical information was provided for the
7 out of 21 informative invasive cases that showed LOH. Five
IGF2R
1GF2R
IGF2R
IGF2R
aj?)
IGF2R
R e c e iv e d 2 5
R e v is e d
A c c e d e d
A p n l 1996
19 M a y
5
J u n e
C o rr e so o n c e n c e
1997
1997
to :
JA Shaw
1
2 SA Chappell el al
ductal carcinoma in situ (DCIS) cases that showed LOH were
screened for mutations in the remaining allele. Two of these, both
comedo-type (high grade) cases, showed missense mutations
(Hankins et al. 1996) supporting the hypothesis that
allelic
loss may be an early event in the aetiology of some breast cancers.
Small, mammographically-detected breast cancers form a useful
group for study o f the involvement of tumour-suppressor genes in
the developm ent and earlier stages of progression of breast cancer.
We have previously identified frequent LOH at 6q25-27 in a group
of ‘early’ invasive carcinomas and preinvasive cases of DCIS
(Chappell et al, 1997), confirming distal chromosome 6q as a major
site for genetic change in the early stages of development of some
sporadic breast cancers. The purpose of this study was to investigate
whether LOH occurs as frequently at the candidate tumoursuppressor gene
in these early tumours, and if so. if there is
any correlation with tumour type. We analysed a highly informative
dinucleotide repeat/tetranucleotide deletion/insertion polymorphism
(Hoi et al. 1992) within the 3' untranslated region of the
gene
in m ultiple tum our foci prepared by microdissection for each of 40
‘early’ invasive carcinomas and 22 cases of pure DCIS.
IGF2R
1GF2R
1GF2R
MATERIALS A N D M E T H O D S
P atients
A total o f 40 invasive breast carcinomas that were impalpable and
detected by mammography were studied. All were from the preva­
lent round o f screening and were detected by the Leicestershire
Breast Screening Service. Cases of 15 mm or less were examined.
All had either axillary node sampling or axillary dissection. None
o f the tum ours were from women with either a strong family
history o f breast cancer or any known inherited predisposition to
the developm ent of tumours. All but two were well- or m oderately
differentiated and all were node negative. A total of 35 were infil­
trating ductal carcinomas with the remainder comprising three
tubular carcinom as and two infiltrating lobular carcinomas.
A total o f 22 cases o f pure DCIS were studied. These com prised
ten high-grade, three intermediate-grade and nine low-grade cases.
T is s u e s an d h istology
All tissues were fixed in 4% formaldehyde in saline for 18-36 h.
A fter a review of haematoxylin and eosin stained sections, repre­
sentative blocks were chosen for further study. All carcinom as
were reported according to the NHS Breast Screening Program m e
N ational C oordinating Group for Breast Screening Pathology
G uidelines (1995). Infiltrating ductal carcinomas were graded
using the modified Bloom and Richardson system (Elston and
Ellis, 1991). Cases of DCIS were graded as low-, interm ediate- or
high-nuclear grade. All histology was undertaken by RA Walker.
DNA extraction and m icrodissection from paraffin
em b ed d ed s e c tio n s
Form alin-fixed, paraffin-embedded tissue from breast tum our
sam ples and non-involved lymph nodes or normal breast served as
the source o f tumour and normal DNA respectively. For each
tum our-norm al pair. DNA was extracted from non-tumour tissue
and m icrodissected tumour foci prepared from lO jim paraffinem bedded sections as described previously (Shaw et al. 1996;
Chappell et al. 1997).
British Journal of Cancer (1997) 00(0), 000-000
Table 1 Clinicopathological features of early breast cancers showing
alterations at I G F 2 R
C ase
Type
Grade
A lteration
D2
D4
D13
D14
55
DCIS
DCIS
DCIS
DCIS
IDC
High
Low
High
High
II
LOH
LOH
LOH
LOH
MSI
DCIS. ductal carcinoma in situ; IDC. infiltrating ductal carcinoma; LOH, loss
of heterozygosity; MSI, microsateilite instability.
PCR an alysis at IGF2R
PCR reaction components were as follows; 45 m M Tris hydrochloric
acid, pH 8.8; 11 mM ammonium sulphate; 4.5 m M magnesium
chloride, 200 jim dTTP; dCTP; dGTP; 25 pM dATP (Pharmacia,
UK); 0.3 pi [a-3SS]deoxyadenosine-5'-triphosphate (600 Ci mmol-1,
10 mCi ml*1(ICN Pharmaceuticals, UK); 113 pg ml*1 bovine serum
album in (Boehringher Mannheim); 6.7 mM (3-mercaptoethanol;
4.4 pM EDTA, pH 8.0; 10 pmol of both the forward (GTA TCA
TG A GAA CCT GAA GAG) and the reverse primer (TTG CCG
G C T G GT GAA TTC AA) (Hoi et al, 1992); 100 ng of DNA or 2 pi
o f microdissected DNA and 1 unit of
DNA polymerase (Gibco
BRL. UK) in a total volume of 25 pi. Hot-start PCR was carried out
using the following: 5 min denaturation at 94°C, followed by 30
cycles o f 1 min denaturation at 94°C, 1 min annealing at 65°C, and 1
min extension at 72°C with a final extension of 7 min at 72°C on a
DNA Thermal Cycler (Perkin Elmer Cetus, UK). Analysis o f PCR
products and interpretation of LOH were as described previously
(Shaw et al, 1996, Chappell et al, 1997).
Tcuj
RESULTS
We have analysed 62 ‘early’ breast carcinomas com prising 40
invasive carcinomas and 22 cases o f pure DCIS for LOH at
Because of the complex heterogeneity o f the disease and
the presence of non-tumour cells, we carried out LOH analysis on
DNA extracted from different foci that had been microdissected
from within the same tumour tissue section. Cases that were
homozygous at the
polymorphic repeat (Figure 1A) were
considered uninformative. The frequency of heterozygosity varied
betw een the two groups: 25 of the ‘early’ invasive cases (62.5% )
and 18 of the 22 preinvasive cases o f DCIS (82%) were inform a­
tive. O f the total 43 cases (69%) that were informative five show ed
alterations at
The clinicopathological features o f these
cases are summarized in Table 1.
The group of ‘early’ invasive carcinomas were predom inantly
well- or moderately differentiated cases (95% grade 1 or grade 2).
None o f these tumours showed any evidence o f LOH at
,
although one cases showed clear evidence o f microsateilite insta­
bility (Figure IB). In contrast, 4 of the 18 informative DCIS cases
(22% ) showed clear evidence of LOH. For example. Figure 1C
shows DCIS cases 2 that exhibited loss of both the upper and lower
allele in separate microdissected tumour ducts. Figure 1D shows
DCIS case 4 with clear loss of the upper allele in all ducts exam ­
ined. Three of the four cases of DCIS with LOH were high nucleargrade cases, and one case was low grade (Table 1). suggesting an
association between LOH at
and poor differentiation in the
early stages of breast cancer development and progression.
IGF2R.
1GF2R
1GF2R.
IGF2R
IGF2R
© Cancer Research Campaign 1997
LOH at IGF2R in early breast cancers
3
IGF2R
The
gene was first identified as a tum our-suppressor
gene in hepatocellular tumours (De Souza et al, 19956) and the
presence of LOH at the
locus in adenomas suggests that
inactivation may be an early event in liver carcinogenesis. The
data reported in this paper, taken together with that o f Hankins et al
(1996), would support a similar early involvement o f
inac­
tivation in certain pathways in the development and progression of
breast cancer. Additional evidence, suggesting a role for the
gene in mammary carcinogenesis, comes from two other
key investigations. Jirtle et al (1993) first demonstrated that
steady-state
mRNA levels in rat mammary tumours,
regressing in response to d-limonene, increased twofold when
compared with untreated tumours and that in unresponsive
tumours expression o f
was unaltered. More recently, Ellis
et al (1996), have shown that the affinity o f IGF2 for
inhibits IGF2 activity in MCF-7 breast cancer cells. Cellular
proliferation, receptor tyrosine kinase-dependent signalling and
extracellular IGF2 protein accumulation were all reduced specifi­
cally in the presence of IGF2R affinity. Therefore, by operating as
an IGF2 anatagonist the
gene has tum our suppressor-like
properties.
The lack o f detection of LOH in the early invasive group o f
cases does not appear to result from technical problem s in inter­
pretation o f LOH data (e.g. masking of any lost alleles by contam ­
inating non-tumour material) as the same DNA sam ples prepared
by microdissection from these carcinomas show clear LOH with
three other microsateilite markers (ESR, D6S186, D 6S193) that
map to 6q25.1-q27 (Chappell et al. 1997). Moreover, the group of^
DCIS cases has previously been studied and also show ed more
frequent LOH (50%) with the three other markers. The frequent
LOH detected within 6q25.1-27 could therefore indicate the
critical inactivation o f other unknown tum our-suppressor genes
within this chromosomal interval (Chappell et al, 1997).
One factor that might interfere with the detection o f LOH in this
study is when polymerase amplification of dinucleotide repeats
produces slippage bands below the true allele (Louis et al, 1992).
Given that for a high proportion o f the cases heterozygous for the
dinucleotide repeat, the two alleles differed by only 2 bp in
length, any slippage bands would tend to mask loss o f the smaller
allele and hence reduce the true frequency o f allelic loss. However,
one of our DCIS cases that showed LOH (Figure 1C) had clearly
lost both the upper and lower allele in different microdissected
tumour ducts. Other mechanisms that do not involve LOH but also
lead to inactivation o f tumour-suppressor genes might also be
critical for inactivation o f
in breast carcinogenesis. For*
example, aberrant hypermethylation of 5' CpG islands within
proximal prom oter regions has been implicated as a mechanism by
w hich tumour-suppressor genes can be inactivated. This has been
dem onstrated for
(Graff et al. 1995) and for the
and
tumour-suppressor genes (Herman et al, 1994; M erlo et al,
1995). Therefore, it would be of interest to investigate the CpG
island methylation status within the 5' regulatory region of the
gene.
O ur study has been concerned with breast cancers at an ‘early’
stage: small, node-negative invasive cases that have features asso­
ciated with a good prognosis appear to show no evidence of LOH
at
whereas high-grade cases of DCIS although at a pre­
invasive stage show evidence of LOH. These data provide good
evidence that
acts as a tumour-suppressor gene in the
development o f some early breast cancers associated with a more
aggressive disease type.
IGF2R
IGF2R
1GF2R
IGF2R
IGF2R
IGF2R
IGF2R
Figure 1 Representative examples of LOH at the IGF2R locus in 'early1
breast cancers. N. normal tissue; T, tumour tissue. (A) Two non-informative
invasive carcinoma patients; (B) informative eany invasive carcinoma showing
microsateilite instability; (C and D) DCIS cases 2 and 4. respectively, with LOH
DISCUSSION
In this study we examined ‘early’ breast carcinomas for loss o f
heterozygosity at a polymorphic microsateilite locus within the 3 '
untranslated region o f the
gene on chrom osome 6q26-27.
DNA extracted from microdissected tum our foci prepared from
10 urn paraffin-em bedded tumour tissue sections were exam ined
separately. We found a difference in the frequency o f LOH
betw een the two groups o f early lesions studied. None o f the 40
early invasive carcinomas showed any evidence o f allele loss,
although one case showed microsateilite instability. This case has
been reported previously and shows instability at nine out o f ten
o ther loci tested (Shaw et al, 1996). The group of ‘early’ invasive
carcinom as were predominantly grade 1 or grade 2 cases (95% ),
suggesting no involvement or inactivation o f
in well- to
m oderately differentiated tumours. In contrast, 4 out o f 18 infor­
m ative cases of DCIS showed clear evidence o f LOH. Although
the DCIS lesions represent a pre-invasive stage o f breast cancer.
10 out o f the 22 cases examined were high nuclear grade, which is
recognized as a more aggressive form o f the disease (Lagios,
1990). T hree of the four DCIS tumours that showed LOH were
from the ten high-grade cases, suggesting an association between
inactivation o f
and poorly differentiated in situ lesions.
T hese data support the findings o f Hankins et al (1996), w ho
reported LOH with missense mutations in the remaining allele in
two com edo-type (high-grade) DCIS cases. In combination, these
data suggest that inactivation of
may occur only w ithin
certain more aggressive subgroups (poorly differentiated cases) o f
these ‘early ’ breast cancers. Unfortunately, we were not able to
undertake sequence analysis of the remaining allele in those cases
that show ed LOH because o f the paucity of available D NA .
prepared by microdissection, for analysis o f the large
gene.
IGF2R
IGF2R
IGF2R
IGF2R
IGF2R
Cancer Research Campaign 1997
IGF2R
1GF2R
pi6
E-cadherin
VHL
IGF2R
IGF2R,
IGF2R
British Journal of Cancer (1997) 00(0), 000-000
4
SA Chappell et al
ACKNOWLEDGEMENTS
Tom Walsh is a PhD student supported by the Department o f
Pathology and the University o f Leicester. This work was
supported by funding from the Glenfield Hospital Research
Comm ittee.
REFERENCES
Chappell SA. Walsh T. Walker RA and Shaw JA (1997) Loss of heterozygosity at
chromosome 6q in preinvasive and early invasive breast carcinomas. Br J
Cancer IS. 1324-1329
Cullen KJ. Yee D. Sly WS. Pardue J. Hampton B. Lippman ME and Rosen N (1990)
Insulin-like growth factor expression and function in human breast cancer.
Cancer Res 50: 48-53
De Leon DD. Bakker B. Wilson DM. Hitz RL and Rosenfeld RG (1988)
Demonstration o f insulin-like growth factor (1GF-I and -Q) receptors and
binding protein in human breast cancer cell lines. Biochem Biophys Res
Common 152. 398-105
Dennis PA and Rifkin DB (1991) Cellular activation o f latent transforming growth
factor 0 requires binding to the canon-independent mannose 6phosphate/insulin-like growth factor type □ receptor. Proc Nail Acad Sci USA
88: 580-584
De Souza AT. Hankins GR. Washington MK. Fine RL, Orton TC and Jirtle RL
(1995c) Frequent loss o f heterozygosity on 6q at the mannose 6phosphate/insulm-like growth factor □ locus in human hepatocellular tumours.
Oncogene 10: 1725-1729
De Souza AT. Hankins GR. Washington MK, Orton TC and Jirtle RL (19955).
M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of
heterozygosity. Nature Genet 11: 447—149
Devilee P. van Vliet M. van Sloun P. Kuipers Dijkshoorn N, Hermans J, Pearson PL
and Comelisse CJ (1991) Allelotype of human breast carcinoma: a second
major site for loss of heterozygosity is on chromosome 6q. Oncogene 6:
1705-1711
Ellis MJC. Singer C. Hornby A. Rasmussen A and Cullen KJ (1994) Insulin-like
growth factor mediated stromal-epithelial interactions in human breast cancer.
Breast Cancer Res Treat 31: 249-261
Ellis MJC. Leav BA. Yang Z. Rasmussen A. Pearce A. Zweibel JA. Lippman ME
and Cullen KJ (1996) Affinity for the insulin-Like Growth Factor II (IGF-U)
Receptor inhibits autocrine IGF-U activity in MCF-7 breast cancer cells. Mol
Endocrinol 10: 286-297
Elston CW and Ellis IO (1991) Pathological prognostic factors in breast cancer. L
The value o f histological grade in breast cancer experience from a large study
with long-term follow-up. Histopathology 19: 403-410
Graff JR. Herman JG. Lapidus RG. Chopra H. Xu R. Jarrard DF, Isaacs WB. Pitha
PM. Davidson NE and Baylin SB (1995) E-cadherin expression is silenced by
DNA hypermeihyladon in human breast and prostate carcinomas. Cancer Res
55 5195-5199
Hankins GR. De Souza AT. Bentley RC, Patel MR. Marks JR. Iglehan JD and Jirtle
RL (1996) M6P/1GF2 receptor, a candidate breast tumour suppressor gene.
Oncogene 12: 2003-2009
H ibert E. Herbelin C and Bougnoux P (1994) Analysis o f the IGF-II receptor gene
copy number in breast carcinoma. BrJ Cancer 69: 120-124
Herman JG. L a a f F. Weng Y. Lerman MI. Zbar B. Liu S. Samid D. Duan DSR.
Gnarra JR. Linehan WM and Baylm SB (1994) Silencing of the VHL
British Journal of Cancer (1997) 00(0), 000-000
tumour-suppressor gene by DNA meihylaiion in renal carcinoma. Proc Natl
Acad Sci USA 91: 9700-9704
Hoi FA. Geurds MP. Hamel BC and Man man ECM (1992) Improving the
polymorphism content of the 3' UTR of the human IGF2R gene. Hum Mol
Genet 1: 347
Jirtle R L Haag JD. Ariazi EA and Gould MN (1993) Increased mannose
phosphate/insulin-like growth factor II receptor and transforming growth factor
^1 levels dunng monoterpene-induced regression of mammary tumours.
Cancer Res S3: 3849-3852
Kornfeld S (1992) Structure and function of the mannose 6-phosphate/insulin like
growth factor II receptor. Anno Rev Biochem 61: 307-330
Lagios MD (1990) Duct carcinoma in situ: pathology and treatment. Surg Clin N Am
70: 853-871
Laureys G. Barton DE. Ullrich A and Francke U (1988) Chromosomal mapping of
the gene for the type II insulin-like growth factor receptor/cation-independent
mannose 6-phosphate receptor in man and mouse. Genomics 3: 224-229
Louis DN. von Deimling A and Seizinger BR (1992) A (CA)mdinucleotide repeat
assay for evaluating loss of allelic heterozygosity in small and archival human
brain tumour specimens. Am J Pathol 141: 777-782
Menasce LP. Orphanous V. Sanobanez-Koref M, Boyle JM and Harrison CJ (1994)
Common region of deledon on the long arm of chromosome 6 in nonHodgkin's lymphoma and acute lymphoblasbc leukaemia. Genes Chrom
Cancer 10: 286-288
Merlo A. Gabrielson E. Mabry M. Vollmer R. Baylin SB and Sidransky D (1994)
Homozygous deledon on chromosome 9p and loss of heterozygosity on 9q,
6p. and 6q in primary human small cell lung cancer. Cancer Res 54:
2322-2326
Merlo A. Herman JG, Mao L Lee DJ, Gabrielson E. Burger PC, Baylin SB and
Sidransky D (1995) 5' CpG island methylation is associated with
transcnpoonal silencing of the tumour suppressor p 16/CDKN2/MTS1 in
human cancers. Nature Med 1: 686-692
Millikin D. Meese E, Vogelstein B. Witkowski C and Trent J (1991) Loss o f
heterozygosity for loci on the long arm of chromosome 6 in human malignant
melanoma. Cancer Res 51: 5449-5453
M orita R. Saito S. Ishikawa J. Ogawa O, Yoshida 0 . Yamakawa K and Nakamra Y
(1991) Common regions of deledon on chromosome 5q. 6q, and lOq in renal
cell carcinoma. Cancer Res 51: 5817-5820
National Coordinating Group for Breast Screening Pathology (1995) Pathology
Reporting in Breast Cancer Screening, 2nd edn. NHSBSP: Sheffield. UK
Orphanos V. McGown G, Hey Y. Boyle JM and Santibanez-Koref M (1995)
Proximal 6q, a region showing allele loss in primary breast cancer. Br J Cancer
71: 290-293
Paik S (1992) Expression of IGF-I and IGF-U rnRNA in breast tissue. Breast Cancer
Res Treat 22: 31-38
Rodabaugh KJ. Blanchard G. Welch WR, Bell DA. Berkowitz RS and M ok SC
(1995)
Detailed deletion mapping of chromosome 6q in borderline epithelial
ovarian tumors. Cancer Res 55: 2169-2172
Singer C, Rasmussen A. Smith HS, Lippman ME. Lynch HT and Cullen KJ (1995)
Malignant breast epithelium selects for insulin-like growth factor U expression
in breast stroma: evidence of paracrine function. Cancer Res 55: 2448-2454
Shaw JA. Walsh T. Chappell SA. Carey N. Johnson K and Walker RA (1996)
Microsateilite instability in early sporadic breast cancer. B rJ Cancer 73:
1393-1397
Zhoa Y. Escot C, Maudelonde T. Puech C. Rouanet P and Rochefort H (1993)
Correlation between mannose-6-phosphate/IGFU receptor and cathepsin D
RNA levels by in situ hybridization in benign and malignant mammary
tumours. Cancer Res 53: 2901-2905
<£>Cancer Research Campaign 1997
S r r t i s h Journal of Cancar (1996) 73, 1 3 5 3 -1 3 9 7
D 1 5 9 5 S to c k to n P re ss All n g n ts reservec ICG7 - 0 9 2 0 /9 6 312.CO
M i c r o s a t e i l i t e i n s t a b i l i t y in e a r ly s p o r a d i c b r e a s t c a n c e r
JA S h a w 1, T W alsh1, SA Chappell1, N Carey2, K J o h n so n 3 and RA Walker
'Brass: Cancer Research Unit. Department of Pathology. University of Leicester, Clinical Sciences. GlenfieId General Hospital.
Groby Road. Leicester LE3 90P. UK: departments of Biochemistry and Surgery. Charing Cross and Westminster Medical School.
Fulham Palace Road. London W6 8RF. UK: laboratory of Genetics. University of Glasgow. Pontecorvo Building. Anderson
College. 56 Dumbarton Road. Glasgow Gil 6HU. UK.
We have studied the incidence o f microsateilite instability at three trinucleotide repeats and seven
dinucleotide repeats from five chromosomal regions, in a group of 30 mammographicaily detected ‘early’
invasive breast cancers and correlated its occurrence with clinicopatholoccai parameters. The myotonic
dystrophy (DM-1) trinucieoude repeat was analysed in 48 additional cases. In 4 out of 78 (5%) paired
tumour - normal DNA samples we found evidence o f somatic microsateilite instability at DM-l: a novel allele
of a different size was seen in the tum our D N A which was not present in die normal DNA sample. All four
tumours that showed evidence of instability were from the core group of 30 cases (13%) and were well or
moderately differentiated, oestrogen receptor-positive, inriltrating ductal carcinomas. Two of these tumours
were unstable at nine of ten loci studied, b o th trinucleotide and dinucieouce repeats. DNA prepared from
different normal tissues showed no evidence o f instability, for all four instability cases. These data indicate that
microsateilite instability is specific to the ru m our D NA and is an early event :n the genesis of some sporadic
breast cancers.
Sum m ary
Keywords: breast carcinoma;
m a m m o g ra p h y ;
B re a s t c a n c e r is a heterogenous disease, b o th clinically and
w ith r e g a r d to th e genetic aicerattons involved in tum origenesis. H e n c e , m u ltip le som atic and in d en ted genetic changes
t h a t le a d to loss o f g ro w th controi may co n trib u te to the
d e v e lo p m e n t o f b re a st cancer. Despite notab le recent
a d v a n c e s , w ith the clon in g o f
(M iki
1994;
F u tr e a i
1994) a n d m apping o f
(W o o ster
1 9 9 4 aj. th e r e is no c lea r understanding o f the n a tu ra l history
«af th e d is e a s e . T h is c o n tra sts with colorectai carcin o m a where
e x te n s iv e stu d ie s have identified a benign to m alignant
p r o g r e s s io n w ith recognisable m olecuiar changes, frequently
p o m : m u t a ti o n s th a t involve proto-oncogene an d tum ours u p p r e s s o r gene loci (F e aro n and Vogeistein. 1990).
R e c e n tly , a novel alterauon based on D N A repeat
m is a lig n m e n t m utag en esis has been described (A aito nen
1993: T h ib o d e a u
1993: Ionov
1993). This
ty p e o f m u ta g e n e s is occurs in microsateilite D N A sequences
in w h ic h o n e - to six-nucleotide motifs are tanaem iv repeated
a n d a r e o fte n highly polym orphic (W eber and M ay. 1989).
M o n o - , d i- a n d trinucleotide repeats are u nstable in
h e r e d i ta r y n o n -p o ly p o sis coiorectal cancer (H N P C C ) as well
as s p o r a d ic c o lo re c ta l cancer ceils. G erm iine m u tatio n s in
f o u r D N A m ism atch repair genes, including
on
c h r o m o s o m e 2 p 2 ! - 2 2 . have been im plicated as the cause o f
the h e r e d ita r y non -o o iy p o sis syndrome and the associated
m ic ro s a te ilite in sta b ility (Fishet
1993: Leach
1993
M ic ro s a te ilite instability may renect defective function
o f D N A m ism a tc h rep a ir genes and be m anifested when both
c o p ie s o f a m ism atch rep air gene are inactivated (P arsons
9 9 3 •.
I f s im ila r m ism atch repair defects are m voived in the
reiu t:v».\ early stages o f breast cancer, then m icrosateilite
m s tu o ib ty -mould be fo u n d ;r. 'eariy' carcinom as an d ‘at risk’
ie sic r.s. In a p relim in ary >tudy we 'nave detected somatic
m ic r c s u te ilite in stab ility at the myotonic dystrophy iD M -U
a s s c c iu te d - C T G rep eat ;n 'eariy' mam mograDhicaiiy detected
b re a s t c a n c e rs iShaw
1995). We now report o u r
n n c .n a s fro m analysis o f ten polym orphic m arkers, three
t n r .- c .e c t i d e rep eats an d >c\en dinudeotido repeats in a
BRCAl
et al..
BRCA2
et al..
e: al..
al..
et al..
microsateilite instability
g ro u p o f 30 ’early ’ sporadic breast cancers together w ith the
a n a ly s is o f D M -l in a totai o f "3 cases. Tne m arkers analysed
m a o to five ch ro m o so m ai restens: o d (SC A -l). 6q (E R T A
a n d D 6 S 1 9 3 ), I6q (D16S2S9. D ’:oS400. D16S402. D16S413),
19a iD M - l a n d X75b) and X c AR). The oestrogen receptor
m a p s to 6q25 (M enasce
.993) and D6S193 m aps to
6 c 2 “ - S a ito
1992). Tne enromosome 16q m arkers span
o v e r 50 c M (W eissenbach
1992) and D M -l and X 75b
re s id e w ith in 90 kb o f each other (Jansen
1992). We
h a v e an a ly se d the frequency ana type o f microsateilite
in s ta b ility a n d co rrelated these data with clinicopathological
fin d in g s. Since the m arkers sreeled are highly polym orphic
c o n c u r r e n t assessm ent o f aileiic loss was also possible.
et al..
et at..
et ai..
et al..
et
et al..
hMSH2
et al..
et al..
et
al..
et ai..
M a te r ia ls and methods
Patterns
A to ta i o f 78 invasive breast carcinomas which were
im p a lp a b le an d detected by mammography were studied.
A il w e re from the prevalent round o f screening and were
d e te c te d by the Leicestershire 3r;ast Screening Service. Cases
'.5 m m o r iess in maxim um diameter were examined. Ail had
e ith e r a x iilary node sampling :r lxiilary dissection. N one o f
th e tu m o u rs were from women with a strong family history
o f b r e a s t cancer.
Tissues
A ll tissu e s were fixed in 4% formaldehyde :n saiine for 1 3 36 h. A fte r slicing, selected blocks were processed th ro u g h
g r a c e d uico h o is and xylene to paraffin wax. Following review
h a e m a to x y lin an d eosin stained sections representative
b lo c k s w ere chosen for further study. A dditional norm al
f.a su e fro m hysterectom y specimens were retrieved from the
p a th o lo g y files o f the Leicester Royai Infirmary. These were
s a m p le d a n d processed at different time periods to the
o n g m a i tu m o u r m aterial.
z:
Histnhtgy
C u r";
r ' “. c e n c s :
3 o n e r:
L a :..-
a*
JA
Shaw .
D e p a rtm e n t o f
K iin a in c k
R o y u i In tirm a ry .
Pathology.
U n iv e rs ity o f
O m e a i S c ie n c e s B u n d in g . P O
LK.
I Decamrer ivlu5: accepted 4
L c tc o tc r L E I ” L .\.
~ March i9*#5: rcMvu
B ox b e .
The
c a rc in o m a s were reported according to the Royal
o f
Pathologists w o r k . rig party guidelines ( 1 9 9 0 ) .
in f iltra tin g d u ctal carcinom as were graded using the m odified
C o lle g e
3 !c<
:r.
an d
R ic h a rd s o n
s y s te m
E ls to n
and
E llis .
1 4 9 1 ).
A ll
Microsateilite instability in breast cancer
JA Shaw et al
h is to l o g y w as u n d e rta k e n by RAW . The clinicopathological
f e a t u r e s a re sh o w n in T ab le I.
O e s tr o g e n recepcor w as determ ined im m unohistochem icaily
u s in g a n tig e n retrie v al a n d ID 5 m onoclonal an tib o d y (D ako)
( R a j a k a n a r a n d W a lk e r, 1995).
se c tio n s w ere dew axed and renydrated by sequential addition,
m ix in g a n d rem oval o f 2 x 1 ml xylene, 2 x I ml 99% ethanol
a n d 2 x 1 m l 95% ethanol. Air-dried pellets were resuspended
in 250 /il p ro tein ase K solution (I mg m l-1 in 50 mM Tris
H C i. p H S.0, 1% sodium dodecyl sulphate), and incubated
o v e rn ig h t a t 37°C. Sampies were then extracted twice with
p h e n o l- c h lo r o f o r m , precipitated w ith ethanol and resus­
p e n d e d in distilled w ater.
D S A extraction
PCR analysis
F o r m a iin - iix e d . pararfin-em bedded tissue from b reast tum our
s a m p ie s a n d n o n -in v o lv e d lym ph nodes served as the sources
o f t u m o u r a n d n o rm a l D N A respectively. F o r each sam ple.
D N A w a s e x tr a c te d from 7 um paraffin-em bedded tissue
s e c tio n s o r m a te ria l p repared by m icrodissection. Briefly.
M ic ro sa te ilite repeats were analysed by polym erase chain
re a c tio n (P C R ). P rim er pairs and amplification conditions
w e re as describ ed in previous reports. Trinucleotide repeats
c o m p ris e d : D M -l (B rook
1992), SC A -l (O rr
1995') a n d A R (L a S pada
1991). D inucleotide repeats
w ere: X 7 5 b (Jansen
1992), a (TA )n repeat in the
u c s tr e a m region o f the human oestrogen receptor gene
lE R T A ) (D el Senno
1992), D6S193 (Saito
19921. D 16S289 (Shen
1992), D16S400, D16S402 and
D 1 6 S 4 1 3 (W eissenbach
1992). T he PC R products were
la b e lle d by the ad d itio n o f 3 uCi o f [<x-3,S]dATP to the
re a c tio n . T h e labelled PC R products were eiectropnoresed
th r o u g h d en a tu rin g 6% polyacrylamide gels a t 70 W for 1 3 h d e p e n d in g on the fragment size. Gels were dried and
e x p o s e d to radiographic film for 1 - 4 days. C om parison o f
th e m ig ra tio n o f alleles from paired norm al and tum our
D N A sam ples th a t showed the appearance o f alleies o f
a lte re d len g th in tu m o u r DNA served to indicate microsa te iiite instability. Where instability was detected, the
a n a ly s e s were repeated using freshly prepared D N A using
a d ja c e n t sections prepared from the paraffin blocks. Allele
sizes w ere estim ated by comparison w ith a M 13m p 18 D N A
seq u e n c e ladder.
Oestrogen receptor
T able I
Clinicopathological features of 73 ‘early- sporadic breast
cancers
Grade
Type
Sumoer of Tumour Sumoer of
cases sice mm) cases
<10
15
13
T ub
n
17
10
Lob* T u b
4
3
11
Idc. lie
3
12
3
lie
6
13
I
22 (2)
Idc
14
3
31 (1)
II
Idc
2<
4
15
III
Idc
73
78
T o tai
T u o . tu o u iar carcinoma; Lob/Tuo. ioouiar ana tubular carcinoma;
Idc. lie . infiltrating ductal with infiltrating Ioouiar carcinoma; lie.
infiltrating oouiar carcinoma; Idc. infiitraung ductal carcinoma;
nu m b ers ;n brackets, node-positive cases.
TNBECMO
T N
et al.,
et al..
et al..
et al..
et al..
et al..
TN
et al.,
et al.,
t n
s a m p l e s f r o m p a i r e d r .c r m a l l y m p h n o d e ( N l a n d
I
M i c r o s i l d l i t e in sta b ility in e ariy b r e a s t c a n c e r. G e r .o — c D \ A
o r u :> 'c :c :c d tu m o u r i T i -a m o ie s a c r e c o m p a re d by P C ? . - r r . r d c a t i o n . e l e c t r o p h o r e s i s o n t*1-# s e q u e n c i n g e e ls a n d
: c o .s e 2 w i t h a n a l y s i s o f > i\ n o r m a l t is s u e s : l y m p n n o d e
: * . c io ji r a n h v m i E x p a n sio n a t D M -! in m tc ro u is s c c te d t u t r o u r
- ' r - i .i : " r e a 't <B». e n d o m e triu m <E'. cerv ix l O . rrty o e n d o ir.e ir:.. — '■■I1 a n d o v a r y t O ) . (bi C o n t r a c t i o n a t X 7 5 b in c a s e I. ic )
■ ;r . - . c t i o n a t S C A - l i n c a s e I C c - e n u m b e r s r e f e r t o T a b l e II.
• - ..c t i o n a l \ R in c ase 2. dl C o n tra c tio n a t D M - l in c a s e - e '
- s u e D N A i n d i c a t i n g s o m e ; c - n ic r o s a te ilite in s ta b ility .
in d ic a te a lte re d le n g th aileics in a m o u r c o m p a re d w ith - o r r
y jr e
:
M icro sateilite instability
JA Sbaw er 3l
inbreast cancer
1395
R e s u lts
D isc u s sio n
T e n p o ly m o rp h ic m icrosateilite markers, three trinucleotide
re p e a ts a n d seven dinucleo tid e repeats, from five ch ro m o so ­
m a l re g io n s w ere am plified from 30 tu m o u r-n o rm a l D N A
p a ir s u s in g th e P C R . T h e D M -l (CTG) repeat was analysed
th r o u g h - 8 a d d itio n a l tu m o u r-n o rm a l D N A pairs.
T h e a p p e a ra n c e o f alleles o f altered length in tu m o u r
D N A in d ic a te d an a lte ra tio n in microsateilite size (F ig u re I
a n d T a b le II). M icrosateilite instability was m axim ally
d e te c te d a t th e D M -l trinucleotide repeat in fo u r o f 78
(5 % ) tu m o u r s . T w o o f these four tumours show ed instability
a t n in e lo c i, b o th tnnucleotide and dinucieotide repeats.
T h e s e d a t a w e re rep licated firstly wtth freshly prep ared D N A
s a m p ie s fr o m a d ja c e n t sections from the paraffin blocks a n d
s e c o n d ly w ith D N A sam pies prepared by m icrodissection o f
s m a il a r e a s o f tu m o u r w ithin a section.
I n o r c e c to verify whether these D N A changes are
r e s tr ic te d to th e tu m o u r DNA, we next analysed D N A
p r e p a r e d fr o m o th e r n o rm al tissues for these instability cases.
D N A p r e p a r e d fro m umnvoived breast, en d o m etriu m a n d
ce rv ix s h o w e d n o evidence o f microsateilite instability fo r all
fo u r c a s e s . F ig u re la show s microsateilite instability a t D M -l
in t u m o u r 2. N o n e o f six normal tissues analysed (ly m p h
n o d e , a is to io g ic a ily n o rm al breast, en d o m etn u m . cervix,
m y c e n d c m e tn u m a n d ovary) showed any evidence o f
m ic ro s a te ilite in stab ility , suggesting that instability is indeed
sp ecific to d ie tu m o u r ceil population.
F o r th e D M - l rep eat, ail o f the novei alleles seen in
tu m o u r D N A lie w ithin the normal p o p u la tio n range,
a lt h o u g h th e new allele sizes differed by up to 16 re p e a t
u n its f r o m th e alleles seen in normal D NA . T hese d a ta have
b e e n c o n firm e d for tw o o f the cases by c io n in g a n d
s e q u e n c in g o f th e altered length aiieies (data n o t show n).
T h e in c id e n c e o f microsateilite instability in tu m o u r D N A
w as less fre q u e n t a t the two other tn n u cleo tid e rep eats
s tu d ie d . T w o o f the fo u r cases showing instability a t D M -l
s n o w e d in s ta b ility a t the SCA-l and A R repeats (e.g. F ig u re
ic a n d
T n e size o f the novel aiieies seen in these tu m o u rs
lies w ith in th e n o rm a l population range o f p o ly m o rp h ism s, as
fo r D M - l . T h e tw o tum ours that snowed instab ility a t ail
th re e tr in u c le o tid e repeats also showed instability a t six o f th e
sev en d in u c ie o tid e repeats. No instability was d etec ted a t the
D l 6 5 2 3 9 re p e a t in the core group of 30 tum ours studied.
T i e c lin ic o p a th o lo g ic a l features of the four cases sh ow ing
m ic ro s a te ilite in stab ility are listed in Tabie II. Ail were n o d en e g a tiv e in filtra tin g du ctal carcinomas, well o r m o d erately
d if f e r e n tia te d a n d oestrogen receptor (ER) positive (R a ja k a r .a r a n d W a lk e r. 1995). However, when we analysed D M -l
t h r o u g h a to ta i o f
cases, none of 53 o th e r in filtratin g
d u c ta : c a rc in o m a s . 13 tubuiar carcinomas, two tu b u la r a n d
lo o u i a r c a rc in o m a s a n d three infiltrating lo b u lar ca rc in o m a s
s t u c ie a s h o w e d any evtaence of DNA instability.
E x p a n s io n o f specific trinucleotide repeats was first n o ted in
s e v e r a l h eritab le neuromuscular diseases including fragile X
s y n d r o m e , m yotonic dystrophy and H untington's disease
( M iw a . 1994). All these repeats are polym orphic in n orm al
p o p u la t io n s as a result of variation in the n um ber o f
tr in u c le o tid e repeat units. Although instability o f these
r e p e a t s is a feature o f expanded disease-specific alleles,
’s m e a r in g ' o f the signal from a single allele o f trinucleotide
r e p e a t g en es has not been reported in norm al individuals,
in d ic a tin g th a t somatic microsateilite instability is uncom m on
in th e n o rm a l population.
W e h a v e detected somatic microsateilite instability a t the
D M - i (C T G )n repeat in-four of 78 (5% ) ’early’ sporadic
b r e a s t c an cers. Two tumours showed instability a t m ultiple
lo c i: th e D M - l. SCA1 and AR trinucleotide repeats and six
o f se v e n dinucleoude repeats. It seems unlikely th a t the
in s ta b ility seen in breast tumours represents a ran d o m
b a c k g r o u n d instability for this reason. Analysis o f D N A
s a m p ie s p rep ared from different norm al tissues (uninvclved
b r e a s t, cervix and uterus) showed no evidence o f m icro­
s a te ilite instability for the four instability cases. These d a ta
p r o v id e firm evidence that the instability seen was specific to
th e b r e a s t tu m o u r DNA. Parsons
al. (1995) reported
re c e n tly th a t rare ceils in a normal tissue p o pulation from
K N P C C p atients may harbour microsateilite alterations. O u r
d a t a f r o m analysis o f different normai tissues do no t conflict
w ith th is finding. The analysis of totai D N A prepared from a
7 a m n o rm a l tissue section wouid not be sufficiently sensitive
to d e te c t a rare variant normal ceil and m icrosateilite
in s ta b ility in these breast cancers may have arisen by a
d if f e r e n t m echanism to that seer, in H NPCC.
O u r d a ta snow a lower level o f microsateilite instability
( 5 3V i
th a n other published reports. This may reflect
d iffe re n c e s between the groups of tum ours studied, w ith o u r
s t u d y b ein g restricted to ’eariy' mam mographicaily detected
c a s e s , a n d variable frequencies of instability for the different
m a r k e r s studied. O ur data are most simiiar to the findings o f
W o o s t e r et ai. (19946) who noted instability at trinucleotide
r e p e a t s in 10% o f 100 breast cancers, with oniy rare
in s ta b ility a t dinucieotide repeats. However, in their study
la r g e r D*M-t alleles were preferentially unstable, whereas in
o u r g ro u p the small five iC T G u repeat aileie was m ost
f r e q u e n tly altered. O ther studies of breast cancer have noted
h ig h e r ’.eveis o f instability, but fewer cases studied. F o u r o f
20 » 2 0 % ) sporadic breast careers showed som atic m icro­
1994). G iebov
s a te ilite instability at several loci iYee
t 1994) noted differences in instability between tu m o u r
D N A fro m patients with a family history o f breast cancer
( F H B C ) an d sporadic breast cancers. Fifteen o f 18 FH B C
tu m o u r s show ed instability at muitipie loci whereas sporadic
b r e a s t can cers showed infrequent instability at specific loci.
P a te :
(1994) examined .3 primary breast cancers an d
n o te d h ig h levels o f both instability and loss o f hetero­
z y g o s ity fo r specific loci on chromosomes 2d. 3p and IOp.
M o s t coiorectal tum ours that display instability reveal
a .ie ie s o f altered length at muitipie loci that are frequently
d in u c ie o tid e repeats i.Aaitoner.
1993: Ionov
al..
993 . T w o o f the four breast tum ours th at displayed
m ic ro s a te ilite instability, revealed altered length aiieies a t
m u itip ie loci and therefore appear to reflect the p attern ot
.r.s ta o iiity seen in coiorectai cancer. These tum ours are
w o r th y o f investigation for m utation in candidate DNA
r e o a i r loct. However, two other tum ours displayed rare
•in stab ility , oniy detected for :.ie D M -l trinucleotide repeat,
a n d m o re closely refieet the pattern o f instability observed by
Vv .'o s t e r
(19946). Other markers need to be analysed in
tn e s e cases to confirm whether the instability is indeed
-e s tr .c te d to specific trinucleotide repeat loci.
T h e v ariatio n in frequency of instability <een lor the ten
■ e re c ts studied impiv that some ioci may be m ore unstable
tr.u n .'trie rs F or o ur tnnucleotide repeat data. D M -! a p p e a r s
■: ~c a m o re '>onNitive' locus t h a n cither AR o r SCA -l lo r
rS
Clinicopathological ana microsateilite instability data for
e a r l y sooracic oreas; cancers
Tabie II
C--C
:
Tumour
zR
Grade size mm H'-eore
?e
.”
uc
ii
v.c
Ii
- •-
MSI detected at
markers
D M -I. S C A -l. A R .
X "5b. E R T A . D 6S 193.
D lftS A O O . D I 6 S 4 0 2 .
D I6 S 4 I3
:
cc
15
ii
133
D M -I. S C A - l. A R .
\~ 5 b . E R T A . D 6 S I9 3 .
D lf tS - W ). D I 6 S 4 0 2
D lh S 4 l3
:
-C
II
II
-
ca
I
14
• v v .r .u in s ! d u c iu i
.
.
;i
. —• i i *
with
_ ..r c :n o m .i.
i^a
231
D M -I
D M -l
.n r i iir .;tin s : l o o u i a r c a r c i n o m a : I d c .
E R . o e s tr o g e n
\ i i c u n c s n iH ic n e g a ti v e .
re c e p to r:
M S I. m ic r o -
et
et al..
ai.
et
et ai.
et ai..
'.
jt ai.
et
Microsateilite instability in breast cancer
JA Shaw e t al
s tu d y in g m ic ro sa te ilite instability. This appears not to be a
f u n c t io n o f re p e a t len g th , as the num ber o f repeats at the A R
lo c u s f o r e x a m p le ten d s to be longer than at D M -l. Six o f
s e v e n d in u c ie o tid e re p e ats studied showed evidence o f
i n s ta b ility in o ne tu m o u r and again the variation in
fre q u e n c y
a p p e a rs n o t to be a function of repeat length.
T h e s e d a t a suggest th a t som e chrom osom al regions are m ore
u n s t a b le th a n o th ers. B o th chrom osom e 6q (Devilee
1 9 9 1> a n d I6 q (S ato
1991) have been shown previously
to h a r b o u r a re a s o f loss o f heterozygosity in breast cancer.
O u r d a t a p ro v id e o th e r evidence for genomic instability in
th e s e c h r o m o s o m a l regions and for specific tnnucleotide
r e p e a ts o n 6 p, I9q a n d X q in breast cancer. Any stru c tu ral
p e r t u r b a t i o n o f these chrom osom ai regions may alter the
f u n c t io n o f gene(s) h a rb o u re d on the specific chrom osom es.
T h e D N A in stab ility observed in the four breast cancers
c o u ld b e a m a n ife sta tio n o f errors in D N A repair as has been
t o u n d f o r H N P C C (F ish el
1993: Leach
1993).
T h e r e ia x e d genom e stability, observed as m icrosateilite
in s ta b ility , c o u ld be in itia ted by alteration o f genes involved
in e it h e r D N A rep lic a tio n or repair and would be an early
e v e n t in c a rc in o g en esis (L oeb. 1 9 9 4 ) . Such unstable c an cer
ceil g e n o m e s c o u ld p ro m o te a cascade o f m utations som e o f
w h ic h e n a b le the c a n c er ceils to bypass the host reg u lato ry
p ro c e s s . S im ilarly the allele instability observed in o u r series
o f 'e a r i y ' b re a s t can c e rs may be a sensiuve in d icato r o f
g e n o m ic h y p e rm u ta tio n in these tum ours. A lthough the D M l a n d A R m icrosatellites are expressed, it is uniikely th a t
th e s e lo c i them seives co n trib u te :o the developm ent o f b re a s t
c a n c e r sin c e all o f th e unstable aiieies lie well w ithin th e
n o r m a i p o p u la tio n ran g e and their sizes are com m on in the
n o r m a l p o p u la tio n . H ow ever, different length repeat alleles,
e v e r , w ith in the n o rm a l range, m a y have subtle influences o n
c e llu la r m e ta b o lism , w hich m a y m anifest in breast cancer.
T h e c lin ic o p a th o lo g ic a l features o f the four breast tu m o u rs
th a t d is p ia y in stab ility were examined for possible co rre la tio n
ei al.,
et al..
et al.,
et al.,
w ith th is p h e n o ty p e. All were infiltrating ductal carcinom as,
w e ll o r m o d e ra tely differentiated, and node-negative. O estro­
g e n re c e p to r w as detected in all at moderate to high levels
( R a ja k a r ia r a n d W alker. 1995). No evidence o f instability a t
D M - l w as d etec ted for 13 tubular, two mixed lobular a n d
tu b u l a r cases a n d three infiltrating lobular carcinomas from a
t o t a i o f 78 tu m o u rs th at were screened. Linell
(1980)
h a v e su g g ested th a t tubular carcinomas may progress to less
d if fe re n tia te d carcinom as if left untreated, and the tu b u la r
m ix e d c arc in o m a s described by Ellis
(1992) m ay lend
s u p p o r t to this. If this is the case, our findings would suggest
e it h e r th a t instability occurs at a particular stage o f
d e v e lo p m e n t a n d progression or only with certain pathw ays
o f d e v e lo p m e n t a n d progression.
I n su m m a ry , we have detected somatic m icrosateilite
in s ta b ility in 5% o f 78 ‘early’ sporadic breast cancers. T hese
d a t a fo r ‘e a rly ’ breast cancers support the suggestion th a t
m ic ro s a te ilite instability may be an early event in the genesis
o f s o m e sp o ra d ic breast cancers (Y
1994). M oreover,
o u r d a ta d e m o n stra te th at instability is not found betw een
d if fe re n t n o rm a l tissues from the same individual, b u t
a p p e a r s to be specific to DNA prepared from w ithin a
t u m o u r . A n extended study on a larger range o f lesions
in c lu d in g a d d itio n a l tubular and lobular carcinomas, cases o f
d u c ta l c a rc in o m a
and ‘at risk’ lesions (e.g. florid a n d
a ty p ic a l h yperplasia) wiil be im portant to verify these
o b s e r v a tio n s a n d to determine the role o f these D N A
c h a n g e s in the n a tu ra l history o f breast cancer.
et al.
et al.
ee et al.,
in situ
A c k n o w le d g e m e n ts
T W alsh and S Chappell are undertaking PhD studies supported
by th e Royal Society and the University of Leicester. X Johnson is
s u p p o rte d by the Muscular Dystrophy Group of Great 3ritam and
N o rth e rn Ireiand. grant numcer R A 3 .T 2 6 / 3 . We are grateful to
M rs S D eanng for technical support.
R eferences
A A LTO N EN
NEN
L A . P E L T O M A K I ?. L E A C H F S . S IS T O N E N P. P Y L K K A -
1.
M E C K L IN
JP .
JA R V fN E N
H.
PO W ELL
SM .
JE N
M AN
T E R R Y L . J H A N W A R S. 3 E R C H U C X A . I G L E H A R T J D . M A R K S J .
JO .
M E.
ABURATAN!
T H IR IO N
JF . H U D SO N
RU N O LE
SA. CR O W
HARLEY
H.
HUNTER
HG.
BUCKLER
A J.
3 .A L L IN G E R
K.
STANTO N
VP.
W IS E M A N
RG.
o v a r ia n c a rc in o m a s .
T . S O H N R. Z E M E L M A N
8. S N E L L
S . D A V I E S J. S H E L 3 0 U R N E P. 3 U X T O N
J.
11992). Moiecuiar basis of myotonic dystrophy:
ex p a n sio n o f a trinucleotice ;CTG) repeat at the 3* end o f a
tr a r.a c n o t encoding a protein kinase famiiy member.
68.
IO N O V
SEN N O
L. A G U IA R I
GL
AND P !V a
'L E E
P. v a n
HERM ANS
JA N S E N
GALEA
HG.
new
m e c h a n ism
fo r
c o lo n ic
363. 5 5 3 -5 6 1 .
P J. A .M E M IY A C . AS L A N D IS C . S H A W
BROOK
C J.
(1 9 9 1 ).
D.
F E N W IC K
R.
D J.
KORNELUK
RG.
M. BROUGHTON
N. L O C K E R
A. B L A .M E Y
RW
ELSTON
HJ
AND
T.ttin•irngy. 19.403 —410
S N ER A N D V O G ELSTE IN 3 .1
i A g e n e tic m o d e l f o r
—e c . a l t u m o u r t genesis. CWx. 61. ” 59 - "ft”
_ ? L E S C O E M K . RA O MRS. COPF.LAND NG. JE N K IN S
NA.
• ■» 3 = 3 r KANF. M \.ND K O LO D N FR R (1 9 9 3 ). T h e h u m a n
- c er.o h o m o io g MSH2 - n o its usM H iiation w ith h e r e d ita r y '
''•''i> colon cancer . .
"5. irtE" • Id.'N
W E IR IN G E R
8. (1 9 9 2 ). P h y s ic a l a n d
g e n e tic
c h a r a c t e r i s a t i o n o f th e d is ta i s e g m e n t o f th e m y o to n i c d y s t r o p h y
u rea o n
Genomics.
I9 q .
SPA D A
AR.
W IL S O N
13. 5 0 9 - 5 1 7
EM . LU B A H N
DB. H A R D IN G
AE
AND
r : 3 C H B E C K K H . ( 1991). A r .c r o g e n r e c e p t o r g e n e m u t a t i o n s i n X : r . x e d s p in a l a n d b u lb a r m u sc u ia r a tr o p n y .
LEACH
FS.
PA RSO N S
N I C O L A ID E S
R.
Suture.
352. 7 7 -7 9 .
N C . P A P A D O P O U L O S N . L IU
PELTOM AK I
N Y S T R O M -L A H T I
)
a
AND
s im p le
T S I L F I D I S C . S H U T L E R G . H E R M E N S R. W O R M S K A M P N G M .
AND CO RN EL2SSE
E L S " ' - T ’ v \ n d E L L I S lO. i !9 « ) i. P a th o lo g ic a l p r o g n o s tic f a c t o r s
-. - r e a s : c a n c e r . I. T h e ’ .tree o f h is to io g ic a i g ra d e in b r e a s t
e x p e r ie n c e fr o m a larg e s tu u y w ith io n g -te r m f o ll o w - u p .
- E •.
..
r
D
SM EETS
PEA RSO N
Histnpatlwlmey.
-f ;
rev eai
Suture.
DE J O N G
S. S H IB A T A
J.
CW
( 1 9 9 2 ) . Patnoiogicai prognostic f a c t o r s in
:anccr. (I. Histoiogicai :>re. Reiationsnip with survival in
_ _-~e -tuuy wuh long-term :'oiiow-up.
20. 4 7 9 AND
G.
M ALK HO SYA N
U b i q u i t o u s s o m a t i c m u t a t i o n s in
V L I E T M . v a n S L O L 'N P. K U I P E R S - D I J K S H O O R N
LA
2
sequences
HARLEY
A .le-.cr/pe of human breast carcinoma: a second major site for
css
heterozygosity is on chromosome bq. Oncogene. 6. 1705E L _ IS
p r im a ry b r e a s t a n d
5 4 . 5 “ 0 3 —3 7 0 9 .
M A.
M . (1 9 9 3 ).
c a rc in o g e n e s is.
R. (1 9 9 2 ).
Hunt. '-/al. Genet..
V
P E IN A D O
re p e a te d
Dinucieotide
re c e u : polym orphism in the nutnan estrogen receptor i ESR) gene.
I. 354.
D E'
Cancer Res..
Y.
PERUCHO
- = c _ ?08.
DEL
8RCAI M u t a t i o n s in
Science. 266. 1 2 0 - 1 2 2 .
F r e q u e n t p 5 3 g e n e m u t a t i o n s a n d n o v e l a ii e ie s in f a m i l i a l b r e a s t
c an ce r.
DE.
Ceil.
D G . 3 A R R 5 T T J C . S K O L N IC K M H . K A M B A A N D
R. (1 9 9 4 ).
O L E 3 0 V O K . M C K E N Z I E K E . W H IT E C A A N D S U K U M A R S. ( 1 9 9 4 ) .
C . J U V O N E N V. JO H N S O N X . H A R P E R PS. S H A W D J A N D
H O U SEM A N
K. M C C LU R E M . F R Y E C .
'- V E A V E R - F E L D H A U S J . D IN G W . G H O L A M l Z . S O D E R K V I S T P .
Science. 260. 312-316.
M CCURRACH
D.
J. M IK I Y. E D D IN G T O N
A.
fam ilial coiorectal cancer.
JO N E S
k
S. 3 E N N E T 7 L.M. H A U G E N - S T R A N O
SW EN SEN
C H U R C H
'
K . T A V T IG IA N
Ha m i l t o n s r . p e t e r se n c m . k in z l e r k w . v o g e l st e in a
AND c= _A C H A P E L L E A . (19*3). Clues to the pathogenesis o f
3R C C K
t
F U T R E A L PA . L I U Q . S H A T T U C K - E I D E N S D . C O C H R A N C . H .A R S H -
J.
P.
M . G U .A N
S IS T O N E N
X -Y .
P.
B. J E N
AALTONEN
F O U R N IE R
J.
LA.
REK. T O D D
S.
L E W I S T . L E A C H R J. NAY LO R S L . W E IS S E N B A C H J . M E C K L I N J?
J .A R V I N E N H . P E T E R S E N G M . H A M IL T O N S R . G R E E N J . J A S S J .
A ATSON
P.
K IN Z L E R
KW
f’ttr5
L iN E L L
LYNCH
\N D
HT.
TRENT
V O G E L S T E IN
JM .
OE
LA C H A P E L L E
3. iI 9 9 3 ) .
M u ta tio n s
h o m o io g in h e re d ita r y n o n p o iy p o s is c o lo r e c ta l c a n c e r .
F.
c a rc in o m a
L JU N B E R G
a s p e c ts
O
of
and
eariy
AN DERSSON
s ta g e s ,
I.
1 1 9 8 0 ).
p ro g re s s io n
and
A.
o f
a
Cell.
B re a st
re la te d
- r o n i e m s . A c ta Putiiol. Micrnhiol. Scunti. 2 7 2 ( s u p p l . ) . 6 3 - 1 0 1 .
MicrovateU'rta instability in breast cancer
JA Shaw -r al
1397
LOE3 LA. (1994). M icrosateilite instability: marker of a mutator
p h e n o ty p e in cancer. Cancer Res.. 5 4 , 5059-5063.
M E N A S C E L P . W H I T E G R M . H A R R IS O N C J A N D B O Y L E J M . (1 9 9 3 ).
L o c a lisa tio n o f the estrogen receptor (ESR) to chromosome 6q25.
I bv F IS H and a simple post-FISH banding technique. Genomics.
17. 263 —265.
M I K I Y . S W E N S E N J. S H A T T U C K -E ID E N S D . F U T R E A L PA . H A R S H M A N K - T A V T I G I A N S . L IU Q . C O C H R A N C . B E N N E T T L M . D IN G
W . BELL
R . R O S E N T H A L J. H U SSEY C. T R A N T. M C C L U R E M .
S A IT O S . O K U I K . T O K IN O T . O S H IM U R A M A N D N A K A M U R A Y.
Isolation and mapping of 6 8 RFLP markers on human
chrom osom e 6 . Am. J. Hum. Genet., 5 0 , 65-70.
(1 9 9 2 ).
S A T O T . A K IY A M A F, SA K A M O T O G . K A SU M I F A N D N A K A M U R A
1991). Accumulation of genetic alterations and progression of
p rim ary breast cancer. Cancer Res., 5 1 , 5794-5799.
S H A W I A. EVANS J G . JOH N SON K A N D W A LK ER R. (1995).
M icrosateilite instability in early breast cancer. J. Pathol.. 1 7 5 ,
Y. (
109.
F R Y E C . H A T T I E R T . P H E L P S R . H A U G E N -S T R A N O A . K A T C H E R
S H E N Y . T H O M P S O N A T . H O LM A N K, C A L L E N D F . S U T H E R L A N D
H.
AND
R I C H A R D S R I . (1992). Four dinucieotide repeat
polym orphism s on human chromosome 16 at D16S289.
D16S318. D16S319 and D16S320. Hum. Mol. Genet., 1,773.
T H I B O D E A U S N . B R E N G A N D S C H A I D D . (1993). Microsateilite
instab ility in cancer o f the proximal colon. Science. 2 6 0 , 316-819.
W E B E R J L .A N D M A Y P E . ( 1 9 8 9 ) . Abundant class of human DNA
polym orphism s which can be typed using the polymerase chain
reaction. Am. J. Hum. Genet.. 4 4 , 3 8 8 - 3 9 6 .
Y A K U M O
K . G H O L A M I Z. S H A F F E R D . S T O N E S . B A Y E R S.
W R A Y C . B O G D E N R . D A Y A N A T H P . W A R D J. T O N I N P. N A R O D
S.
B R IS T O W
RO STECK
PK .
P.
M O R R IS O N
P.
L A I M . B A R R E T T JC . L E W IS C . N E U H A U S E N
N O R R IS
S.
C A N N O N -A L B R IG H T
A N D
S K O L N IC K
F.
H E L V E R IN G
L. G O L D G A R
L.
D . W IS E M A N
R. K A M B
A
M H . (1 9 9 4 ). A s tro n g c a n d id a te f o r th e b r e a s t
a n d o v a r i a n c a n c e r s u s c e p tib ility g en e
BRCA1. Science,
266, 6 6 -
71.
M IW .a S . ( 1 9 9 4 ) . T r i p l e t r e p e a t s s tr ik e a g a in .
ORR
HT. C H U N G
3EA U D ET
ZO G H 31
C A G
AL.
HY.
M CCALL
(1 9 9 3 ).
r e tr e a t in
Sature Genet.,
6, 3 - 4 .
M Y . B A N F I S. K W IA T K O W S K I T J. S E R V A D IO A .
AE.
D U V IC K
E x p a n s io n
of
LA.
an
RANUM
u n s ta b le
s p in o c e r e b e lla r a ta x ia ty p e
I.
LP
AND
trin u c le o tid e
Sature Genet..
4,
P A R S O N S R . L I G - M . L O N G L E Y M J. F A N G W -H . P A P A D O P O U L O S N .
J . DE LA C H A P E L L E A . K I N Z L E R K W . V O G E L S T E I N B A N D
M O D R IC H
P.
(1 9 9 3 ).
d e d c te n c y in R E R PA RSO NS
H y p e rm u ta o iiity
t u m o u r c e ils .
Ceil.
and
m is m a tc h
SR.
K IN Z L E R
KW
AND
re p a ir
V O G E L S T E IN
B. B E R K T .
B.
(1 9 9 5 ).
M i s m a t c h r e p a i r d e f ic ie n c y in p h e a o ty p ic a ily n o r m a l h u m a n c e lls .
Science.
268. 7 3 8 -7 4 0 .
P A T E L U . G R U N D F E S T -B R O N I A T O W S K I S. G U P T A M A N D B A N E R J E E S . ( 1 9 9 4 ) . Microsateilite instabilities at five chromosomes in
p rim a ry b reast tum ours. Oncogene. 9 . 3695-3700.
R a J A K a R L a R R A N D W A L K E R R a . (1995). Pathoiogicai and
bio lo g ical features o f mammographicailv detected invasive
b re a s t carcinom as. Br. J. Cancer. 7 1 . 150- 154.
ROYAL
J . G Y A P A Y G . D IB C . V I G N A L A . M O R I S E T T E J .
M I L L .A S S E .A U
P . V A Y S S IE X
G
AND
LATHROP
M . (1 9 9 2 ).
A
second generation linkage map of the human genome. Sature,
359. 7 9 4 -8 0 1 .
W O O STER
R . N E U H A U S E N S. M A N G IO N J. Q U IR K Y . F O R D
D.
m a r s h a l l
g
. m a r o d s . l e n o ir g m . l y n c h h . f e u n t e u n j .
D E V I L E E P . C O R N E L IS S E C J. M E N K O F H . D A L Y P A . O R M IS T O N
W . M C M A N U S R . P Y E C . L E W IS C M . C A N N O N - .A L B R I G H T L A .
P E T O J . P O N D E R B A J. S K O L N IC K M H . E A S T O N D F . G O L D G A R
75, 1 2 2 7 - 1236.
R . L I G -M . L O N G L E Y M . M O D R IC H P. L IU
H A M IL T O N
W E IS S E N B A C H
C O L L I N S N . N G U Y E N K . S E A L S. T R A N T . A V E R I L L D . F I E L D S P .
2 2 1 -2 2 6 .
JE N
G R
C O L L E G E O F P A T H O L O G IS T S W O R K IN G G R O U P . ( 1 9 9 0 ) .
S H S 3reast Screening Programme: Pathology Reporting in Breast
• • Cancer Screening. Royal Coilege of Pathologists: London.
(1994a). Localisation of a breast cancer
susceotibility gene. 3RCA2, to chromosome 13a 12-13. Science,
D E A N D STRA TTO N M R.
2 6 5 . 2 0 8 8 -2 0 9 0 .
W O O STER
R . C L E N T O N -JA N S E N A M . C O L L IN S N . M A N G IO N J,
C O R N E L I S R S . C O O P E R C S . G U S T E R S O N B A . P O N D E R B A J . VON
D E IM L IN G A
STRATTON
W E IS T L E R O D . C O R N E L IS S E C J. D E V IL E E P A N D
M R.
(19946).
I n s ta b ility
' m ic r o s a t e ll it e s ) in h u m a n c a n c e rs .
of
sh o rt
ta n d e m
re p e a ts
Sature Genet.. 6, 152-156.
Y E E C . \ R O O D ! N . V E R R I E R C S A ND P A R L F F . (1994). M i c r o s a t e i l i t e
in s ta b ility
Res..
and
lo ss
5 4 . 1 6 4 1 -1 6 4 4 .
of
h e te r o z y g o s ity
in b r e a s t c a n c e r .
Cancer
JOURNAL OF PATHOLOGY. VOL. 185: 18-24 (1998)
MICROSATELLITE INSTABILITY IN DUCTAL
CARCINOMA
I N S IT U OF THE
TOM
W A L S H 1. S T E P H E N A . C H A P P E L L 2 , J A C Q U I A . S H A W 1 A N D
R O S E M A R Y A . W A L K E R 1*
1Breast Cancer Research Unit. Department of Pathology. University of Leicester. Glenfield Hospital. Gruby Road.
Leicester LE3 9QP. U.K.
-Department of Biochemistry. University of Leicester. Leicester LEI ~RH. U.K.
SUM M A R Y
M icrosateilite instability (M I"-) is associated with defects in mismatch repair, resulting in a "mutator' phenotype and the development
and progression of cancer. M I+ has been documented in invasive breast carcinomas. This study was undertaken to determine whether
M I* is found in the early non-invasive form of breast cancer, ductal carcinoma
(D C IS). We examined microdissected ducts from
23 cases of D CIS with 11 m arkers comprising mono-, di-, and trinucleotide repeats from six chromosomal regions. Five tum ours (22 per
cent) displayed M I* at two or more loci, in all ducts examined. A further seven (30 per cent) tumours showed alterations at a single locus
(the D M -l trinucleotide), and for two of these, heterogeneity between ducts was observed. Alterations at m icrosateilite repeat motifs in
the coding regions of four cancer-associated genes (
, and
were not observed. Immunohistochemistry
revealed that there was no loss o f reactivity for the mismatch repair proteins, M L H 1, .VISH2, and PV1S2, in the D C IS cases. In general.
M I* tum ours and those with alterations a t the D M -l microsateilite were predominantly of higher nuclear grade and expressing
suggesting that aberrations in DNA repair functions may lead to the acquisition of a more aggressive phenotype in breast
cancer. C 1998 Joh n W iley
Sons, Ltd.
insitu
TGFpRlL IGFIIR, BAX
c-erbB-2,
E2F-4)
&
J . P a th o l. 185: 1 8 -2 4 , 1 998.
KEY w o r d s —
ductal carcinoma
insitu: microdissection; m icrosateilite instability; D M -l
IN T R O D U C T IO N
Instability of microsateilite DNA sequences (Ml*),
also known as replication error phenotype, character­
ized by the presence of random contractions or expan­
sions in the length of simple sequence repeats, is a key
feature of the hereditary' non-polyposis colorectal cancer
(HNPCC) syndrome.1 It has been shown in HNPCC
kindreds that the defect associates with inherited
mutations in human homologues of the bacterial mis­
match repair (MMR) genes h M S H 2 , h M L H l, h P X tS l,
and h P X fS 2.2 Examination of protein expression by
immunohistochemistry, on paraffin-embedded samples,
has been shown to be a rapid method for pre-screening
tumours for mutations in the MMR genes.3 MI* has
also been observed in tumours not associated with
HNPCC, suggesting that the phenotype is common
in many sporadic cancers.4 Several studies have
documented instability of microsatellites in cohorts of
invasive breast carcinomas.5-15 The incidence of MI*
varies considerably; for example, Peltomaki et al.15
found no evidence of MI* in 84 breast carcinomas,
whereas Patel et al? detected instability in all of 13
tumours studied. In contrast, the analogous investiga­
tion of non-invasive lesions is in its infancy. As a
consequence, it is unclear at what stage of breast cancer
•Correspondence to: Rosemary A. Walker, Breast Cancer Research
Unit. Department of Pathology. University of Leicester, Glenfield
Hospital, Grobv Road. Leicester LE3 9QP. U.K.
Contract grant sponsors: University of Leicester: Glenfield Hospital
Trust Research Committee.
C C C 0022-3417/98/050018-07 S I7.50
c 1998 Jo h n Wiley & Sons, Lid.
alterations; mismatch repair protein expression
development and progression microsateilite instability
becomes manifest.
We therefore focused our study on non-invasive
lesions. Ductal carcinoma in situ (DCIS) of the breast is
a malignant proliferation confined‘to the mammary duct
and is one of the earliest recognizable forms of breast
cancer. DCIS is a heterogeneous group of pathological
and biological subtypes, which may differ in behaviour.
The lesions can be categorized by their nuclear
morphology into high, intermediate, and low nuclear
grade.16 High-grade lesions may be at greater risk of
local recurrence if incompletely excised and may be
associated with the development of invasion.17 We
investigated the subtypes of DCIS for MI* and MMR
protein expression to determine the chronology of
genomic instability in breast cancer development and to
ascertain whether potential differences in biological
behaviour of the subtypes are reflected in distinct
molecular changes. Overexpression of the oncogene
c-erbB-2 is found with high-grade, comedo type,18
whereas alterations to p53, although predominantly
identified in high grade, can be found in low-grade
DCIS.19 We therefore wished to correlate the MI*
status in these cases with p53 and c-erbB-2 alterations.
The existence of a mutator phenotype, characterized
by MI*, may be responsible for aberrations in critical
cancer-associated loci.20 .Among the targets identified to
date are repetitive sequences in the coding regions of the
TGF/? type II receptor {TGFflRIT)?] the insulin-like
growth factor II receptor (IGFIIR)?223 BAX?4 and the
E2F-4 gene.25-26 Each of the genes has tumoursuppressive functions relating to the resistance and
Received 31 July 1997
Accepted28 October 1997
19
MIC RO SATELLITE INSTABILITY IN DC IS
activation o f T G F // (T G F fiR lI and IG FIIR, respect­
ively). the m aintenance o f ap optosis {B A X ), and
co-ordin ating control o f the cell cycle {E2F-4). These
m utational 'hotspots' have been screened for alterations
in order to determ ine w hether they are o f significance in
breast cancer and to increase our know ledge about
genom ic instability in early breast cancer lesions.
M A T E R IA L S A N D M E T H O D S
6q (D 6S186. D 6S193), and chrom osom e 19q (X75b).
The (TA)„ dinucieotide (ESR i. located 1 kb upstream o f
the oestrogen receptor, was also investigated. The trinu­
cleotide m icrosatellites were D M -l (C T G );| located in
the 3' untranslated region o f the m yotonic dystrophy
protein kinase gene, and a (CAG)„ repeat which is
located in'the coding region o f the spinocerebellar ataxia
type I (S C A -l) gene. The PCR primer sequences and
P A G E conditions, have been described p reviously."
except for the BAT 25 and 40 markers.27
Tissues
M aterial from 23 patients w as exam ined: 13 cases
were identified by m am m ographic screening and ten had
presented clinically. All tissues had been fixed in 4 per
cent form aldehyde in saline for 18-24 h. blocks selected,
and processed through graded alcoh ols to paraffin wax.
For tw o cases, lymph node tissue was available which
was used as a source o f normal D N A , w hereas for 21
cases norm al breast tissue, at least 3 cm aw ay from the
area o f in situ carcinom a, was the source o f norm al
D N A . All tum our and normal sam ples had received the
sam e fixation and processing.
H aem atoxylin and eosin-stained sections were exam ­
ined and D C IS was categorized for grade and architec­
ture using the criteria in the N H SB SP guidelines, second
edition (R A W ).
X ticrodissection and D N A extraction
A m icrodissection procedure was used to analyse
different ducts within the same tum our section and to
separate tum our cells from adjacent norm al strom a.
Briefly, serial 10 am sections were m ounted on glass
slides and deparaffinized twice with xylene, rinsed twice
with 99 per cent ethanol, once with 95 per cent ethanol,
stained with eosin. and air-dried. The adjacent section
was routinely stained with haem atoxylin and eosin, and
m ounted: ducts for m icrodissection were marked for use
as a reference. U p to three individual n on-con tiguou s
ducts in volved by D C IS were m icrodissected from the
eosin slides using a drawn-out capillary tube and a
d issecting m icroscope (Am erican Optical C orp.). Pro­
cured cells were resuspended in 25 //I o f digestion buffer
co n tain in g 0 1 m T ris-H C l (pH 8 -8 ). 0-05 m E D T A , and
2 0 0 /ig /m l proteinase K and incubated at 56°C for 3 h.
F ollow in g proteinase inactivation at 95°C for 10 min.
5 //I o f the solu tion w as used as tem plate in polym erase
chain reaction (P C R ) am plification. N orm al D N A
tem plate was prepared from sections o f norm al breast
and lym ph node in the sam e manner.
A n alysis o f XII*
T h e m icrosateilite markers studied were categorized
into three groups based on the repeat unit. The m o n o ­
nucleotide repeats were BAT25 and BA T40, which span
polym orphic (A)„ repeats, located in introns o f the c -kit
o n cogen e and 3-/f-hydroxysteroid dehydrogenase gene,
respectively. D inucieotide (CA)„ m icrosatellites m apped
to ch rom osom e 16q (D 16S413, D 16S400), ch rom osom e
< 1998 John Wiley & Sons. Ltd.
M u tation al analysis
M utations o f TG FfiRlI (A ) 10. IGFIIR (G )8. B A X (G )8,
and E2F-4 (CAG)„ were analysed using a PCR-based
assay. Primers which span each o f the coding region
repeat m otifs have previously been described.20 25 A ber­
rantly sized PCR products were sequenced to confirm
that the size difference was due to an alteration in the
repeat m otif.
Im m unohistochem istry
Each o f the tum ours was analysed for the expression
o f p53, c-erbB-2. and three o f the m ism atch repair
proteins, M LH 1, PM S2, and M SH2.
p53 was detected using the polyclonal antiserum CM 1
(N ovocastra). Sections were exposed to tw o cycles each
o f 5 min o f m icrowave exposure in 10 m.vt citric acid
buffer, pH 6 0, using an 800 W m icrow ave oven at full
power. CM1 was applied at 1:800 dilution overnight at
4°C and detected using biotinylated swine anti-rabbit
im m unoglobulin antiserum , follow ed by streptavidin
b iotin-peroxid ase com plex. C ontrols were the om ission
o f the primary antibody and the inclusion o f a known
positive control with each staining batch. Ten ducts
were selected and the percentage o f positive cells was
determ ined by counting all cells within them.
c-erbB-2 was detected using NCL-CB11 m onoclonal
antibody (N ovocastra). Sections were incubated with the
antibody diluted 1:80 overnight at 4°C. follow ed by
biotinylated rabbit anti-m ouse im m unoglobulin and
streptavidin biotin-peroxidase com plex. C ontrols were
as for p53. R eactivity was determined as negative;
positive, if all ducts showed membrane staining; or
heterogeneous, if a proportion o f ducts show ed staining.
PM S2 was detected using the C lone 9 m onoclonal
antibody (O ncogene Science). This antibody detects a
96 kD protein by Western blot. The portion o f the
PM S2 m olecule from which the antibody has been raised
has not been characterized. MLH1 was detected with the
C lone 14 m onoclonal antibody (O ncogene Science). This
antibody reacts with a 87 kD protein on W estern b lot­
ting and with a nuclear antigen on paraffin-em bedded
tissues. M SH 2 was detected using the F E 1 1 m onoclonal
antibody (O ncogene Science). The antibody was raised
against a carboxyl terminal region o f the M S H 2 protein
and detects a 100 kD protein on W estern blots. Sections
were pretreated by pressure cooking. Briefly, dew axed
sections were immersed in a pressure cook er (Prestige,
M odel 6189) containing boiling lOmvt citric acid
JOURNAL OF PATHOLOGY, VOL.
185: IS—24 (1998)
20
T. W ALSH E T AL.
i
T a b le I— M ic r o s a te llite in sta b ility in D C IS
T um ou r
M ierosatellite instability
G rad e
1
High
High
High
High
High
High
High
High
High
High
High
Inter.
Inter.
Inter.
Low
Low
LowLow
Low
LowLow
Low
Low
5
6
7
11
12
13
14
16
17
9
10
15
8
3
4
18
19
20
21
■>*»
23
D M -1, D6S193
DM-1
D6S193. D6S186
DM-1
DM-1
DM-1
DM-1
D M -1. D 6SI93. D16S400. D I6S413
D M -1. D6S193
DM-1
DM-1
BAT25. BAT40
p53
e-er/>B-2
_
_
-
Het. +
85%
+
—
—
Het. +
Het. +
50%
20%
80%
+
+
Het. +
-
+
+
+
—
—
—
—
—
—
45%
—
—
—
—
—
—
—
30%
N orm al D N A
NB
NB
NB
NB
NB
NB
LN
NB
NB
NB
NB
NB
NB
NB
NB
NB
NB
NB
NB
NB
LN
NB
NB
Inter. = interm ediate: H et. = heterogeneous: NB = norm al breast; LN = lymph node: — = negative: + = positive.
buffer. pH 6 0. The sections were incubated for 1 min at
m axim um pressure. After com p letion , slides were im m e­
diately transferred to phosphate-buffered saline (PBS)
prior to im m unostaining. C lone 9 and F E l l were
applied at 1:40 dilutions, and C lone 14 at 1:10, overnight
at 4*C, follow ed by biotinylated rabbit anti-m ouse
im m unoglobulin antiserum and streptavid in-b iotin per­
oxidase com plex. Staining o f D C IS was com pared with
that seen in normal breast tissue in the sam e section, in
relation to presence and intensity.
S ta tistic a l analysis
Statistical com parisons were perform ed using the
M an tel-H aenszel test.
R ESU L TS
M icrosatellite in stability
W e tested m ultiple ducts in 23 cases o f D C IS with 11
m icrosatellite markers, from six chrom osom al regions,
and correlated our findings with p athological and
m arker data (Table I). MI + w as characterized by the
appearance o f alleles o f altered length in tum our
sam ples, indicating an alteration in m icrosatellite size
(Fig. 1). A tum our was classified as M I+ when altera­
tions in m icrosatellite size at tw o or more independent
gen om ic sites were observed (as previously defined28).
MI + sam ples were verified by m icrodissecting the sam e
duct from a serial tum our section and carrying out an
independent PCR.
C 1998 John Wiley & Sons. Ltd.
M I* was detected in 5 o f 23 cases (22 per cent). In
three o f these, both trinucleotide and dinucleotide
repeats were affected. One tum our show ed M I* at two
dinucleotides. The rem aining case showed alterations at
the long m ononucleotide repeats (BA T 25 and 40). For
all five cases, M I+ was detected for all ducts dissected,
for all markers. A futher seven cases (30 per cent)
showed instability exclusively at the DM-1 repeat, pre­
viously shown to be a sensitive indicator o f M I * .11 Tw o
o f these dem onstrated intratum oural heterogeneity, with
different patterns o f instability between different ducts
analysed from within the same tum our section.
M I* tum ours and those with alterations to DM -1
were then screened for alterations in the repeat tracts o f
the TGF/JRII. IG FIIR. BA A', and E2F-4 genes. All
tum our sam ples were screened in the presence o f a
wild-type control for each gene. Positive controls, for
wild-type alleles, were established by am plifying a series
o f normal D N A s and com paring the size o f PCR
products with an M 13 D N A sequence ladder (data not
show n). The expected sizes o f the PCR products for the
TGF/1RII. IG F IIR . BAX. and E2F-4 genes are 73, 111,
96 and 135 bp, respectively. Each PCR product w as then
sequenced directly to confirm that it contained the
w ild-tvpe number o f repeats.
One case (tum our 14) showed two distinct E2F-4 PCR
products differing in size by 3 bp. Sequencing o f these
products show ed that the size difference was due to the
presence o f tw o alleles, containing 13 and 14 cop ies o f
the C A G repeat, respectively. This alteration w as also
present in the corresponding norm al D N A . and so was
not a tum our-specific alteration.
JOLRNAL OF PATHOLOGY. VOL.
185: 18-24 (1998)
21
MICROS. VTKLL1TK INSTABILITY IN DCIS
N
T
T
N
T
T
T
T
N
T
N
T
P
N
T
&
id
lb)
la)
)
17
id .
17
w 17
16
1 4
N T
T
T
N
T
T
T
I)
id
20
Fig. I —Genomic DNA from paired normal (N) and microdissected ducts (T) from samples o f paraffin-embedded tissues, (a.b) Examples of
alterations at DM-1. in tumours 14 and 16. (c.d.e) Tum our 17 with instability at DM-1 and D16S4U0. and stability at TGFpRIL respectively. [A
positive control (P) containing (A),0 was included.] (0 Tum our 20 showing instability at BAT25. (g) Tum our 13 showing intraductal heterogeneity
at DM -1. Arrows indicate alterations in the electrophoretic mobility o f PCR products from tum our compared with normal DNA. indicating MI*.
The numbering of the tumours is the same as that used in Table I
[ PM S2, \1 L H 1 . and \1 S H 2
N uclear staining was observed for PMS2 and M L H l.
but in som e o f the tissues there was no reactivity o f the
normal or tum our, or variable staining o f both. This
was taken into account when categorizing the
reactivity o f the DCIS. Tw o cases showed stronger
t 1998 John Wiley & Sons. Ltd.
PM S2 staining in the DCIS than normal, but for the
others reactivity was similar to normal breast tissue in
the sam e section.
The staining for M SH 2 was stronger and generally
there was clear nuclear reactivity in normal breast and
D C IS (Fig. 2). Overall, there was no difference in the
jo l r n a l o f pa t h o l o g y , v o l .
185: 18-24 (1998)
r. W \LSH AT (A.
Fig. 2 -Im m unohistochem ical staining o f DCIS with MSH2 antibody.
Duct showing ductal carcinoma in situ with two normal acini adjacent.
There is staining o f the nuclei of all cells
reactivity for PM S2. M L H I. and M SH 2 between those
cases show ing MI"- and those not.
c-erbB-2 and p53
c-erbB -2 w as detected in ten cases, four show ing
heterogeneous labelling. Three were M I* and five
show ed instability at DM -1; this w as a significant a sso ci­
ation ( P = 0 02). All were high grade and com prised 10 o f
the 11 high-grade cases (Table I).
p53 was detected in 20-85 per cent o f cells in six cases.
Five o f these show ed instability; four had instability at
D M -1 only and one (tum our 17) w as M I* . The five
p53-positive cases that showed instability were all high
grade and the sixth p53-positive case was low grade. The
17 rem aining tum ours that show ed no im m unoreactive
p53 included four o f the five tum ours that were M I* and
three other cases show ing instability at DM -1 only.
D C IS type
Eleven cases were o f high nuclear grade, three o f
interm ediate grade, and nine low grade. M I* and in­
stability at DM -1 were found in high grade (8/11) and
interm ediate grade (2/3) at a higher frequency than in
the low -grade cases (2/9), but this was not significant
(P = 0 0 6 ).
Loss o f h eterozygosity ( L O H )
LOH analysis could not be undertaken because in­
stability obscures the determ ination o f allelic loss, since
m utant alleles formed as a result o f M I* m ay m atch
exactly one o f the w ild-type alleles present in norm al
D N A o f the constitutionally heterozygous cases.
D IS C U S S IO N
Previous studies28-29 have suggested that carcinom as
displaying M I* can be grouped into four classes;
0 1998 John Wiley & Sons. Ltd.
tum ours associated with H N PC C that show alterations
in m ultiple loci; sporadic colorectal cancers that also
show instability at m ultiple loci; other types o f sporadic
cancers (including breast) that show fewer and less
dram atic m icrosatellite alterations; and tum ours which
display instability at m ononucleotide repeats.
iFive (22 per cent) tum ours showed alterations at
m ultiple loci, four o f which were unstable at d inu cleo­
tide and trinucleotide repeats and may be sim ilar to the
second class o f M I* , described above. The fifth case was
unstable at the long m ononucleotide repeats and
resembles the fourth type o f instability. T hese data
support the previously published study which described
a similar phenotype in 13 per cent (3/23) o f D C IS. '0 The
cases that showed instability at m ultiple loci were
w orthy o f m utation analysis o f candidate M M R genes,
particularly in the light o f the recent identification o f a
frameshift m utation in h M L H l in a breast cancer sh ow ­
ing widespread M I * .31 This was the first evidence that
breast cancer, albeit as an integral tum our within the
H N PC C syndrom e, may result from the inheritance o f a
m utant M M R gene. Exam ination o f protein expression
by
im m unohistochem istry.
on
paraffin-em bedded
sam ples, has been shown to be a rapid m ethod for
pre-screening tum ours for m utations in the M M R genes,
in both hereditary and sporadic tum ours. T o date, the
m ajority o f inactivating m utations o f h M S H 2 have led
to a lack o f expression, or the expression o f a truncated
protein not detectable by the antibody used in this
stud y.32 The antibody to M SH 2 gave clearly defined
staining whereas the detection o f PM S2 and M L H I was
m ore variable, which may relate to the ability o f the
antibodies to detect the proteins; they are less well
characterized. A ll the M I* tum ours in this study dem ­
onstrated expression o f M SH 2. M L H I, and PM S2, and
there were no significant differences between staining for
M M R protein between cases show ing M I* and those
w ithout alterations to m icrosatellites. It is possible that
aberrations in the other M M R genes are responsible for
the observed M I*.
The cases in this study that show ed less widespread
instability, that is. alterations only at the D M -1 repeat,
and those in our previous stu d y 11 with a sim ilar alter­
ation. are unlikely to reflect aberrations in the h M SH 2
and h M L H l genes, since H N P C C tum ours with M M R
gene m utations show widespread instability at all types
o f loci exam ined. However, colorectal cell lines that
contain m utations in h M SH 2 and h M L H l sh ow very
similar or identical DM-1 repeat lengths to those o f
m ism atch proficient cell lines,33 suggesting that instabil­
ity at this locus is unusual in H N P C C . Other candidate
m utator loci need to be investigated in these breast
cancers. Based on the enzym ology o f m ism atch repair in
yeast, inactivating m utations in the hum an h om ologu e
o f M S H 3 (h M S H 3 ) might be predicted, since m utations
in M S H 3 d o not cause the extrem e m icrosatellite in­
stability and spontaneous m utability observed in M S H 2
m utants.34
It is also possible that the D M -1-specific instability
indicates a role for D M P K in breast cancer. T he DM -1
repeat is located in a 3' region o f the D M P K gene which
is transcribed but not translated.35 A lterations in the
JOURNAL OF PATHOLOGY. VOL.
185: 18-24 (1998)
23
M ICROSATELLITE i n s t a b i l i t y IN DCIS
num ber o f CTG„ repeats may alter m R N A stability or
the translational controls that norm ally function to
regulate D M P K . This may be w orthy o f further
investigation, particularly as D M P K has been identified
as an orthologue o f the Drosophila wartslluts tumour
suppressor gene. '6 and an association between defects
in human m ismatch repair genes and those o f
transcription-coupled excision repair has been recently
described.37
M ost human colon cancers with the m utator pheno­
type. and som e other M I* tum our types, have
fram eshift m utations in the m ononucleotide repeats
within the TCF/iRII. IG FIIR, and B A X coding regions
which lead to truncation o f gene products and result in
non-fu nctionin g proteins.21 24 In the present study,
insertions or deletions were not observed at these
repeats, even in the case o f D C IS with alterations to
other m ononucleotide repeats. The E2F-4 repeat can
also be altered in M I* gastrointestinal tum ours.25-26
This repeat is polym orphic in the norm al human popu­
lation. and therefore a functional effect o f m utation in
this region cannot be assum ed. H ow ever, no tumourspecific alterations o f the repeat were observed, even in
the in situ carcinom as with alterations at the polym or­
phic DM -1 trinucleotide repeat. It is possible that breast
cancers, with this phenotype, target different cancerassociated genes from those described in gastrointestinal
cancers. T he oestrogen receptor (ER ) represents a par­
ticularly good candidate gene in breast cancer. A recent
study has reported that m ononucleotide and trinucleo­
tide repeats in exon 1 o f the ER gene are not subject to
instability in M I* gastrointestinal cancers.37
T he biological behaviour o f D C IS appears to differ
between the histological subtypes and relate to nuclear
grade. There is a significant association between high
nuclear grade and the likelihood o f recurrence, and
progression to invasive d isease.16 M icrosatellite instabil­
ity w as detected at a higher frequency in the high- and
interm ediate-grade cases, compared with the low -grade
cases. Recent studies have observed that M I* in invasive
breast carcinom as correlated with indicators com m only
associated with poor disease p rogn osis.13 ,4 In our pre­
v io u s stu d y 11 o f m amm ographically detected, sm all,
n ode-n egative, grade I and II invasive carcinom as, there
w as a low frequency o f M I* (5 per cent). The overall
findings indicate that in situ and invasive breast carci­
nom as. instability is associated with high nuclear grade
and an aggressive phenotype. This is supported by the
finding o f an association between c-erbB-2 and MI in the
present study, since there is a strong correlation between
c-erbB -2 in D C IS and high grade. Im m unohistochem ically reactive p53 represents stabilized p53, which
in those cases with greater reactivity is m ore likely to be
due to a m utation.39 Six o f the tum ours did display
stron g nuclear reactivity and five o f these showed in­
stability, but four o f the five M I* cases identified
show ed no im munoreactive p53. This is consistent with
colorectal studies which have previously dem onstrated
that M I* cancers have significantly fewer p53 m utations
than MI “ cancers.40
W e chose to study pure non-invasive ductal carci­
nom a since it represents an early stage o f the disease.
C 1998 John Wiley & Sons. Ltd.
N ot only is it a heterogeneous condition, but within an
individual case there may be differences in the m orpho­
logical and biological features o f individual ducts. This
is illustrated by our findings o f heterogeneity in the
detection o f c-erbB-2 and p53. O f interest was the
observation that for the five cases which had multiple
instability, it was detected in all ducts exam ined per
carcinom a, suggesting that this alteration must be early
in the developm ent o f malignancy in these patients.
There were two cases with instability only at DM-1 that
showed heterogeneity in instability between the ducts
examined. This was not reflected in the findings for
c-erbB-2 and p53, where the cases showed h om ogeneous
or heterogeneous reactivity, or were negative. This is
further evidence for DM-1 instability being different
from M I*. The finding o f heterogeneity suggests that it
occurs at a stage whereby individual foci in in situ
cancers develop genetically divergent populations 41
In conclusion, we have identified m icrosatellite in­
stability (M I*) at multiple foci in 3 o f 23 cases o f DC IS,
but have not found changes in M M R expression or
alterations at microsatellite repeat m otifs in the coding
regions o f four cancer-associated genes, suggesting that
M I* in breast differs from that found in colorectal
cancers.
ACKNOW LEDGEM ENTS
This work was supported by the U niversity o f
Leicester and the Glenfield H ospital Trust Research
Comm ittee.
R EFER ENC ES
1. Thibodeau SN, Bren G. Schaid D. Microsatellite instability in cancer o f the
proximal colon. Science 1993: 260: 816-819.
2. Karran P Microsatellite instability and DNA mismatch repair in human
cancer. Semin Cancer Biol 1996: 7: 15-24.
3. Thibodeau SN, French AJ. Roche PC. et at. Altered expression of hM SH2
and hM LH l in tum ours with microsatellite instability and genetic alter­
ations in mismatch repair genes. Cancer Res 1996; 4836-4840.
4. Eshleman JR. M arkowitz SD. Microsatellite instability in inherited and
sporadic neoplasms. Curr Opin Oncol 1995; 7: 83-89.
5. Yee CJ. Roodi N. Verrier CS. Pari FF. Microsatellite instability and loss o f
heterozy gosity in breast cancer. Cancer Res 1994: 54: 1641-1644
6. W ooster R. Cleton-Jansen A-M , Collins N, et al. Instability o f short tandem
repeats (microsatelliies) in human cancers. Suture Genet 1994; 6: 152-156.
7. Patel U. Grundfest-Broniatowski S. G upta M, Banerjee S. Microsatellite
instabilities at five chromosomes in primary breast tumours. Oncogene 1994:
9: 3695-3700.
8. Glebov OK. McKenzie KE. White CA. Sukumar S. Frequent p53 gene
m utations and novel alleles in familial breast cancer. Cancer Res 1994; 54:
3703-3709.
9. Contegiacomo A. Palm irotta R. DeMarchis L, et al. Microsatellite instabil­
ity and pathological aspects o f breast cancer. Int J Cancer 1995; 64:
264-268.
10. Jonsson M. Johannsson O. Borg A. Infrequent occurrence o f microsatellite
instability in sporadic and familial breast cancer. Eur J Cancer 1995; 31A:
2330-2334.
11. Shaw JA. Walsh T, Chappell SA. Carey N, Johnson K. Walker RA.
Microsatellite instability in early sporadic breast cancer. Br J Cancer 1996;
73: 1393-1397.
12. Toyam a T. Iwase H. Yamashita H. et al. Microsatellite instability in
sporadic human breast cancers. Int J Cancer 1996: 68: 447-451.
13. Paulson TG. Wright FA. Parker BA. Russack V, Wahl G M . Microsatellite
instability correlates with reduced survival and poor disease prognosis in
breast cancer. Cancer Res 1996; 56: 1641-1644
14. De Marchis L. Contegiacomo A. D Amico C. et al. Microsatellite instability
is correlated with lymph nodc-posim e breast cancer. Clin Cancer Res 1997;
3: 241-248.
JOLRNAL OF PATHOLOGY, VOL.
185: 18-24 (1998)
24
T. W ALSH E T AL .
15. Pcltomaki P. Lothe RA. Aaltonen LA. et al. M icrosatcllite instability is
associated with tum ours that characterise the hereditary non-polyposis
colorectal carcinoma syndrome. Cancer Res 1993: 53: 5853-5855.
16. Bobrow LG. Happcrtield LC. Gregory W M. Millis RR. D uctal carcinom a
hi situ: assessment o f necrosis and nuclear morphology and their association
with biological m arkers. J Pathol 1995: 171: 333-341.
17. Lagios MD. W estdahl PR. M argolin FR . Rose MR. R elationship o f extent
o f non-invasivc disease to the frequency o f occult invasion, multicentricity .
Ivmph node metastasis and short term treatm ent failures. Cancer 1982: 50:
1309-1314.
18. Bartkova J. Barnes DM . Millis RR. Gullick WJ. Imm unohistochemical
dem onstration o f c-erhB-2 protein in mam m ary ductal carcinom a in situ.
Hum Pathol 1990; 21: 1164-1167.
19. M unn KE. W alker RA. Menascc L. Varley JM . M utation o f the TP53 gene
and imbalance at chrom osom e I7 p l3 in ductal carcinom a in situ. Br J
Cancer 1996: 74: 1578-1585
20. Loeb LA. M icrosatcllite instability: m arker o f a m u tato r phenotype in
cancer. Cancer Res 1994; 54: 5059-5063
21. MyerotT LL. Parsons R. Kim S-J. et al. A transform ing grow th factor
receptor type II gene common in colon and gastric but rare in endom etrial
cancers with microsatellite instability. Cancer Res 1995: 55: 5545-5547.
22. Souza RF. Appel R. Yin J. et al. M icrosatellite instability in the insulin-like
growth factor II receptor « n e in gastrointestinal tum ours. Suture Genet
'|996: 14: 255-257.
23. O uyang H. Shiwaku H. Hagiwara H. et al. The insulin-like grow th factor II
receptor gene is m utated in genetically unstable cancers o f the endom etrium ,
stomach and colorectal. Cancer Res 1997; 57: 1851-1854
24. R am pino N. Yam am oto H. Ionov Y. et al. Som atic frameshift m utations in
the BAX sienc in colon cancers o f the microsatcllite m u tato r phenotvpe.
Science 1997; 275: 967-969.
25. Y oshitaka T. M atsubara N. Ikeda M. el al. M utations o f the E2F-4
trinucleotide repeat in colorectal cancer with microsatellite instability.
Biochem Biophvs Res Commun 1996; 227: 553-557.
26. Souza R F. Yin J. Smolinski KN. et al. Frequent m utation o f the E2F-3 cell
c v d e gene in pnm arv hum an gastrointestinal tum ours. Cancer Res 1997: 57:
2350-2353.
27. Liu B. Farrington SM . Petersen G M . et al. Genetic instability in the
majority o f voung patients with colorectal cancer. Suture Med 1995: I:
348-352.
C 1998 John Wiley & Sons. Ltd.
28. Aaltonen LA, Pcltomaki P. Leach FS. et al. Clues to the pathogenesis o f
familial colorectal cancer. Science 1993. 260: 812-816.
29. Papadopoulos N. Nicolaides NC. Liu B, et al. M utations o f GTBP in
genetically unstable cells. Science 1995: 268: 1915-1917.
30. Aldaz MC, C hen T. Sahin A. C unningham J. Bondy M. C om parative
allelotype o f in situ and invasive hum an breast cancer: high frequency of
miqrosatellite instabilitv in lobular breast carcinomas. Cancer Res 1995: 55:
3976-3981.
31. Risinger JI. Barrett CJ, W atson P. Lynch HT. Boyd J. M olecular genetic
evidence of the occurrence o f breast cancer as an integral tum our in patients
with the hereditarv nonpolvposis colorectal carcinom a syndrome. Cancer
1996: 77: 1836-1843.
32. Leach FS, Polyak K. Burrell M. et al. Expression o f the human mismatch
repair gene hMSH2 in normal and neoplastic tissues. Cancer Res 1996; 56:
235-240.
33. Kram er PR. Pearson CE, Sinden RR. Stability o f triplet repeats o f
m votonic-dystrophv and fragile-X loci in hum an m utator mismatch repair
cell-lines. Hum Genet 1996: 98: 151-157.
34. Johnson RE, Kovvali G K . Prakash L. Prakash S. Requirement o f the yeast
MSH3 and MSH6 nenes for MSH2-dependent genomic stability. J Biol
them 1996: 271: 7285-7288.
35. Brook DJ. M cCurrach ME. Harley H G . et al. M olecular basis o f m yotonic
dystrophy: expansion o f a trinucleotide (CTG) repeat at the 3’ end o f a
transcript encoding a protein kinase family member. Cell 1992: 68: 799-808.
36. W atson KL. Drosophila W A RTS-lum our suppressor and member o f the
myotonic dystrophy protein kinase family. Bioessays 1995: 17: 673-676.
37. Mellon I. Rajpal DK. Koi M. Boland RC. C ham pe G N . Transcriptioncoupled repair. Science 1996: 272: 557-560.
38. Simms LA. Zou T-T. Young J. et al. A pparent protection from instability of
repeat sequences in cancer-related genes in replication error positive gas­
trointestinal cancers. Oncogene 1997; 14: 2613-2618.
39. G retasdottir S. Tryggvadottir L. Jonasson ?, et al. TP53 m utation analyses
on breast carcinomas: a studv o f paraffin-embedded archival material. Br J
Cancer 1996; 77: 555-561.
40. Kim H. Jen J. Vogelstein B. H am ilton SR. Clinical and pathological
characteristics o f sporadic colorectal carcinom as with DN A replication
errors in microsatellites. Am J Pathol 1994: 145: 148-156.
41. Fujii H. M arsh C. Cairns P. Sidransky D. G abrielson E. Genetic divergence
in the clonal evolution o f breast cancer. Cancer Res 1996: 56: 1493-1497.
JOLRNAL OF PATHOLOGY. VOL.
185: 18-24 (1998)
C a n c e r a n d M etastasis Review s 16: 5-27. 1997.
£ 1997 K lu w e r A ca d em ic Publishers. P rin ted in the N eth erla n d s.
M o l e c u l a r p a th o l o g y o f b r e a s t c a n c e r a n d its a p p lic a tio n to c lin ic a l
m anagem ent
R o s e m a r y A . W alker. J. L o u ise J o n e s. S te p h e n C h a p p e ll. T om W alsh and Jacqueline A . Shaw
D e p a r tm e n t o f P athology, U n iv e rsity o f L eicester. B r e a s t C a n cer R esearch Unit. C linical Sciences. G len field
H o s p ita l N H S Trust, G r o b y R o a d , L e ic e s te r L E 3 9 Q P , U K
K e y w o r d s: b reast cancer, g e n e tic a n a ly sis, s te r o id r e c e p to r , ep id erm al grow th factor receptor, cell ad hesion,
p r o te a s e s
A b s tr a c t
B r e a s t c a n c e r is a m ajor cau se o f m o r b id ity a n d m o r ta lity in w o m e n in m any parts o f the w orld. Breast carcino­
m a s are h eter o g en o u s in th eir b io lo g ic a l an d c lin ic a l b eh a v io u r and a greater understanding o f h o w they
d e v e lo p an d progress could lea d to m o r e d ir e c te d fo r m s o f screen in g and therapy. It is im portant to d eterm in e
th e m o le c u la r m echanism s u n d e r ly in g the n a tu ra l h is to r y o f breast cancer.
D e v e lo p m e n ts in the tec h n iq u es fo r m o le c u la r a n a ly sis h a v e m eant that they can now be applied to a large
r a n g e o f clinical m aterial su ch as c y to lo g ic a l p r e p a r a tio n s and fixed, em b ed ded m aterial, so increasing the
p o te n t ia l for relating any m o le c u la r a lter a tio n s to c lin ic a l b eh aviou r and response to therapy.
In th is review w e co n sid er r e c e n t d e v e lo p m e n ts in th r ee areas o f im portance to breast cancer: gen etic
a n a ly sis - o n co g en es, tum our su p p r e sso r g e n e s , lo s s o f h eterozygosity, m icrosatellite instability, fam ilial
b r e a s t can cer: steroid recep tors, o e s tr o g e n r e g u la te d p r o te in s, epiderm al grow th factor receptor, grow th fac­
to rs p articu larly transform ing g r o w th factor b eta : a n d c e ll ad h esion , in vasion and m etastasis - E -cadherin.
in te g r in s. p roteases. T h ese are d isc u sse d in r e la tio n to p o te n tia l for screen ing, prognosis and treatm ent.
I n tr o d u c tio n
B r e a s t c a rc in o m a is the m ost c o m m o n m a lig n a n c y
in w o m e n in N o rth A m e ric a a n d W e s te rn E u r o p e
a n d is a m a jo r cause of m orb id ity a n d m ortality. A l ­
t h o u g h th e r e have been advances in t r e a tm e n t o n ly
m o d e s t im prov e m e nts in survival have b e e n
a c h ie v e d . O n e of the im p orta n t p o in ts to a p p r e c ia te
a b o u t b r e a s t cancers is that they ar e h e te r o g e n e o u s
'in th e w av they behave both biologically a n d c lin ­
ically. It is. therefore, essential th a t th e n a tu r e o f in ­
d iv id u a l tu m o u rs is characterised a n d used to p la n
th e m o s t a p prop ria te therapy. A g r e a te r u n d e r ­
s t a n d i n g of how breast cancers d e v e lo p a n d h o w
th e y p ro g re s s could lead to m o re d ir e c te d forms o f
s c r e e n in g a n d therapy. This is w hy it is necessary' to
d eterm in e the molecular mechanisms underlying
th e natural history of breast cancer.
T h e techniques w'hich have been available for
analysis have changed considerably so that many
can now be applied to routine, diagnostic histopathological material and to aspirates/small biopsies.
T h e various approaches which can be used will be
re fe rre d to in the different sections where relevant.
A sum m ary of the possible wavs of analysing breast
cancers is given in Table 1.
It is not possible to cover every aspect of breast
ca n c e r and we will concentrate on three areas.
H ow ever, for each of these the concern is to consid­
e r the following:
• can alterations be identified which represen t
early events and have potential for screening
6
•
•
d o a n y a lter a tio n s/a b n o rm a lities p ro v id e in fo r­
e v e n t in the d evelopm en t o f tum ours [2], the ev i­
m a tio n a b o u t h ow individual ca n ce rs w ill b e h a v e
d e n c e from hum an breast carcinom as is less conclu­
c a n th e m o le c u la r alterations b e u sed for the s e ­
siv e . A ltera tio n s to the c-m yc g en e, predom inantly
le c t io n o f th e m ost ap propriate form o f tr e a t­
a m p lifica tio n , have been found in approxim ately
m e n t o r b e targets for n ew form s o f therapy.
25% o f carcinom as but whilst B on illa et al. [3] con­
sid e r e d this to be associated w ith the d evelopm ent
o f b reast cancer, other studies have found c-m yc al­
G e n e t ic a lte r a tio n s
te r a tio n s to correlate with aggressive features and/
o r p o o r p rogn osis [4-6]. W h en c-m yc m R N A and
O ncogen es
p r o te in is d eterm in ed in carcinom as by N orthern
a n a ly sis, in-situ hybridisation, im m unohistochem is-
A lt e r a tio n s to a n um ber o f p r o to -o n c o g e n e s a p p ea r
try and flow cytom etry it is ev id en t that there is a
to b e o f im p o r ta n c e in a p ro p o rtio n o f b reast c a n ­
ce rs .
p o o r correlation b etw een lev els and extent o f ex­
p ressio n and am plification o f the gene [7-8] and
h e n c e th ese approaches are o f n o value in determ in­
in g prognosis.
c-m y c
T h is e n c o d e s a n uclear p h o sp h o p r o te in w h ich acts
as a tra n scr ip tio n a l regulator, c o n tr o llin g c e ll p ro lif­
ra s
e r a tio n , d iffer en tia tio n and a p o p to sis [1], W h ilst
A lth o u g h activation of the ras proto-on cogene fam ­
s tu d ie s o f c -m y c in m ouse sy stem s su g g e st that a l­
ily is im portant in rodent m am m ary tumours the
te r a tio n s to th e g en e m ay be an im p o r ta n t early
r o le o f ras g en es in human breast cancer is less clear.
Table1. T he various techniques which can be applied to breast cancer sam ples of different types
DNA
M eth o d
D e tec t
Application
S o u th ern blotting
A m p lificatio n , d e le tio n , allelic loss.
im balance
A m plification, d e le tio n allelic loss.
im balance, m ic ro sa tellite instability.
Fresh, frozen tissue.
P C R amplification
Flow cytom etry
Ploidv status.
F lu o rescen t and non-radioactive in-situ D eletions, allelic loss, aneusomv.
hybridisation
RNA
P ro te in
.
N o rth e rn blotting
R T -PC R
O ver e x p ressio n , m u tatio n
E xpression ( can be quantified)
R T -PC R with SSCPJ C D G E 5
V ariants. M u ta tio n s.
In -situ hybridisation
E xpression w ith localisation.
M olecular w eights of proteins, plus
relative levels.
Im m unohistochem istry. with o r w ithout L ocalisation of p ro tein . Extent of
reactivity e.g. receptors.
antigen retrieval
Fresh, frozen tissue. Fixed, paraffin
embedded tissue. Fine needle
aspiration.
Fresh, frozen tissue, fixed, paraffin
embedded tissue. Aspirates (if
sufficient cells).
Fixed, paraffin em bedded tissue. Fine
needle aspirates.
Fresh, frozen tissue.
Fresh, frozen tissue. Fine needle
aspirates.
Fresh, frozen tissue. Fine needle
aspirates.
Fresh, frozen tissue. Fixed, embedded
tissue.
W estern blotting
Fresh, frozen tissue. Fixed, embedded
tissue. Fine needle aspirates.
A b b re v ia tio n s : J SSCP - Single strand conform ational polym orphism : ' C D G E - Constant denaturing gel electrophoresis.
7
M u ta tio n s h a v e rarely b een id en tified [9]. a lth o u g h
lo s s o f o n e H-ras-1 allele has b e e n co r re la ted w ith
su lt in a greater incidence o f low level staining and
a g g r e ssiv e fea tu res [10].
so m e stu d ies failed to find a relationship b etw een
Im m u n o h isto ch em ic a l stu d ies give co n flic tin g r e ­
su lts re g a rd in g ras protein ex p re ssio n , so m e o f
o v er-ex p ressio n and prognosis [21]. or only for n o d e
p o sitiv e cases [25]. many have found c-erbB -2 over­
lo ss o f the correlation with am plification. A lth o u gh
w h ic h m ay resu lt from the lack o f sp ec ificity o f th e
e x p re ssio n to be an independent predictor o f p o o r­
a n tib o d y u se d by several grou p s. A lth o u g h so m e
er d isea se free interval and survival [26-30] and w e
s tu d ie s fo u n d greater exp ression in ca rcin o m a s [11—
c o n tin u e to d o so. There is n ow a w ealth o f an tib od ­
12]. w e d id n o t [13]. G oin g et al. [14] fo u n d an in ­
ie s availab le and a detailed analysis of them foun d
c r e a s e in sta in in g from n orm al to in -situ ca r c in o ­
m a s. b u t surprisingly foun d g rea ter sta in in g in
b le, w hich m ay account for som e o f the differen ces
m y o e p ith e lia l cells.
c o n cern in g the relationship b etw een c-erbB -2 and
O f g r e a te r sign ifican ce from a sc re en in g p o in t o f
th eir ability to detect over-expression to be varia­
p ro g n o sis [31].
v ie w c o m e s from the finding o f the tight lin k a g e o f
A m p lifica tio n and over-expression o f c-erb B -2
H -ras-1 to a m in isatellite lo c u s w hich c o n sists o f
d o e s ap pear to be associated with an aggressive
fo u r c o m m o n a lleles and sev era l rare a lle le s. T h e r e
fo rm o f breast cancer, so it was surprising that in
is a sig n ifica n t association o f the rare a lle le s w ith
c a n c e r an d as m any as 1 in 11 b reast can cers m ig h t b e
im m u n oh istoch em ical studies the protein could b e
d e te c te d in ductal carcinoma in situ [32]. M ore e x ­
a ttr ib u te d to this [15].
ten siv e stu d ies found c-erbB -2 expression in 4 0 60% o f cases but always associated with the high
c -e r b B -2
g rad e (co m ed o ) type [33-35]. A llred et al. [36] p ro ­
T h is p r o to -o n c o g e n e has p ro v ed to b e o f p articu la r
p o s e d that eith er over-expression of c-erbB -2 d e ­
in te r e st in h u m an breast can cer. A ls o c a lle d n eu o r
c rea ses as carcinom as evolve from in-situ to in va­
H E R 2 , it e n c o d e s a 1S5 k D tra n sm e m b ra n e g ly c o ­
p r o te in . that has exten sive h o m o lo g y w ith e p id e r ­
siv e or that m any invasive carcinom as arise d e n o v o
m a l g ro w th factor receptor and is a p u ta tiv e g ro w th
p ro p o se that there are several m olecular pathw ays
fa c to r r e c e p to r [16].
b y w hich b reast carcinomas arise and alteration to
A m p lific a tio n o f the g e n e is fo u n d in 2 0 -3 0 % in
b y m echan ism s not involving c-erbB -2. W e w ould
c-e rb B -2 is just one of them.
in v a s iv e ca rcin o m a s [5. 17-19] an d a c o r r e la tio n
K n o w led g e o f c-erbB-2 status can be o f value in
h a s b e e n fo u n d b etw een a m p lific a tio n a n d a g g r e s ­
d eterm in in g therapy since there is clear ev id en ce
s iv e fe a tu r e s and poor sh ort term p r o g n o sis, a l­
th at c-erb B -2 positive tum ours show a poor r e ­
t h o u g h n o t by all [20-21]. T h e vast m a jo rity o f
sp o n se to endocrine therapy [37-39].
t h e s e s tu d ie s h ave used S o u th e rn b lo ttin g t e c h ­
e n tia l p o ly m e r a s e chain re a ctio n sin ce it can b e a p ­
CCNDl
In terest in g en es on chrom osom e llqi.3 cam e origi­
p lie d to fix e d , em b ed d e d tissu e. L iu et al. [22] id e n ­
n a lly from retroviral studies in mice, w here the o n ­
t ifie d a m p lific a tio n in 21% o f b reast ca n c e r s b u t
c o g e n ic effect o f mouse mamm ary tum our virus is
n iq u e s b ut m o r e recent w ork has e m p lo y e d d iffe r ­
H u b b a r d et al. [23] found it at a m uch h ig h e r in ­
d u e to the transcriptional activation o f the cellu lar
c id e n c e .
Im p ortan tly, there is a g o o d co n c o r d a n c e b e ­
p r o to -o n co g e n e s :m-l and int-2 (now W n t-1 and
F gF-3). Initially, the ;m-2/FGF3 gene, which is at
tw e e n c -e r b B -2 gen e am p lification and o v e r -e x ­
llq l3 . was studied and shown to be am plified in 1 0 -
p r e ssio n o f m R N A and p rotein as d e te c te d b y
20% o f breast cancers. A n association was fou n d
N o r th e r n an alysis, in situ hybridisation and im m u n o h isto c h e m is tr y [7. 19. 24], C -erb B -2 can be read -
w ith the presence of oestrogen receptor [40] a l­
th ou gh others found that am plification was a sso c i­
:iy d e te c te d in fixed, paraffin em b ed d e d tissu e u sin g
im m u n o h isto ch em istry . It is our e x p er ie n c e that an-
ly show n that FGF3 is not expressed in hum an
tiuen retrieval should not be u sed , sin ce this can r e ­
b reast cancers and so the am plification was just a
ated with p oor prognosis [41.42], It was su b se q u en t­
s
u s e fu l in d ic a to r o f am plification o f a n oth er gen e in
[54] an d the o th er high expression with allele loss
th a t r e g io n o f th e ch rom osom e.
C C N D 1 is at l l q l 3 and en c o d e s cyclin D i [43].
[55]. O f sig n ifica n ce was that alterations were found
e ith e r in ad v a n ced cases [54] or in aneuploid. high S
T h is , w h e n c o m p le x e d with its a sso cia ted cyclin -d e-
p h a s e can cers [55]. suggesting that alterations to
p e n d e n t k in a se , con trols cell cycle p rogression in
R B are n o t an initiating event in breast cancer but is
G 1 b y p h o sp h o r y la tin g retin ob lastom a p rotein [44],
a n e v e n t occurring in an unstable genom e.
O v e r -e x p r e s s io n o f cyclin D l has b e e n fou n d in
b r e a s t c a n c e r c e ll lin es in both the p resen ce and ab ­
p53
s e n c e o f a m p lific a tio n and d v sregu lation o f ex p re s­
T h e r e is su bstan tial evidence that alterations to p53
s io n o f cy clin D l m ay be a p o te n tia l factor in th e
c a n p ro v id e inform ation about many aspects o f
b r e a st cancer.
p a th o g e n e s is o f b reast cancer [45].
A n t ib o d ie s h a v e b een gen era ted again st re co m ­
G e rm lin e p53 m utations have b een found in fam ­
b in a n t h u m a n cy clin D l. O ver-ex p ressio n has b e e n
ilie s w ith the Li-Fraum eni Syndrom e [56], a rare
id e n tifie d in b rea st cancers, b oth w ith an d w ith o u t
sy n d r o m e in w hich there is young onset sarcom a as­
a m p lific a tio n [46. 47]. In on e stu d y p a tie n ts w h o se
s o c ia te d w ith breast cancer, primary brain tum our
c a r c in o m a s c o -e x p r e sse d cyclin D l w ith ep id erm a l
o r le u k a e m ia in a first degree relative under the age
g r o w th fa cto r r e ce p to r w ere fo u n d to h ave a p o o rer
o f 45 years. T h e m utations w ere initially found in
p r o g n o s is [48]. B a rn es et al. [49] fo u n d a g o o d cor­
e x o n 7. but h a v e subsequently b een found in other
r e la tio n b e tw e e n im m u n o h isto c h e m ic a lly d e te c ta ­
a r e a s in the con served region. H ow ever, only about
b le c y c lin D l an d o estro g en re cep to r an d n o te d th at
h a lf th e L i-F raum eni fam ilies have p53 m utations
c y c lin D l w as a v ery g o o d m arker o f lik ely re sp o n se
[57] an d g erm lin e p53 m utations are rarely found in
to e n d o c r in e therapy. U sin g a d iffer en t m o n o c lo n a l
c a s e s o f ea rlv -o n set breast cancer and those w ith a
a n tib o d y an d p ressu re-co o k er a n tig en retrieval
s tr o n g fam ily history [58. 59].
m e t h o d s , w e h a v e foun d a sim ilar co rrela tio n b e ­
A lth o u g h allelic loss on 17p and in the region o f
t w e e n cyclin D l and o estro g en r e ce p to r (p < 0.001).
th e p 53 g en e has b een reported by many groups, it
It c o u ld clea rly form an altern ative to p ro g este ro n e
w o u ld ap pear that the presence and nature of m uta­
r e c e p to r or p S 2 as a m arker o f h o r m o n e re sp o n siv e ­
tio n s p ro v id e m ore inform ation that is o f clinical
n e ss.
v a lu e.
S e v e r a l approaches have been m ade for id en tify­
in g p 53 m u tation s. T hese have included generating
T u m o u r s u p p r e s s o r genes
c D N A from R N A or starting from D N A : su b se­
q u e n t sin gle-stran d ed conform ation polym orphism
R e tin o b la s to m a gen e
a n a ly sis or co n sta n t denaturing gel electrophoresis
R e t in o b la s to m a g e n e ( R B 1) is the classical e x a m p le
w ith se q u en cin g to confirm potential m utations, or
o f a tu m o u r su p p re sso r gene. It has b e e n lo c a te d o n
c D N A seq u en cin g . Most studies have focused on
1 3 q l4 [50] a n d e n co d es for a M rl0 5 0 0 0 p ro tein
s e q u e n c e ch an ges in exons 5. 6. 7 and 8 with highly
w h ic h in its u n p h o sp h o ry ia ted form restricts cell c y ­
c o n se r v e d d om ain s [60-67] although com plete s e ­
c le p r o g r e s s io n in G l. by interacting w ith E 2F tran ­
q u e n c in g has b een undertaken [68]. The oth er
sc r ip tio n fa cto r [51].
It w o u ld p o te n tia lly appear to be a ca n d id a te
te c t m u ta tio n s was im munohistochemistry. It is n ow
g e n e to b e a lte r e d in the early stages o f the d isea se.
e v id e n t that an increase in p53 expression can occur
A lt e r a tio n s to ch ro m o so m e 13q have b e e n foun d at
in re sp o n se to D N A damage and can be d etected by
im m u n oh istoch em istry ;69]. The incidence o f m u­
se v e r a l loci in b reast carcinom as [52] as have stru c­
m e th o d u sed which was initially considered to d e ­
e ith e r by im m u n o h isto ch em istry or im m u n o b lo t-
ta tio n s and im m unoreactive protein in the sam e
g ro u p o f carcinom as does differ [60. 70]. The m ain
a d v a n ta g e o f im m unohistochem istry is that m any
tin g . w ith o n e find in g loss o f p rotein w ith allele loss
a n tib o d ies can be applied to fixed, em bedded tissue
tu ra l a lte r a tio n s in the RBL gen e [53]. Tw o stu d ies
h a v e c o m p a r e d a llele loss and p rotein ex p re ssio n ,
9
a n d th e u se o f an tigen retrieval (m icro w a v e, p r e s­
su re c o o k e r ) im p ro ves results, but care has to b e
b a la n c es appear to be late events, w hereas changes
a ffectin g 7p. 16q, 17p and 17q appear to be early ab­
ta k e n in in te rp re ta tio n and it is b est to c o n sid er
n o rm a lities sin ce they have b een found in 25-30%
sta in in g to re p resen t stabilised, p o ssib ly m u tan t,
p r o te in .
o f d uctal carcinom a in situ cases [80]. Clearly id en ti­
fica tio n o f such changes at very early stages o f the
T h e r e is an associa tio n b etw e en the p resen ce o f
d ise a se cou ld be o f value as use as markers in w om ­
m u ta tio n s an d aggressive features w ith in b rea st
e n w h o are at increased risk.
c a r c in o m a s e.g. lack o f o estro g en r e ce p to r [62. 6 5 .
W e h ave b e e n studying ductal carcinom a in situ
6 6 ]. h igh S p h a se in d ex [64], A n d e r se n et al. [65]
a n d sm all, im palpable lymph n od e negative inva­
fo u n d a sig n ifica n t association b e tw e e n p53 m u ta ­
s iv e
tio n s an d d ise a s e -fr e e and overall survival. B erg h e t
sc re en in g , concentrating on ch rom osom e 6q since it
al. [68] w h o sc r e e n e d the w h ole c o d in g s e q u e n c e
is th e se co n d m ost frequent site for allelic loss. T h e
fo u n d d iffe r e n c e s in the sites o f m u ta tio n s b e tw e e n
6 q 2 5-27 region has been analysed using PC R analy­
n o d e p o sitiv e and n o d e n egative ca ses. M u ta tio n s in
sis o f four polym orphic m icrosatellite markers. Tu­
c o n s e r v e d re g io n s II and V w ere a sso c ia ted w ith a
m o u r fo ci and individual ducts from ductal carcino­
sig n ific a n tly w o rse progn osis. A lth o u g h im m u n o ­
m a in situ w ere m icrodissected from form alin-fixed,
h isto c h e m istr y m ay n ot alw ays id en tify m u ta tio n s
p araffin em b ed d ed sections and D N A was extract­
th e resu lts a lso relate to tum our ch aracteristics. It is
e d from fixed , em bedded corresponding norm al tis­
im p o r ta n t th at there are clearly d e fin e d c u t-o ff
su e for com parison.
carcinom as
detected
by
m am m ographic
p o in ts for d e fin in g p ositive and n e g a tiv e ca ses [71].
T h e m icrodissection ensures that there is no co n ­
P r o m in e n t rea ctivity for p53 is a sso c ia ted w ith lack
ta m in a tio n from non-tum our cells and can also
o f o e s tr o g e n receptor, p oor d iffer en tia tio n , h ig h
d em o n stra te heterogeneity w ithin individual tu­
p r o life r a tio n
rates, and p resen ce o f ep id er m a l
m o u rs. W e have found LO H in 48% of invasive car­
g ro w th fa cto r recep tor [70. 7 2 -7 4 ]. It has also b e e n
cin o m a s and 50% o f ductal carcinom a in situ, w ith
s h o w n to b e an in d ep en d en t m ark er o f p ro g n o sis
lo ss at sin gle and m ultiple loci. C hanges were ob ­
[72, 75. 76]. alth ou gh seq u en cin g has b e e n fo u n d to
se r v e d in all types of invasive and in situ carcinom as
2 iv e
w h ich su ggests that inactivation o f tum our suppres­
b e tte r p ro gn ostic in form ation th an the m o n o ­
c lo n a l a n tib o d y PablSO l [77].
B e s id e s b e in g o f value for the p red ictio n o f p r o g ­
n o sis. p 53 ca n aid in the se lec tio n o f therapy. B e rg h
so r g en es in this region on chrom osom e 6q could
o cc u r relatively early and w ould be an area w orthy
o f stu d y in 'at risk’ lesions.
e t a l. [68] fo u n d adjuvant tam o x ifen th era p y to b e o f
le ss v a lu e in p53 m utation lym ph n o d e p o sitiv e
c a s e s. R e s p o n s e to ch em o th era p y and ra d io th era p y
.V icroscitellite instability
ca n b e a ffe c te d by altered p53 fu n ction , d ue to its
r o le in re g u la tin g D N A dam age resp o n se [78].
T h e sim p le random repeat D N A sequences o f
m o n o -, di- and trinucleotides represent a very com ­
m o n and highly polymorphic class of genetic e le ­
L o s s o f h e te r o z y g o s ity
m en ts. w hich are used in gene m apping and linkage
T h e fr e q u e n t loss o f h etero zy g o sity (L O H ) at a c e r ­
ex p a n sio n or contraction o f repeat elem ents was
ta in c h r o m o so m a l locus in tum ours in d icates th a t
first reported in neoplasms in colorectal tum ours
an alysis. M icrosatellite instability, dem onstrated by
th is c o u ld be the sight o f a tum our su p p resso r g e n e .
[81]. In hereditary nonpolyposis colorectal carcino­
T h e r e h a v e b een m any studies w hich h ave e x a m ­
m a (H N P C C ) this instability has been shown to be
c a u sed by inherited and som atic m utations in D N A
in e d a v a rietv o f ch rom osom es tor L O H in series o f
b rea st ca n cers. S pecific allelic losses h ave b een r e ­
p o r te d for m an y ch rom osom es (re v iew ed by D e v ile e an d C o r n e lisse [79]). C ertain a llelic losses or im ­
m ism atch repair genes.
M icrosatellite instability has been found in breast
carcinom as. W ooster et al. [82] concluded that in­
10
s ta b ility o cc u r re d alm ost ex c lu siv e ly at h igher o rd er
S te ro id and grow th factor receptors, oestrogen
tri- a n d te tra n u c leo tid e rep eats and that in stab ility
re la ted regulated proteins and grow th factors
at m o r e th a n o n e locus w as rare. O th e r stu d ies [83.
S4] h a v e id e n tifie d tum ours sh o w in g in stab ility at
O estro g en recep to r
m o r e th an o n e locus, in clu d in g d in u cleo tid e r e ­
p e a t s . w h ich is rem iniscent o f that s e e n in H N P C C
D ete rm in a tio n
k in d r e d s. Y e e et al. [83] su g g ested that m icr o sa tel­
T h e ev a lu a tio n o f the oestrogen receptor status o f a
lit e in sta b ility is an early e v e n t in m am m ary tu m o u -
b rea st can cer has been used for over 20 years to
r ig e n e s is . w h ilst L O H m ay occu r at a later stage.
h elp d eterm in e the likely response to horm onal
W e h a v e u n d ertak en an an alysis o f m icr o sa tellite
therapy. O estro g en receptor (E R ) can be d etected
in sta b ility in early, m a m m o g ra p h ica lly -d etec te d .
in 6 0 -7 0 % o f breast cancers, o f w hom about a half
in v a s iv e ca r cin o m a s [85] and in situ carcin om as, u s­
(i.e. a third overall) will respond to horm onal m a­
in g m u ltip le m arkers. A sm all n u m b er o f in v a siv e
n ip u la tio n [91]. T h e predom inant assay m ethod has
c a r c in o m a s sh o w e d instability w ith 7% sh o w in g in ­
b e e n ligand binding in nature, in which radiola­
s ta b ility at m u ltip le loci. T h ere w as n o in stab ility in
b e lle d steroid is added to h o m ogen ised breast tu­
t u b u la r ca rcin o m a s, including ca ses w h ich p r e se n t­
m o u r cy to so l and binding d eterm ined after rem oval
e d clin ica lly . M icrosatellite in stab ility w as d e te c te d
o f free steroid by dextran-coated charcoal. It has
in in situ ca rcin o m a but o n ly w h e n m ic r o d isse c tio n
b e e n replaced in som e laboratories by an enzym e
im m u n oassay, utilising an antibody generated
w a s p e r fo r m e d . T he in cid en ce w as g rea ter in th e
h ig h g ra d e typ e. D iffe re n c es b e tw e e n d ucts fro m
a gain st the receptor. Both these approaches can be
t h e sa m e c a se w as found in 3 o f 23 ca ses. T h e d a ta
q u a n tified , but they do require fresh tissue which
s u g g e s t th at m icrosatellite in sta b ility is an ea r ly
can b e a lim iting factor.
e v e n t in th e g en esis o f so m e sp o ra d ic b reast ca n cers
T h e introduction of antibodies to E R has m eant
a n d th at it c o u ld have p o ten tia l as a sc re en in g to o l.
that im m unohistochem istry can be used. The initial
a n tib o d ies w ere used predom inantly on frozen se c­
tio n s [92], although with various enzym e pretreat­
m en ts. reactivity could be achieved in fixed m ateri­
F a m ilia l b r e a st cancer
al [93]. Further antibodies have b een generated and
T h e m a jo rity o f breast can cers are d u e to acq u ired
are availab le com m ercially which work on form a­
m u ta tio n s , w ith only 5 to 8% o f b reast can cer p a ­
lin -fix ed m aterial. T hese require antigen retrieval,
t ie n ts h a v in g a strong fam ily history, in d icative o f
e ith er by m icrow aving or pressure cooking in ci­
in h e r ita n c e o f m utations.
In h e r ite d early on set breast ca n cer has b een lin k ­
w ith q uan titative enzym e im m unoassay [94]. T h ey
e d to tw o g e n e s . B R C A 1 [86] and B R C A 2 [87]. T h e
clea rly have the advantage in that they can be ap­
g e n o m ic stru ctu re o f B R C A 1 is c o m p le x . C u rren tly
p lied to routine pathology m aterial and also to fine
th e p r o te in tru ncation test p rom ises to b e c o m e a
n e e d le aspirates or small biopsies. H ow ever, ev a l­
trate buffer, but give results which are com parable
v a lu a b le te c h n iq u e for d etectin g m u tation s. H o g e r -
u a tio n o f reactivity and defining cu t-off points
v o r st e t al. [88] used it to screen for m u tation s in
w hich w ill provide information about response to
e x o n 11 w h ich en co d es 61% o f B R C A 1 . H o w e v e r ,
h o rm o n e therapy, can cause problem s. It can only
t h e r e is n o hard evid en ce that th e se g e n e s are im ­
be sem iq u antitative. there are problem s in repro­
p o r ta n t in sp o ra d ic breast cancers. B eck m a n n et al.
d u cib ility [95] and there can be problem s in achiev­
8 9] in v e stig a te d LO H o f B R C A 1 and B R C A 2 in
sp o r a d ic
can cers
using
P C R -b ased
flu o r esc en t
D N A te c h n o lo g y and found that it w as n ot o f th e
ing con sisten cy in staining intensity. With care both
the m ore com p lex H score and sim pler scoring sys­
tem s can provide useful clinical inform ation [96].
sa m e p r o g n o stic vaiue as for fam ilial cancer. N o
3 R C A 1 m u ta tio n s have b een fou n d in sp orad ic
5 / 'zn ift can ce o f E R
c a s e s [90].
W hilst the main interest in oestrogen receptor has
11
b e e n its r o le in the clinical m a n a g e m e n t o f b r e a s t
to failure to express ER. H o w ev er Yaich et al. [103]
c a n c e r , th e re are several im portan t q u e stio n s su r­
fa iled to find any such association. A n d ersen et al.
r o u n d in g th e sign ifican ce o f E R in b reast can cer:
[104] analysed B st V I and P v u ll p olym orphism s
•
d o r e c e p to r s in norm al breast ep ith e lia l c e lls a n d
an d foun d alleles with the PvitW restriction site
m a lig n a n t cells have the sa m e stru ctu re a n d
fu n c tio n ?
w er e m ore frequent in patients with p rogesteron e
•
w h y are a prop ortion o f carcin om as E R n e g a ­
tiv e ?
a sso cia tio n betw een any polym orphism s and E R
statu s.
•
w h y d o a p rop ortion o f E R p o sitiv e ca n cers fa il
•
•
re ce p to r negative tumours, although there was n o
O n e m echanism s that could block transcription
to r e sp o n d to en d ocrine therapy?
o f th e E R g en e in E R negative tum ours w ithou t
w h y d o a sm all num ber o f E R n e g a tiv e tu m o u r s
structural alterations in the g en e is m ethvlation o f
r e sp o n se to horm one therapy?
cv to sin e-rich areas. CpG islands, in the 5' regulato­
w h a t is th e ro le o f E R in b reast ca n ce r d e v e lo p ­
ry reg io n o f the gene. H vp erm eth ylation has b een
m e n t an d p rogression ?
fo u n d in E R negative tum ours [105. 106] and ce ll
T h e c lo n in g o f E R D N A [97] has m a d e p o s s ib le
lin es [107]. Supporting the co n cep t that m ethvlation
stru ctu r a l an d functional an a ly ses o f E R w h ic h
can affect expression is the finding that d em eth yla-
c o u ld c o n tr ib u te to our u n d erstan d in g o f th e a b o v e
tio n o f the E R gene in E R n egative breast cancer
p r o b le m s.
ce ils can reactivate expression [108].
E R h as 6 co n serv ed dom ain s. A - F [98] w ith d if­
H o w ev er, the major focus o f research into the g e ­
fe r e n t fu n ction s: transcriptional a ctiv a tio n an d r e ­
n e tic analysis of E R has b een con cern ed with the
p r e s s io n . n u c lea r localisation . D N A b in d in g a n d
id en tifica tio n o f m utations or variants which m ay
h o r m o n e b ind in g. The D N A b in d in g d o m a in is r e ­
p lay a role in receptor dysfunction. T h e basis for
s p o n s ib le for the recognition o f sp e c ific e n h a n c e r
so m e o f these studies is because Sluvser [109] p ro ­
s e q u e n c e s fo u n d in h o rm o n e -r esp o n siv e g e n e s , e.g .
b e tw e e n
p o s e d that m utated or truncated form s o f the s te ­
roid recep tor family have o n cogen ic potential, w ith
d iffe r e n t h o rm o n e resp onse e le m e n ts is d e te r ­
ab errant form s com pleting w ith norm al receptor
p r o g e s te r o n e
receptor. D isc r im in a tio n
m in e d b y th ree am ino acids at th e b a se o f th e first
for binding to horm one response elem en ts, so in ­
z in c fin g e r o f the D N A binding r e g io n [99] so m u ta ­
terferin g with normal transcription m echanism s.
t io n s c o u ld h ave im portant fu n ctio n a l c o n s e q u e n c ­
O n e o f the first studies [110] used R N ase p ro tec­
e s. T h e E r e g io n , which con tain s b o th th e h o r m o n e
tio n assays to screen primary breast carcinom as and
b in d in g d o m a in and the region re q u ir ed for sta b le
id en tifie d m odifications in the B region o f E R
d im e r iz a tio n . is very com p lex an d m u ta tio n s w ith in
w h ich correlated with low levels o f oestrogen b ind ­
th is r e g io n co u ld have p rofoun d c o n s e q u e n c e s o n
ing. H o w ev er, this was subsequently found at a sim ­
r e c e p to r fu n ctio n .
ilar frequ en cy in both ER positive and E R negative
A lte r a tio n s to E R
sign ifican ce.
T h e r e a r e s o m e alterations which h ave b e e n id e n t i ­
fied w h ic h can contribute to explaining so m e o f th e
q u e s tio n s o u tlin e d above.
« A n a ly sis of D N A has shown no g o o d e v id e n c e
for a m p lific a tio n of the.gene in b re a s t cancers [100].
U sin g S o u th e r n blotting tech niqu es W atts et al.
[ L01 ] failed to find any gross r e a r ra n g e m e n ts .
S e v era l re p o rts have suggested that a p o ly m o r ­
p h is m asso c ia te d with the restriction e n d o n u c le a s e
P v u U was link ed to ER expression in h u m a n b r e a s t
c a n c e r [102] with absence of one allele being re la te d
A lteratio ns to the DNA binding domain could be
im p o r ta n t in determining receptor function. Scott
et al. [112] used gel-retardation assays to m easure
E R - D N A binding capacity and reported that two
thirds of cancers with high ER levels retain D N A
binding ability white most tumours with low or in­
term e d iate levels have lost this activity. Also, they
fo und an association between E R - D N A binding
a n d progesterone receptor. Both D N A binding an d
non D N A binding receptors could be found in the
sam e tum our demonstrating the heterogeneity o f
tu m o u rs [111], suggesting it may be of no functional
12
b r e a s t ca n c e r s. Fuqua et al. [113] id en tifie d a varian t
p a rticu la r phenotype. They also studied tum ours o f
w h ic h r e su lte d from a d ele tio n o f e x o n 3 o f th e
th e E R + P g R - phenotype and id en tified a truncat­
e d E R lackin g E xon 7. which was unable to induce
D N A - b in d in g d om ain such that the se co n d zin c fin ­
g e r w a s m issin g . A lth ou gh this w o u ld ap p ear to b e
im p o r ta n t as a d om in ant n eg a tiv e recep to r, it m a y
tra n scrip tio n and which p revented norm al E R
fu n ctio n .
b e a n a tu ra l alternative sp liced m ista k e w h ich is
f u n c tio n a lly insignificant.
a n ts/m u ta tio n s in tam oxifen resistance. K am ik et
M u r p h y a n d D o tz la w [114] id e n tifie d variant E R
al. [120] screen ed tam oxifen resistant and tam oxi­
m R N A s in b rea st cancer b io p sies, an d the sm a lle r
fe n se n sitiv e tum ours for m utations and identified a
v a r ia n ts a p p e a r e d to be m isin g so m e or all o f th e E
b a se-p a ir replacem ent and a base pair d eletion in
a n d F d o m a in s o f the receptor. S in c e th e se w o u ld
h a v e a lte r e d h o rm o n e bind in g so m e o f th em w er e
e x o n 6 in 2 o f 20 tam oxifen resistant tum ours. Since
th e se w ere at a low frequency they suggested that
c lo n e d a n d se q u e n c e d [115]. O n e, C lo n e 4, w h ich
m u ta tio n s did not account for tam oxifen resistance.
T h er e has b een interest about the role o f E R vari­
w a s p r e s e n t in m ultiple tu m ou rs, h a d se q u e n c e s
D a ffa d a et al. [121] investigated the exon 5 deletion
id e n tic a l to w ild type exon s 1 an d 2 b u t th en d i­
sp lic e variant in a similar group o f tum ours using
v e r g e d . T o d eter m in e w h eth er this varian t is o f im ­
R T -P C R . H igh er levels o f the variant w ere found in
p o r ta n c e . 106 carcinom as w h ich w er e ER-f- PgR-K
th e E R - P gR + or pS2-h group and there was signif­
E R r P g R - an d E R - P g R - w er e a ssa y e d u sin g an
ic a n tly greater variant exon 5 m R N A expression in
R N a s e p r o te c tio n assay. S ig n ifica n tly h igh er le v e ls
th e E R + p S2+ tam oxifen-resistant group. W hile the
o f c lo n e 4 varian t m R N A w ere fo u n d in P g R n e g a ­
v a ria n t is unlikely to be responsible for tam oxifen
tiv e tu m o u r s and in tum ours w ith m ark ers o f p o o r
re sista n ce in m ost breast cancers, it m ay be im por­
p r o g n o s is [116]. T his particular varian t co u ld b e im ­
ta n t in som e.
p o r ta n t in th e p rogression o f b rea st ca n cer fro m
h o r m o n e d e p e n d e n c e to in d e p e n d e n c e .
R o o d i et al. [122] screened D N A from E R p osi­
tiv e and n egative tumours for deletions/insertions
F u q u a e t al. [117] d ev elo p e d a se n sitiv e p o ly m ­
o r p o in t m utations, amplifying exons 1 through to 8.
e r a s e ch a in reaction iP C R ) a ssay to d e te c t E R
N o d eletion s/in sertion s w ere identified and only 2
m R N A u sin g sm all am ounts o f R N A . so as to se m i-
m u ta tio n s w ere found in the sam e E R negative tu­
q u a n tita te E R exp ression an d to id e n tify rare tran ­
m o u r. A polym orphism in codon 325 show ed a
sc r ip ts. U s in g this approach th ey id e n tifie d a varia ­
stro n g association with family history and this war­
n t la c k in g e x o n 5 o f the h o rm o n e b in d in g d o m a in ,
ran ts further study.
w h ic h w as th e p red om inan t E R R N A e x p r e sse d in
T h e stud ies to date go towards explaining som e
E R - P g R - tu m ou rs [11S] and co u ld a cc o u n t for th is
o f the q u estio n s posed e.g. the E R - PgR-f- pheno-
T a b le 2. S um m ary of the alterations identified in the E R gene
A lte ra tio n
Significance
Reference No
P o ly m o rp h ism
P v u II
A b sen c e o f o n e allele, failure to
e x p ress E R .
A b sen c e o f o n e allele, no effect.
A sso c ia tio n w ith PgR negative tumours.
A s so c ia ted w ith family history.
No fu n c tio n al significance.
N o fu n c tio n al significance.
H ig h er levels in PgR negative tumour, tumours with
m a rk e rs o f p o o r prognosis.
A sso c ia ted with E R - PgR+ phenotype
A sso c ia ted w ith tam oxifen resistance in E R - PS2tu m o u rs
102
C o d o n 325
M o d ifica tio n to B region
D e le tio n of exon 3
W iid tvpe exon 1 and 2 then em ergence (clone 4)
E xon 5 v ariants
103
104
122
111
113
lib '
I IS
121
F igure2. Four carcinomas (lanes A to D) analysed for the second
C ). T h e D N A binding domain has been analysed by R T -PC R
a n d single stranded conform ational polym orphism analysis.
T h e re is a b a n d shift in carcinom a B (to be sequenced).
h alf of the horm one binding domain by RT-PCR. The negative
control (E) and 100 bp ladder is shown. Carcinoma A has a weak
( upper) wild type band and a stronger lower variant band, carci­
nom a B only has the variant, and carcinomas C and D have both
wild type (600 bp) and variant.
ty p e , b u t th ere is still a lot to u n d erstan d ab o u t E R
lator. It is a useful predictor o f response to en d o ­
in n orm al cells, and E R fu n ction . T h e p u b lish ed d a ­
crin e therapy [91].
F igure I. F our carcinom as (lanes A . B. D. E) and E R cD N A (lane
ta regard in g the E R gen e is su m m arised in T a b le 2.
P gR can b e d etected by binding assays, similar to
W e h a v e in vestigated a grou p o f can cers d iffer en t
th o se used for E R providing a synthetic p rogesto­
fr o m th o se in published series. T h e se have b e e n
m a m m o g ra p h ica lly -d etected sm all in vasive ca rci­
g en is used. M onoclonal antibodies have been d e ­
v e lo p e d and these can be used reliably on form alin
n o m a s w h ich have a m uch h ig h e r in cid en ce o f E R
fix ed m aterial [123]. and on aspirates.
p o sitiv ity [123]. R N A has b e e n extra cted from fr o ­
A b se n c e o f PgR in breast cancers may be due to
z e n se c tio n s using a D y n a b e a d m eth o d , re v e r se -
d efec ts in E R function, as described above, or to
m olecu lar alterations in PgR. Fuqua et al. [124] did
tra n scrib ed and regions o f th e D N A b ind in g a n d
a m p lifie d by P C R .
n o t find any m ajor gene rearrangements to PgR in a
T h e s e h ave b een analysed by sin g le stran ded c o n ­
fo r m a tio n a l polym orphism an alysis. U sin g p rim ers
stu d y o f 132 tum ours and concluded that this m ech­
anism could not explain the lack o f PgR expression
to a m p lify a region covering th e D N A b ind in g d o ­
m a in in clu d in g the zinc fin gers, band shifts h a v e
for the m ajority o f breast tumours. A lthough L O H
has b een d etected within the PgR gene it did not
b e e n id e n tifie d in seven o f 45 carcin om as (F igu re 1).
co rrelate w ith loss of im m unoreactive PgR [125].
O n e carcin om a show ed a b an d sh ift w hen e x o n 5
L o ss o f P gR m ore likely relates to altered E R fun c­
w a s am p lified . Interestingly, 11 o f the 45 c a s e s
tion.
h orm one
binding d om ains
s h o w e d sp lice variants in the se c o n d h alf o f th e h o r ­
m o n e b in d in g dom ain w hen a region sp an n in g e x ­
o n s 6 to 8 w as am plified (F igu re 2). We are in th e
O estro g en regulated proteins
p r o c e ss o f seq u en cin g all ab n orm al patterns an d r e ­
la tin g th e abnorm alities to ch aracteristics o f th e tu ­
m o u r s su ch as PgR and p roliferation statu s. It
w o u ld b e o f value also to screen n on -in vasiv e ca r ci­
n o m a s an d ’at-risk' lesions such as atypical d u c ta l
h y p e r p la sia to determ ine w h e th er sim ilar a lter­
V arious o estrogen regulated proteins have b een
id en tified from the MCF-7 breast cancer cell line,
eith er by analysis o f culture medium after o estro ­
g en stim ulation or by differential screening o f
c D N A from horm one treated and untreated cells.
a tio n s are foun d in these ev e n earlier ch an ges.
T h e pS2 gen e was identified by the latter approach
P ro g e s te ro n e receptor
[126] and subsequently identified by other groups
and called pN R 2 [12_] and Md2 [128]. Expression o f
pS2 m R N A and protein is related to oestrogen re ­
cep to r w ithin breast carcinomas [129-131] and is a
P r o g e ste r o n e receptor (P g R ) is regulated by o e s ­
tro g e n actin g through ER and is itself a g en e regu -
u seful m arker o f potential hormone resp onsiven ess
[132]. W estlev’s group has developed a com p etitive
14
r e v e r s e tran scrip tion -p olym erase ch ain reaction for
q u a n tita tio n o f pS2 ( and E R and P g R ) w hich can be
g r o w th o f norm al breast epithelium , acting through
its r e ce p to r (E G F R ) which is a transm em brane gly­
u s e d to m e a su re exp ression in sm all n um bers o f
c o p r o te in . T his has an extra cellular ligand binding
c e lls o b ta in e d b y fine n eed le asp iration [133]. T h is
ty p e o f a p p r o a ch is o f particular v alu e for tum ours
d o m a in and an intracellular tyrosine kinase d o ­
m a in . w ith clo se similarity to v-erb B [145]. High lev ­
in e ld e r ly p a tie n ts.
e ls o f E G F R , as determ ined by binding assays o f
p S 2 is a tr e fo il p ep tid e [134] and its fu n ction r e ­
h o m o g e n a te s
and im m unohistochem istry, have
m a in s u n clea r. It can be found in tissu es w hich are
b e e n sh o w n to correlate with a poor prognosis and
o e s tr o g e n u n r esp o n siv e e.g. gastric m u cosa and
fa ilu r e to respond to horm one therapy [146-149]. In
h e r e it m ay b e in v o lv ed in healin g. T h e g en e is also
m o s t tum ours there is an inverse relationship with
r e s p o n s iv e to ep id erm al grow th factor. T P A . c-H -
E R . E G F R can provide useful clinical inform ation.
r a s an d c-ju n [135]. T h ere m ay be factors oth er than
T h e in creased expression is n ot due to gene am ­
o e s tr o g e n reg u la tin g its ex p re ssio n in breast ca n ­
p lifica tio n . Increased levels o f E G F R m R N A are
c e r s w h ich m ay accou n t for the c o m p le x stain in g
fo u n d [150]. W e have undertaken in-situ hybridisa­
p a tte r n s e e n w ith im m u n o h isto ch em istry and th e
tio n fo r E G F R m R N A in both benign and m alig­
fin d in g o f pS2 in E R n egative tu m ou rs [136].
n a n t b reast tissue using a d igoxigenin-labelled o li­
O th e r p ro tein s which w ere o rig in a lly id en tifie d
as b e in g o e s tr o g e n regulated ap p ea r to b e m arkers
g o n u c le o tid e
probe.
In
norm al/benign
tissue
m R N A is d etectab le to a greater extent than pro­
o f a g g re ssiv e b eh aviour, w hich co n tra sts w ith E R .
te in as dem onstrated by im m unohistochem istry,
C a th e p sin D . a m ajor lysosom al p ro tea se , w as in i­
b u t this difference is more striking in the carcino­
tia lly id e n tifie d in the m ed ium o f M C F -7 cells cu l­
m a s. Two thirds o f the cancers assessed had d em on ­
tu r e d in the p resen ce o f o e str o g e n [137]. H o w ev e r,
str a b le m R N A (Figure 3). w hereas only one third
it is c o n stitu tiv e lv overex p ressed in E R n eg a tiv e
h a d d etecta b le protein. Those cases with both w ere
b r e a s t ca n cer cell lines and se v er a l stu d ies h av e
all E R and P gR negative, but the group with E G F R
fa ile d to find a relationship b e tw e e n cath ep sin D
m R N A but no protein com prised a m ixture o f E R +
a n d E R in prim ary breast carcin om as [131.138-140].
P g R ^ . E R - P g R - cases, and E R -P gR . It has b een
T r a n sfe c tio n o f cath ep sin D re su lte d in tran sform ed
su g g e ste d [151] that regulation o f the E G FR gene
c e lls in cr ea sin g their m etastatic ca p a city [141]. C lin ­
ica l stu d ie s, u sing im m u n oassays, h a v e su g g ested
in v o lv e s a num ber o f interactions b etw een positive
re g u la to r y factors and repressors (som e of which
th a t c a th e p sin D is a m arker o f p o o r p ro g n o sis [138—
m a y b e oestro g en regulated) with binding sites in
140] b u t im m u n o h isto ch em ica l stu d ies h ave sh o w n
b o th the prom oter and first intron and that there is a
th a t th e c a th ep sin D may be in the' strom al c o m p o ­
p ro g re ssio n from an E R positive with low levels o f
n e n t. rath er than the tum our cells and that this m ay
E G F R , con sid ered negative, to an ER negative.
E G F R overexpressing tumour. We would suggest
be a r e fle c tio n o f m acrophage in filtrate [142].
A n o t h e r o e str o g e n regulated p ro tein , h eat sh o ck
th a t there could be other reasons and that it is im ­
str e ss r e sp o n se p rotein (Srp-27) has also b een a sso ­
p o r ta n t to consider EG FR expression in the differ­
c ia te d w ith tu m ou rs being aggressive, d esp ite a cor­
e n t cell types present in normal breast e.g. could
r e la tio n w ith E R [143]. M anning et al. [144] h av e
th e r e have b een developm ent from an ER negative
id e n tifie d
E G F R p ositive m yoepithelial cell.
an
oestro g en -reg u la ted g en e. p L IV l.
w h ich is a sso c ia te d with lym ph n o d e in v o lv em e n t
a n d m av re p resen t a candidate g e n e for m etastatic
s c r e a d in E R - cancers.
G r o w t h factors
P e p tid e growth factors act in an autocrine/para­
E o i d e r m a l g r o w th factor receptor
c r in e manner. Primary breast carcinom as consist o f
E o id e r m a l urow th factor i E G F ) is n ecessary for the
an d interactions between these are im portant for
ep ith eliu m , strom a, vascular and other elem en ts
15
tu m o u r grow th . A ltered cellular ex p ression a n d /o r
r e s p o n s e to d ifferen t growth factors by e p ith e lia
a n d /o r stro m a l cells could clearly be im portan t in
th e d e v e lo p m e n t and progression o f breast can cers.
V a rio u s grow th factors have b een stu d ied e.g. in ­
su lin g ro w th factor 1 and 2. transform ing g ro w th
fa c to r alp h a , fibroblast grow th factors. O ur o w n
stu d ie s h a v e b een con cern ed w ith tran sform in g
g r o w th fa cto r b eta, and this w ill be co n sid er ed in
m o r e d eta il.
T ra n sfo rm in g gro w th fa c to r beta
T r a n sfo rm in g grow th factor b eta (TGF-(3) co m p ris­
e s a g ro u p o f m ultifun ction al regulatory p ro tein s
w h ic h h a v e effec ts on m any p ro cesses w hich co u ld
b e o f im p o r ta n ce in the overall b eh aviou r b rea st
c a n c e r [152]. T h ese include p roliferation , d iffe r e n ­
Figure 3. Breast carcinoma hybridised with digoxigenin-labelled
oligonucleotide probe to EGFR. showing strong labelling of tu­
m our cells. EG FR protein was detected immunohistochemicallv.
tia tio n , stim u la tio n o f extracellular m atrix fo rm a ­
tio n . c e ll m igration , an giogen esis and im m u n e fu n c ­
tio n .
T h e r e are so m e conflicting find in gs regarding th e
r o le ( s ) o f TGF-(3 in breast carcinom as. TGF-{3 has
b e e n r e p o r ted to inhibit [153] and stim u late [154]
p r o life r a tio n o f breast cancer cells. T am oxifen has
b e e n sh o w n to m odify the ex p ressio n o f T G F -p ,
[155] an d it has b een p rop osed that this m ay b e a
se c o n d a r y m echan ism for the m ed ia tio n o f its a n ti­
tu m o u r effec ts. In this in vivo stud y [155] stro m a l
fib r o b la sts w ere d em onstrated as th e sou rce o f e x ­
tra c ellu la r TGF-(3. which w ou ld then act in a p a ra ­
c r in e fa sh io n on tum our ceils. S u b seq u en t in vitro
stu d ie s [156] su ggest that syn th esis o f TGF-(3 by fi­
b ro b la sts is in creased by tam oxifen , but n ot n e c e s ­
sa r ily th e se cr etio n which is n e e d e d for the para­
c r in e e ffe c t. M acC allum et al. [157] e x a m in e d
tw e e n in-situ and invasive carcinomas, with fewer
in-situ cases having reactivity. We also dem onstrat­
e d th a t TGF-{3. but not TGF-(3: expression was asso­
c ia te d with lymph node metastasis [161]. It was also
associated with increased reactivity for the stromal
c o m p o n e n ts fibronectin and tenascin and altered
m a c ro ph a ge and T lymphocyte infiltration, all of
w hich may be important in enhancing invasion and
metastasis. In an in-situ hybridisation study, using
digoxigenin-labelled riboprobes. we demonstrated
TGF-(3, m R N A in the epithelial tumour cells and
n o t stromal cells [162] which is contrary to some im­
m unohistochem ical studies.
D u e to the complex, multifunctional properties
o f TGF-(3 further studies are required to elucidate
its value in clinical management.
m R N A for th e three m am m alian TGF-(3 iso fo rm s
in c a r c in o m a s b efore and after tam oxifen trea tm en t
a n d fo u n d that TG F-|3, rather than T G F -p , e x p r e s­
s io n te n d e d to increase. T he role o f T G F -P in the
r e g u la tio n o f proliferation and its relation sh ip to ta ­
m o x ife n is o b v io u sly com p lex and requ ires furth er
stud y.
T h e r e is evidence from several studies, including
o u r ow n. th a t T G F -3 expression by breast c a rc in o ­
m a s is associated with po orer prognostic fe a tu re s
[153-161]. We identified a difference in TGFf3, p r o ­
te in. as d e te c te d by im m unohistochemistry. be-
C ell ad hesion, invasion, metastasis
T his is a large topic which includes cell-cell interac­
tions. cell-stromal interactions, proteases, angioge­
nesis. growth factors, anti-metastasis genes such as
nm23. and interaction of tumour cells with specific
o rg a n environments. Since it is not possible to cover
all of these aspects comprehensively, we will con­
c e n tra te on areas of particular interest to our lab­
oratory.
16
E -c a d /ie r in
16q24.3 and identified L O H ranging from 25-44% ,
d e p e n d in g on the marker. O ne third o f the inform a­
E -c a d h e r in plays a critical role in in itiatin g an d
tiv e ca ses sh ow ed L O H for D 16S400. the closest to
m a in ta in in g cell-cell contacts and is a m em b er o f
th e region o f the E-cadherin gen e but we failed to
th e g r o w in g fam ily o f ca lciu m -d ep en d e n t cad h erin
fin d any relationship b etw een L O H and loss o f E-
a d h e s io n m o le c u le s [163]. E -ca d h erin m o le c u le s
ca d h erin staining. T h ose cases show ing L O H w ere
a re lo c a te d w ith in adherans ju n ctio n s an d are trans-
th e n screen ed for m utations in exon s 6. 7 and 9 o f
m e m b r a n e structures: the cyto p la sm ic reg io n in ter­
th e E -cad h erin gene. O ne had a 1 bp insertion in e x ­
a c ts w ith th e caten in s. w hich are in turn c o n n e c te d
o n 7 but n o other m utations w ere identified. This
to th e a c tin m icrofilam en t n etw ork [164]. E v id e n c e
w a s a w ell d ifferentiated infiltrating ductal carcino­
th a t E -c a d h e r in is im portant in su p p ressin g in va ­
m a w ith h om o g en o u s cytoplasm ic reactivity for E-
s io n c o m e s from studies in w hich this is o b serv ed in
ca d h erin . M utations have b een identified by others
c e lls tr a n sfe c te d with the c D N A [165].
in a sm all num ber o f infiltrating lobular carcinom as
S tu d ie s o f prim ary breast carcin om as have g iv e n
in c o n s is t e n t results. T he p red o m in a n t m eth o d o f
[1 7 2 ,1 7 3 ], but not in infiltrating ductal carcinom as
[174].
a n a ly sis h as b een im m u n oh istoch em istry; so m e
T h e con trol o f E-cadherin in breast carcinom as is
s t u d ie s h a v e u sed frozen m aterial an d so m e h av e
c o m p le x and alterations in the link betw een m em ­
u s e d fix e d tissu e. The in terp retation an d d efin itio n
b ra n e signalling pathways and E-cadherin expres­
o f p o s itiv e /n e g a tiv e has also varied an d so m e o f this
sio n m ay be the key to its role in invasion.
m a y a c c o u n t for the discrepancies.
R e d u c e d E -cad h erin has [166. 167] and has n o t
[168. 169] b e e n related to p o o rer d iffer en tia tio n o f
In te grin s
in filtr a tin g d uctal carcinom as. W e [169] and o th e r
[167] h a v e fo u n d a relationship b e tw e e n red u ced
In teg rin s are cell adhesion m olecules which are in­
m e m b r a n e stain in g and lym ph n o d e m etastasis. W e
v o lv e d in cell-strom al interactions, and possibly
a ls o d e m o n str a te d a highly sig n ifica n t a sso cia tio n
c e ll-c e ll interactions. They are heterodim ers com ­
b e t w e e n cy to p la sm ic reactivity for E -cad h erin in
p o s e d o f n on-covalently liked a and (3 subunits
t u m o u r c e lls and nodal m etastasis. T h is altered E -
w h ich p rovide a transm em brane link b etw een the
c a d h e r in lo c a lisa tio n m ay be the resu lt o f ab n orm a l
c y to sk e le to n and specific extracellular matrix pro­
c a te n in e x p r e ssio n , or be due to lo c a l e n v ir o n m e n ­
te in s [175]. T he integrins are classified according to
tal in flu e n c e s . W e also d em o n stra ted a co rrela tio n
th e ir (3 subunit. T h ey include receptor for collagen,
b e t w e e n th e p resen ce o f E G F R and cy to p la sm ic r e ­
a ctiv ity. T h e r e is evid en ce that p h o sp h o r y la tio n o f
Iam inin and fibronectin.
B e ca u se they play an im portant role in cell ad­
E G F R le a d s to d issociation o f E -ca d h erin /ca ten in
h e sio n and m igration, alterations in expression and/
c o m p le x e s from the cy to sk eleto n [164] and so m a y
o r fu n ction could be important in altered grow th
m o d ify fu n ctio n .
O n e o f th e interesting, and p o te n tia lly u seful d i-
stu d ie s o f integrin expression in primary breast car­
a g n o stic a lly . ob servation s that w e an d o th e r h av e
cin o m a s. using im m unohistochem istry although
a n d in vasion in neoplasm s. There have been several
m a d e is th a t infiltrating lobular carcin om as and lo b -
Z u tte r et al. [176] have exam ined m R N A exp res­
u iar c a r c in o m a in situ do not ex p ress E -cad h erin
sio n o f cq. a 5 and (3, with radioactive-labelled ribo-
'166. 1 6 8-170], T his may be o f re lev a n ce in ex p la in ­
p ro b e in-situ hybridisation. All studies have show n
ing: th e in filtra tiv e nature o f this typ e o f in vasive
red u ced expression at the protein [177-180] and
ca r c in o m a .
T h e E -ca d h erin gene m aps to ch r o m o so m e 16q:
m R N A [176] levels for cq[3:. cq(3, and cq|3.. We [180]
fo u n d no relationship to grade or node status, al­
d e le tio n s o f this region occur freq u en tly and have
th o u g h others have found loss of 0.(3. to be greater
r e e n a s s o c ia te d with distant m eta sta sis [171]. W e
in p oorly differentiated carcinomas [177. 178]. A s
n a v e u sed m icrosatellite m arkers co v e rin g I6q21 to
w ith E-cadherin. we found that reactivity could be
17
c y to p la sm ic rather than m em b ran e [180] an d th is
lo ss o f p o la riza tio n has b een re p o r ted by P ig n a te lli
p o n e n ts and their in ter-relationships. U sin g im ­
m u n o h isto ch em istry uPA can be d etec te d in tu ­
e t al. [181]. O verall alterations in in tegrin e x p r e s ­
m o u r cells, with a greater in ten sity than n orm al/
sio n h a v e n o t b een o f value in p rovid in g p r o g n o stic
in fo r m a tio n .
b en ig n ep ith eliu m and also w eak ly in strom al cells
W e o b s e r v e d an alteration in the sta in in g fo r c^P..
in
n o n -in v o lv e d
tissue
from
[187]. A sim ilar pattern o f stain ing is seen for PA I-1
[187]. R esu lts for the loca liza tio n o f uP A R have
c a n c e r -c o n ta in in g
b e e n con flictin g. Pvke et al. [188] and B ianchi et al.
b r e a sts, in th at tw o thirds o f the ca ses s h o w e d lo s s o f
[189] did not d etect uPA R in n orm al and b en ign
rea ctiv ity , id en tica l to that o f the c o r r e sp o n d in g tu ­
b rea st, but the form er only foun d the receptor o n
m o u r [180]. T h is was not se e n for o th e r in te g rin s
tu m o u r associated m acroph ages, w h ile the latter
a n d su g g e sts that altered a 6P4 in tegrin sta in in g m a y
fo u n d it on tum our cells as w ell in a p rop ortion o f
b e an ea r ly e v e n t in the n eo p la stic p r o c e ss, a n d as
ca ses. In contrast C onstantini et al. [187] d etec te d
su ch , m ay b e o f use as a m arker o f p re-m a lig n a n t
u P A R on norm al and tum oural ep ith elia l cells and
ch a n g e.
stro m a l cells, but to a greater ex te n t in tum ours.
O n e rea so n for som e of the d ifferen ces could b e
d u e to d ifferen ces in the sen sitiv ities o f the a n ti­
P r o te a s e s
b o d ie s used, since the g ly co sy la ted variants o f
u P A R m ay have b een d etec te d to d ifferen t d e ­
P r o te o ly tic en zym es that cou ld d eg ra d e e x tr a c e llu ­
la r m a trix m ay play an im portan t r o le in in v a s io n
and
m eta sta sis. P roteases fall in to
th ree m a in
g rees.
T h e finding o f reactivity in tum our and strom al
ce lls suggests the existance o f co m p lex paracrine in­
g ro u p s: asp artyl. which in clu d es ca th e p sin D . w h ic h
teraction s. H ow ever, staining m ay reflect bound e n ­
h a s a lr ea d y b e e n discussed: serin e, su ch as p la s m i­
zy m e and d oes not necessarily indicate the site o f
n o g e n activators: and m etal d e p e n d e n t, w h ich in ­
syn th esis. A lso , it is not know n w heth er the exp res­
sio n ob served in carcinom as is a primary alteration
c lu d e s m eta llo p ro tein a se s and stro m elv sin s.
or is sim ply a m arker of the effects o f cytokines and
P la s m in o g e n activators
grow th factors on tumour cells. A m ore d etailed un ­
T h e r e are tw o m ain form s o f p la sm in o g e n a c tiv a ­
d erstan ding o f this w ould be o f value so as to id en ti­
to rs, u P A a n d tPA . w hich activate p lasm in : u P A a lso
fy the m ost appropriate targets, for the design o f
p a r tic ip a te s in tissue rem od ellin g u n d er n orm al a n d
n e w therap eu tic agents.
p a th o lo g ic a l con d ition s [182]. T h e u P A p a th w a y is
r e g u la te d b y tw o inhibitors PA I-1 an d P A I-2 an d b y
a c e ll su rfa ce receptor for uPA ( u P A R ) w h ich b in d s
M a trix m etalloproteinases
T h e m atrix m etalloproteinases (M M P) are a fam ily
b o th p r o -u P A and active uPA . T h u s th e ce ll su r fa c e
o f zinc d ep en dent enzym es capable o f degrading
m a y be th e p h ysiological site o f u P A -in itia te d fib r i­
d ifferen t com ponents of the extracellular m atrix
n o ly sis.
T h e r e has b een interest as to w h e th e r th e p la sm i­
w hich can be broadly categorized into collagen ases.
n o g e n a ctiv a to rs could provide p ro g n o stic in fo r m a ­
gela tin a ses and strom elvsins acording to their s p e c ­
[190]. A t least 11 m em bers are now described [191]
tio n an d th ere have been several rep orts that c a r c i­
ificity. The collagenases include MMP-1 ( interstitial
n o m a s w ith high levels o f uPA an d PA I-1 h a v e a
co lla g en a se). MMP-3 (neutrophil collagen ase) and
high risk o f relapse [183-186]. T h e se stu d ie s h a v e
M M P -13 (collagenase 3). and degrade fibrillar in­
u sed E L I S A assays o f tissue h o m o g e n a te s. w h ich
terstitial collage. The gelatinases. also know n as
has the ad van tage of q u an tification but d o e s r e ­
q u ire fresh tissue. C ertainly uPA an d PAI-1 a p p e a r
against denatured collagen and include M M P-2
to be u sefu l prognostic m arkers.
T h e s e ap p roach es though do n ot p rovid e in fo r ­
tinase B. 92 kd gelatinase). Strom elvsins 1. 2 and 3
m a tio n a b o u t cellular location o f the d ifferen t c o m ­
i M M P-3. M M P-10 and M M P -ll respectively) have
type
IV
collagenases.
are
particularly
p o ten t
(gela tin a se A . 72 kd gelatinase) and M M P-9 (g ela -
IS
a b ro a d er su b stra te specificity and d egrad e g ly c o ­
M M P and M T3-M MP. have in com m on a trans­
p r o te in s su ch as lam inin and fibronectin .
m em b ra n e d om ain and are localised at the cell
M M P a c tiv ity can be con trolled at m u ltip le le v e ls
m em b ra n e. B o th MT1-MMP and M T2-M M P are
w h ic h in clu d e transcriptional activation , c o n v e r ­
p o te n t activators o f progelatinase A [207.210] and
s io n o f la ten t precursor to active en zy m e and in h ib ­
so ex p ressio n o f these receptors on tum our cells
it io n via in tera ctio n with the sp ecific tissu e in h ib ­
w o u ld lo ca lise proteolytic activity to the site o f in ­
ito r s o f the M M P s. TIM P-1. 2 and 3 [192.193].
v a sio n . m irroring the strategy displayed by the uPA
B o th im m u n oh istoch em istry and zy m o g r a p h y
h a v e d e m o n str a te d increased reactivity for M M P -2
r e c e p to r [187].' E levated levels o f M T1-M M P has
in b rea st ca rcin o m a s in com p arison to n orm al a n d
b e n ig n b rea st (194.195]. w ith som e stu d ies sh o w in g
212]. W ork in our laboratory has revealed con sis­
a r e la tio n sh ip to lym ph n od e in v o lv em e n t [196].
n o m a s and suggests that com bined expression o f
b e e n d em on strated in various tum our tissues [211,
te n tly high reactivity for M T1-M M P in breast carci­
O u r o w n stu d ie s have confirm ed the rea ctivity b u t
M T 1-M M P and M M P-2 show a better correlation
n o t fo u n d an y relation ship w ith n o d e statu s (m a n u ­
w ith p rogn osis than MMP-2 alone (m anuscript in
sc r ip t in p rep a ra tio n ). M M P-9 im m u n o rea c tiv itv
p rep a ra tio n ).
h a s a lso b e e n sh ow n to correlate w ith n o d a l m e ta s­
T h er e has b een considerable interest in the d e ­
ta sis [197]. M M P-11 or strom elvsin 3 [198] has b e e n
v e lo p m e n t o f agents which could m odify M M P
d e t e c t e d in m o st invasive and high grade in situ car­
fu n c tio n , that w ould have therapeutic applications
c in o m a s b u t is rarely d etected in b en ign b reast d is­
fo r m an y d iseases, including neoplasia. A pplication
e a s e [1 9 9 ,2 0 0 ]. L evels of strom elvsin 3 m R N A m a y
o f th e se in breast cancer w ould clearly be of inter­
b e o f p r o g n o stic value in that o n e stu d y fo u n d r e ­
est.
c u r r e n c e m o re com m on in p atien ts w ith tu m ou rs
h a v in g high as com pared to low le v els [201].
B e c a u s e o f th e interactions b e tw e e n M M P s. th e ir
C o n c lu sio n s
a c tiv a to r s an d inhibitors, assessm en t o f o n e a lo n e
o r e v e n a p a n e l m ay be of lim ited v alu e w ith regard s
It is n o t p ossib le to cover every aspect of the m o le­
to p r o g n o sis. T h is is illustrated by a stu d y u sin g
cu la r p a th o lo g y o f breast cancer which could be o f
q u a n tita tiv e
an d
clin ica l im portance e.g. multi-drug resistance genes.
T I M P -2 m R N A in n ode n egative and p o sitiv e c a se s
H o w e v e r , m any aspects have b een covered which
[2 0 2 ]. W h en le v e ls o f m R N A w ere a n a ly sed s e p a ­
sh o w that our greater understanding o f m olecular
R T -P C R to com p are M M P -2
r a te ly n o re la tio n sh ip was foun d, but w h en the ra ­
m ech a n ism s and the increasing ability to in vesti­
tio s o f M M P -2 m R N A to T IM P -2 m R N A w ere e x ­
g a te th ese in primary breast carcinom as should aid
a m in e d . ra tio s o f 2-4:1 w ere a sso cia ted w ith n o d e
th e clinical m anagem ent of breast cancer. We high­
m e ta sta s is an d < 1:1 with n egative n o d es. A n a ly sis
lig h te d three im portant areas in the developm en t
b y N o r th e r n b lo ttin g did not find any re la tio n sh ip
a n d p rogression o f breast cancer and have id en ti­
b e t w e e n M M P -2 m R N A levels and m align an cy,
fie d alterations which could be o f relevance to three
o n ly h ig h er’le v e ls o f MMP-11 m R N A in ca r cin o m a s
a sp e c ts o f clinical management:
12 0 3 ].
W h ilst im m u n oh istoch em ical stu d ies sh o w sta in ­
E a r ly changes/screening potential
in g o f M M P s in tum our and strom al cells, in -situ h y ­
L o ss o f key chrom osm al regions e.g. 6q.
M icro sa tellite instability as a marker o f D N A repair
b r id iz a tio n stu d ie s have localised M M P tran scripts
p r e d o m in a n tly to the strom al cells [204. 205]. T h is
B R C A 1 and B R C A 2.
d efec ts.
a o D a ren t d iscrep an cy can be ex p lain ed by the p res-
P53 m utations in certain families.
er.ce*of re ce p to rs for M M Ps on tum our cells [206].
C ertain E R variants and polym orphism s.
A ls o a n ew form of MM P has recen tly b een d e ­
A lte r e d a6(34 integrin expression in non involved
sc r ib e d . the m em brane type M M P (M t-M M P )
r2 0 ~ -2 0 9 ]. T h e three m olecules. M T1-M M P. M T 2-
ca n cero u s breast.
19
B e h a v io a r/p rogn os is
C -e rb 2 o v er -ex p ressio n and p rogn osis.
2.
P 53 m u ta tio n s and p rognosis.
E G F R a n d p rognosis.
C a th e p sin D and prognosis.
3.
T G F -(3 l a n d lym ph node m etastasis.
E -C a d h e r in and lym ph n o d e m etastasis.
4.
u P A an d P A I-1 and p rogn osis.
M M P -2 a n d M T1-M M P and p rogn osis.
5.
T re a tm e n t stra teg y /n e w th erapies
C -e r b B -2 a n d p oor resp on se to en d o crin e therap y.
P o te n tia l ta rg et site for n ew drug.
P 5 3 m u ta tio n and poor resp o n se to en d o crin e th er­
6.
ap y; d e v e lo p m e n t o f th era p eu tic agen ts a g a in st
m u ta n ts.
E R an d re sp o n se to en d ocrin e therapy, p o ssib ility
o f v a ria n ts exp lain in g ta m o x ifen resistan ce.
7.
P g R . P S 2 a n d C yclin D l re sp o n se to en d o crin e therapv.
E G F R a n d p o o r resp onse to e n d o crin e therapy.
8.
P la s m in o g e n activators and in h ib itors (u P A a n d
P A I-1 ) an d m eta llo p ro tein a se s as targets for d e sig n
o f n e w th e ra p eu tic agents.
9.
T h e o n ly w a y forw ard is to en su re that b oth b io lo g ­
ic a l an d clin ica l features o f b reast carcin om a are
c o n s id e r e d togeth er since w e n e e d to h ave an u n ­
10.
d e r s ta n d in g o f the form er if w e are to im p rove th e
c lin ic a l o u tc o m e o f b reast can cer. T h ere is still
11.
m u c h w o r k to be done.
A c k n o w le d g e m e n ts
W e are g ra te fu l to Mrs S h eila D ea r in g for h er c o n ­
12.
tin u in g e x c e lle n t technical su p p ort, the se cr eta rie s
o f th e D e p a r tm e n t of P ath ology, p articu larly M rs
B e v e r le y R ich ard son , for secretarial su p p ort, a n d
C ancer
R esea rch
C am paign.
Trent
13.
R e g io n a l
H e a lth . R o y a l Society and the U n iv ersity o f L e ic ­
14.
e s te r for finan cial support.
15.
References
I.
Evan G I. Wyllie AH. G ilbert CS. L ittlew ood TD. L and H .
B rooks M. Waters CM. Penn LZ. H ancock DC: Induction
16.
of apoptosis in fibroblasts by c-m yc protein. Cell 69:1-20.
1992
E dw ards PAW. Ward JL. Bradbury JM: Alteration of mor­
phogenesis by the v-mvc oncogene in transplants of mouse
mammary' gland. Oncogene 2: 407-412.1988
Bonilla M. Ramirez M. Lopez-Cueto J. Gariglio P: In vivo
am plification and rearrangem ents of c-m yc oncogene in
hum an breast tumours. J Natl Cancer Inst 80:665-671.1988
E scot C. Theillet C. Lidereau R. Spvratos F. Champeme
M-H. G est J. Callaham R: Genetic alterations of the c-m yc
proto-oncogene (M Y O in human breast carcinomas. Proc
N atl A cad Sci L'SA 83: 4834-4838.1986
Varlev JM . Swallow JE. Brammar WJ. W hittaker JL.
W alker RA: Alterations to either c-erbB2 (neu) or c-m yc
proto-oncogenes in breast carcinomas correlate with poor
short-term prognosis. Oncogene 1: 423-430.1987
B em s PM JJ. Krijn JGM. Van Putten WLJ. Van Steveren
IL. Portengen H. Foekens JA: C -m yc amplification is a bet­
ter prognostic factor than HER2/new amplification in pri­
m ary breast cancer. Cancer Res 52:1107-1113.1992
W alker RA. Senior P V Jones JL. Critchlev DR. Varlev JM:
A n imm uno-histochemical and in-situ hybridisation study
of c-m yc and c-erbB-2 expression in primary breast tu­
m ours. J Pathol 153: 97-105.1989
L ocker AP. Dowie CS. Ellis IO. Elston CW. Blarney RW.
K ikora K. Evan G. Robins RA: C-m yc oncogene product
expression and prognosis in operable breast cancer. Br J
C ancer 60: 669—6”2 . 19S9
R ochlitz CF. Scott GK. Dodson JM. Liu E. Dolbaum C.
Sm ith HS. Benz CC: Incidence of activating ras oncogene
m utations associated with primary and metastatic breast
cancer. Cancer Res 49: 357-360.1989
T heillet C. Liderau R. Escot C. Hutzell P. Brunet M. Gest J.
Schlom J. Callahan R: Loss of c-H-ras-l allele and aggres­
sive hum an primary breast carcinomas. Cancer Res 46:
4776-4781.1986
O huchi N. T hor A. Page DL. Horan Hand P. H alter S.
Schlom J: Expression of the 21.000 molecular weight ras
protein with a molecular weight ras protein in a spectrum
of benign and malignant human mammary tissues. Cancer
Res 46: 2511-2519.1986
C lair T. Miller WR. Cho-Chung YS: Prognostic signifi­
cance of the expression of a ras protein with a m olecular
weight of 21.000 by human breast cancer. Cancer Res 47:
5290-5293.1986
W alker RA. Wilkinson N: p21 ras protein expression in be­
nign and malignant human breast. J Pathol 156: 147-153.
1988
Going JJ. Anderson TJ. Wyllie AH: Ras p21 in breast tis­
sue: associations with pathology and cellular localisation.
Br J Cancer 65: -5-50. 492
K rontinsT G . De\iin B. Karp DD. Robert JJ. Risch N: An
association between the risk of cancer and mutations in the
H rasl mini-satellite locus. N Engl J Med 329:517-523.1993
Coussens L. Yana-Feng TL. Chen Y-CLE. Gray A.
M cG rath J, Scebura PH. Libermann TA. Schiessinger J.
20
Francke L'. Levinson A. Ullrich A: Tyrosine kinase recep­
tor with extensive homology to EG F receptor shares chro­
mosomal location with
oncogene. Science 230: 1132—
1139.19S5
Slam on DJ. Clark G. Wong SG. Levin WJ. Ullrich A.
M cG uire \VL: Human breast cancer: correlation of relapse
and survival with amplification of the HER-2//ieu onco­
gene. Science 235: 177-182.1987
Z hou D. Battifora H. Yokata J. Yamamoto T. Cline MJ:
A ssociation of multiple copies of the
oncogene
with the spread of breast cancer. Cancer Res 47:6123-6125.
1987
B orresen A-L. Ottestad L. Gaustad A. A ndersenT I. Heikkila R. Jahnsen T. Tveit KM. Nesland JM: Amplification
and protein over-expression of the
proto-oncogene in human breast carcinomas: relationship
to loss of gene sequences on chromosome 17. family history
and prognosis. Br J Cancer 62: 585-590.1990
C lark GM. McGuire WL: Follow-up study of
amplification in primary breast cancer. Cancer Res 51:944—
948.1991
Z hou D-J. Ahuja H. Cline MJ: Proto-oncogene abnorm al­
ities in hum an breast cancer:
amplification does
not correlate with recurrence of disease. Oncogene 4:105108.1989
Liu E. T hor A. He M. Barcos M. Ljung B-M. Benz C: The
H ER-2(c-erbB-2) oncogene is frequently amplified in
carcinomas of the breast. Oncogene 7:1027-1032.1992
H ubbard AL. Doris CP. Thompson AM. Chettv U. Ander­
son TJ: Critical determination of the frequency of c-erbB-2
amplification in breast cancer. Br J Cancer 70:434—139.1994
V enter DJ. Tuzi NL. Kumar S. Gullick WJ: Over-expres­
sion of the c-erbB-2 oncoprotein in human breast carcino­
mas: immunohistological assessment correlates with gene
amplification. Lancet ii: 69—“2.1987
O ’Reilly SM. Barnes DM. Camplejohn RA. Bartkova J.
G regory WM. Richards MA: The relationship between cexpression. S-phase fraction and prognosis in
breast cancer. Br J Cancer o3: 444—146.1991
W alker RA. Gullick WJ. Varlev JM: An evaluation of imm unoreactivitv for c-erbB-2 protein as a m arker of short­
term prognosis in breast cancer. Br J Cancer 60: 426-429.
neu
17.
18.
c-erbB-2
19.
neuJHER-2Jc-erbB-2
20.
21.
HER-2Jneu
c-erbB-2
22.
in-
situ
23.
24.
25.
erbB-2
26.
1989
27. Gullick WJ. Love SB. Wright C. Barnes DM. Gusterson B.
H arris AL. Altman DG:
protein overexpression
in breast cancer is a risk factor in patients with involved and
u n involved lymph nodes. Br J Cancer 63: 434—138. 1991
C-erbB-2
28. L ovekin C. Ellis 10. Locker A. R obertson JF R . Bell J. Ni. eholson R. Gullick WJ. Eiston CW. Blarney RW: C-erbB-2
o n c o p ro te in expression and prognosis in operable breast
cancer. Br J C ancer t>3: 43°——3. 1991
29. W instanlev J. Cooke T. M urray GD. Platt-H iggins A. G eorae WD. H olt S. Mystov M. Scedding A. B arraclough BR.
R u d lan d PS: The long-term prognostic significance of cin prim ary breast cancer. Br J C ancer 63: 447-450.
1991
erhB-2
30. Press MF. Piek MC. Chazin VR. Hung G. Udove JJA. Markowicz M. Danyluk J. Godolphin W. Sliwkowski M. Akita
R. Paterson MC, Slamon DJ: Her-2
expression in node
negative breast cancer: direct tissue quantitation by com­
puterized image analysis and association of over-expres­
sion with increased risk of recurrent disease. Cancer Res
53: 4960-4970. 1993
31. Press MF. Hung G. Godolphin W, Slamon DJ: Sensitivity
of
antibodies in archival tissue samples: poten­
tial source of error in immuno-histochemical studies of on­
cogene expression. Cancer Res 54: 2771-2777,1994
32. Van der Vijver MJ. Peterse JL, Mooi WJ. Wisman P. Lomans J. Dalesio O. Nuse R: .Veu-protein over-expression in
breast cancer: association with comedo-type ductal carci­
noma
and limited prognostic value in Stage II breast
cancer. N Engl J Med 319:1239-1245.1988
33. Bartkova J, Barnes DM. Millis RR. Gullick WJ: Immuno­
histochemical demonstration of c-erbB-2 protein in mam­
mary ductal carcinoma
Hum Pathol 21:1164-1167,
1990
34. Ramachandra S. Machin L. Ashley S. Monaghan P. Gus­
terson BA: Immunohistochemical distribution of c-erbB-2
in in situ breast carcinoma - a detailed morphological anal­
ysis. J Pathol 161: 7-14.1990
35. Lodata RJF, Maguire HC Jr. Greene MI. Weiner DB. LiVolsi VA: Immunohistochemical evaluation of c-erbB-2
oncogene expression in ductal carcinoma
and atyp­
ical ductal hyperplasia of the breast. Mod Pathol 3: 449.
1990
36. Allred DC. Clark GMR. Tandon AK. Schmitt SJ. Gil­
christ KW, Osborne CK. Tormev DC. McGuire WL:
Overexpression of
and its relationship with
other prognostic factors change during the progression of
to invasive breast cancer. Hum Pathol 23:474-479,
1992
37. Wright C. Nicholson S. Angus B. Sainsbury JRC. Famdon
J. Cairns J. Harris AL. Home CHW: Relationship between
protein product expression and response to en­
docrine therapy in advanced breast cancer. Br J Cancer 65:
118-121.1992
38. Nicholson RI. McClelland RA. Finlay P. Eaton CL. Gul­
lick WJ. Dixon AR. Robertson JFR. Ellis IO. Blarney RW:
Relationship between EGF-R. c-erbB-2 protein expres­
sion and Ki67 immunostaining in breast cancer and hor­
mone sensitivity. Eur J Cancer 29A: 1018-1023.1993
39. Klijn JGM. Berns EMMJ. Foekens JA: Prognostic factors
and response to therapy in breast cancer. Cancer Surv 18:
165-198.1993
40. Fantl V. Richards MA. Smith R. Lammie GJA. Johnstone
G. Allen D. Gregory W. Peters G. Dickson C. Barnes DM:
Gene amplification on chromosome band 1Iq 13 and oes­
trogen receptor status :n breast cancer. Eur J Cancer 26:
423-429.1990
41. Borg A. Sigurdsson H. Clark G. Femo M. Fuqua SAW, Olsson H. Killander D. McGuire WL: Association of INT2/
HST1 coamplification :n primary breast cancer with hor­
/neu
HER-2/neu
insitu
insitu.
insitu
HER-1'neu
insitu
c-erbB-2
21
42.
m one-dependent phenotype and poor prognosis. Br J C an­
cer 63: 136-142.1991
H enry JA. Hennessy C. Levett DL. Lennard TWJ. Westlev
BR. May FEB:
amplification in breast cancer: associ­
ation with decreased survival and relationship to amplifi­
cation of c-er&B-2 and
Int J Cancer 53: 774-780.
1993
Xiong Y. Menninger J. Beach D. Ward DC: M olecular
cloning and chromosomal mapping of CCND genes encod­
ing hum an D-tvpe cvclins. Genomics 13: 575-584.1992
Scherr CJ: G1 phase progression: cycling on cue. Cell 79:
551-555.1994
Buckley MF. Sweeney KJE. Hamilton JA. Sini RL, M an­
ning D L. Nicholson Rl. de Fazio A. Watts CKW. Musgrove
E A. Sutherland RL: Expression and amplification of cyclin
genes in human breast cancer. Oncogene 8:2127-2133.1993
B artkova J. Lukas J. Muller H. Lutzhoft D. Strauss M. Bartek J: Cyclin D l protein expression and function in hum an
breast cancer. Int J Cancer 57: 353-361.1994
Gillet C. Fantl V. Smith R. Fisher C. Bartek J. Dickson C.
Barnes DM, Peters G: Amplification and overexpression
of cyclin D l in breast cancer detected by immunohistochemicai staining. Cancer Res 54:1812-1817.1994
M cIntosh GG. Anderson JJ. Milton I. Steward M. P art
A H . Thomas MD. Henry JA. Angus B. Lennard TW J.
H orne CHW: Determination of the prognostic value of cy­
clin D l overexpression in breast cancer. Oncogene 11:885891.1995
Gillet C. Smith P. Gregory W. Richard M. Millis R. Peters
G. Barnes D: Cvciin Dl and prognosis in human breast can ­
cer. Int J Cancer 69: 92-99.1996
Friend SH. Bernards R. Rogelj S. Weinberg RA. R apaport
IM. A lbert DM. Dryja TP: A human D NA segment with
properties of the gene that predisposes to retinoblastom a
and osteosarcoma. Nature 323: 643-646.1986
Chellappan SP. Hierbert S. Mudrvj M. Horowitz JM. Nevins JR: The E2F transcription factor is a cellular target for
the RB protein. Ceil 65:1053-1061.1991
Lundberg C. Skoog L. Cavanee WK. Nordenskjold M:
Loss of heterozygosity in human ductal breast tumors in­
dicates a recessive mutation on chromosome 13. Proc N atl
A cad Set USA 84: 2372-2376. 1987
T 'A ng A. Varlev JM. Chakraborty S. M urphree AL. Fung
Y-KT: Structural rearrangement of the retinoblastom a
gene in human breast carcinoma. Science 242: 263-266.
Int-2
c-myc.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
1988
54. V arlev JM . A rm our J. Swallow JE . Jeffreys A J. P o n d e r
BA J. Tang A. Fung Y-KT. B ram m ar W J. W alker RA: T h e
retin o b lasto m a gene ;s frequently altered leading to loss o f
expression in pnm ary breast tum ours. O ncogene 4: 7 2 5 -
68.
730.1989
55. B ora A. Z hana Q-X. Aim P. O lsson H. Sellberg G:
' retinoblastom a gene :n breast cancer: allele loss is not
related with loss of gene protein expression. C ancer
52: 2991-2994.19°2
50. M alkin D. Li FP. Strong LC. Fraum eni JF Jr. Nelson
69.
The
c o r­
Res
CE.
"0.
Kim DH. Kassel J. Gryka MA. Bischoff FZ. Tainsky MA.
Friend SH: Germ line p53 mutations in a familial syndrome
of breast cancer, sarcomas and other neoplasms. Science
250:1233-1238,1990
Santibanez-Koref MF. Birch JM. Hartley AL. Norris Jones
PH, Craft AH. Eden T. Crowther D. Kelsey AM. Harris M:
p53 germline mutations in Li Fraumeni syndrome. Lancet
338:1490-1491.1991
Sidransky D. Tokins T. Helzisouer K. Rauch G, Zehnbauer
B. Shelton B. Prestiglacomo L. Vogelstein B. Davidson N:
Inherited p53 gene mutations in breast cancer. Cancer Res
52: 2984-2986.1992
Warren W. Eeles RA. Ponder BAJ. Easton DF. Averill D.
Ponder MA. Anderson K. Evans AM. De Mars R. Love R:
No evidence for germline mutations in exons 5-9 of the p53
gene in 25 breast cancer families. Oncogene 7:1043-1046,
1992
Varlev JM. Brammar WJ. Lane DP. Swallow JE, Dolan C,
Walker RA: Loss of chromosome 17pl3 sequences and mu­
tation of p53 in human breast carcinomas. Oncogene 6:
413-421.1991
Osborne RJ. Merlo GR. Mitsudomi T. Venesio T. Liscia
DS. Cappa APM. Chiba I. Takahashi T. Nau MM. Callahan
R. Minna JD: Mutations in the p53 gene in primary human
breast cancers. Cancer Res 51: 6194-6198.1991
Mazars R. Spinardi L. Ben Cheikh M. Simony-Lafontaine
J. Jeanteur P. Theillet C: p53 mutations occur in aggressive
breast cancer. Cancer Res 52: 3918-39213.1992
Coles C. Condie A. Chettv U. Steel CM. Evans HJ. Prosser
J: p53 mutations in breast cancer. Cancer Res 52: 52915298.1992
Merlo GR. Bemardi A. Diella F. Venesio T. Cappa APM.
Callahan R. Liscia DS: In primary human breast carcino­
mas mutations in exons 5 and 6 on the p53 gene are associ­
ated with high S-pnase index. Int J Cancer 53:531-535,1993
Andersen TI. Holm R. Nesiand JM. Heimdel KR. Ottestad
L. Borresen AL: Prognostic significance of TP53 alter­
ations in breast carcinoma. Br J Cancer 68: 540-548.1993
Caleffi M. Teague MW. Jensen RA. Vzencak-Jones CL.
Dupont WD. Pari FF: p53 gene mutations and steroid re­
ceptor status in breast cancer. Cancer 73:1247-1256,1994
Evfjord JE. Thorlacius S. Steinarsdottir M. Valgardsdottir
R. Ogmundsdottir HM. Anamthawat-Jonsson K: p53 ab­
normalities and genomic instability in primary human
breast carcinomas. Cancer Res 55: 646-651.1995
Bergh J. Norbert T. Sjostren S. Lindgren A. Holmberg L:
Complete sequencing of the p53 gene provides prognostic
information in breast cancer patients, particularly in rela­
tion to adjuvant systemic therapy and radiotherapy. Na­
ture Med 1: 1029-1034 1995
Hall PA. McKee PH. Menage H. Du P. Dover R. Lane DP:
High levels of p53 protein in LA‘-irradiated normal human
skin. Oncogene 8: 203-207. 1993
Jacquemier J. Moles JP. Penault-Liorca F. Adelaide J. Torrente M. Viens P. Birnbaum D. Theillet C: p53 immunohis­
tochemical analvsis in breast cancer with tour monoclonal
22
71.
72.
73.
74.
75.
“6.
“7.
~8.
79.
SO.
antibodies: comparison of staining and PCR: SSCP results.
Br J C ancer 69: 846-852.1994
Fisher CJ. Gillett CE. Vojtesek B. Barnes DM. Millis R:
Problem s with p53 immunohistochemical staining: the ef­
fect of fixation and variation in the methods of evaluation.
Br J C ancer 69: 26-31.1994
T hor AD. Moore IDH. Edgerton SM. Kawasski ES. Reihsaus E. Lynch HT. Marcus JN. Schwartz L. Chen L-C. Mavall BH. Smith HS: Accumulation of p53 tum or suppressor
gene protein: an independent marker of prognosis in
breast cancers. J Natl Cancer Inst 84: 845-855. 1992
W alker R A . Dearing SJ. Lane DP. Varlev JM: Expression
of p53 protein in infiltrating and in-situ breast carcinomas.
J Pathol 165: 203-211.1991
Poller DN. Hutchings CE. Galea M. Bell JA. Nicholson
R A . Elston CW. Blarney RW. Ellis IO: p53 protein expres­
sion in breast cancer carcinoma: relationship to expression
of epiderm al growth factor receptor, c-erbB-2 protein
overexpression, and oestrogen receptor. Br J Cancer 66:
583-588.1992
Barnes DM. Dublin EH. Fisher CJ. Levison DA. Millis
RR: Immunohistochemical detection of p53 protein in
m am m ary carcinoma: an important new independent in­
dicator of prognosis? Human Pathol 24: 469-476.1993
Lipponen P. Ji H. Aaltomaa S. Svrjanen S. Svrjanen K: P53
protein expression in breast cancer as related to histopathological characteristics and prognosis. Int J Cancer 55: 5156.1993
Sjogren S. Inganas M. Norberg T. Lindgren A. Nordgren
H. Holm bergh L. Bergh J: The p53 gene in breast cancer:
prognostic value of complementary DNA sequencing ver­
sus immunohistochemistry'. J Natl Cancer Inst 88:173-182.
1996
Lowe SW. Rulev HE. Jacks T. Housman DE: p53 depend­
ent apoptosis modulates the cytotoxicity of anti cancer
agents. Cell 84: 957-967. 1993
Devilee P. Cornelisse CJ: Somatic genetic changes in hu­
man breast cancer. Biochem Biophvs Acta 1198: 113-130.
1994
Stratton MR. Collins N. Lakhani SR. Sloane JP: Loss of
heterozygosity in ductal carcinoma in
of the breast. J
Pathol 175:195-202. 1995
T hibodeau SN. Bren G. Scnaid D: Microsatellite instability
in cancer of the proximal colon. Science 269: 816-819.1993
W ooster R. Cleton-Jansen A-M. Collins N. Mangion J.
Cornelis RS. Cooper CS. Gusterson BA. Ponder BAJ. von
Deimhing A. Wiestler OD. Cornelisse CJ. Devilee P. Strat­
ton MR: Instability of short tandem repeats (microsatel­
lites) in human cancers. Nature Genetics 6: 152-156. 1994
Yee CJ. Roodi N. Verner CS. Pari FP: Microsatellite instabilitv and loss of heterozygosity in breast cancer. Cancer
Res 54: 1641-1644. 1944
-Patel L\ G rundfest-B rom atow ski S. G u p ta M. B anerjee S:
S6.
87.
88.
89.
90.
91.
situ
31.
32.
S3.
34.
M icrosatellite instabilities in 5 chrom osom es in prim ary
breast tum ours. Oncogene 9: 3695-3700. 1994
35. Shaw JA. Walsh T. Chappell SA. Carey N. Johnson K.
92.
Walker RA: Microsatellite instability in early breast can­
cer. Br J Cancer 73: 1393-1397.1996
Miki Y. Swensen J. Shattuck-Eidens JD. Futrea! PA.
Harshman K.Tavtigian S. Liu Q. Cochran C. Bennett LM.
Ding W. Bell R. Rosenthall J. Hussey C. Tran T. McClure
M. Frye C. HattierT. Phelps R. Hangen-Strauo A. Kalchen
H. Yakumo K. Ghoiami Z. Shaffer D. Stone S. Bayer S.
Wray C. Bogden R. Davananth P. Ward J. Tonin P. Narod S.
Bristow PK. Norris FH. Helvering L. Morrison P. Rostek P.
Lai M. Barrett JC. Lewis C. Neuhausen S. Cannon-Albright L. Goldgar D. Wiseman R. Kamb A. Skolnick MH:
A strong candidate for the breast and ovarian cancer sus­
ceptibility gene BRCA1. Science 266: 66-71,1996
W ooster R. Bignell G. Lancaster J. Swift S. Seal S. Mangion
J. Collinss N. Gregory S. Gumbs C. Micklem G. Barfoot R.
Hamoudi R, Pantei S. Rice C, Biggs P. Hashim Y. Smith A.
Connor F. Arason A. Gudmundsson J. Ficenec D, Kelsell
D. Ford D. Tonin P. Bishop DT. Spurr NK. Ponder BAJ.
Eeles R. Peto J. Devilee P. Cornelisse C. Lynch H. Narod S.
Lenoir G. Egilsson V. Barkadottir RB. Lynch H. Easton
DF, Bentley DR. Futreal PA. Ashworth A. Stratton MR:
Identification of the breast cancer gene BRCA2. Nature
378: 789-792.1995
Hogervorst FBL. Cornelis RS. Bout M. Van Vliet M. Oosterwijk JC. Olmer R. Bakker B. Klijn TGM. Vasen HFA.
Meijers-Heijbuer H. Menko FH. Cornelisse CJ. Den Dunnen JT. Devilee P. Van Ommen G-JB: Rapid detection of
BRCA1 mutations by the protein truncation test. Nature
Gen 10: 208-212.1995
Beckmann MW. Picard F. Anlt X. Van Roeyen CRC. Dominik SI. Mosnv DS. Schnurch HG. Bender HG. Niederacher D: Clinical impact of detection of loss of heterozy­
gosity of BRCA1 and BRCA2 markers in sporadic breast
cancer. Br J Cancer "3:1200-1226.1996
Hosking L. Trowsdale J. Nicolai H. Solomon E. Foulkes W.
Stamp G. Signer E. Jeffreys A: A somatic BRCA1 muta­
tion in an ovarian tumour. Nature Gene: 9: 343-344.1995
Rose C. Thorpe SM. Andersen KW on behalf of Danish
Breast Cancer Co-operative Group: Beneficial effect of
adjuvant tamoxifen therapy in primary breast cancer pa­
tients with high oestrogen receptor values. Lancet i: 16-20.
1985
King WJ. Al Sombre ER. Jensen E V. Greene GL: C om par­
ison of immunocytcchemical and steroid-binding assays
for oestrogen receptor in human breast tumors. Cancer
Res 45: 293-304. 1985
93. A ndersen J. Orntoft TF. Poulsen HS: Immunohistochemical dem onstration of estrogen receptor i ER) in formalinfixed. paraffin embedded human breast cancer tissue by
use of monoclonal antibody to ER. J Histochem Cvtochem
36:1553-1560.198S
94. Saccani Jotti G. Johnstone SRD. Salter J. Detre S. Dowsett
M: Com parison of new immunohistochemical assay for
oestrogen receptor :n paraffin wax embedded breast carci­
nom a tissue with quantitative enzyme immunoassay. J Clin
Pathol 47: 900-905. 1994
23
95 . McClelland RA. Wilson D, Leake R. Finlay P. Nicholson
96.
97.
98.
99.
100.
101.
102 .
103.
104.
105.
106.
107.
108.
Rl: A multicentre study into the reliability of steroid recep­
tor immunocytochemical assay quantification. Eur J C an ­
cer 27: 711-715. 1991
B arnes DM. Millis RR: Oestrogen receptors: the history,
the relevance and the methods of evaluation: In: Kirkham
N. Lemoine NR (eds) Progress in Pathology. Vol 2. C hur­
chill Livingstone. Edinburgh. 1995. pp 89-114
W alter P. Green S. Greene G. Krust A. B ornert J-M, Jeltsch
J-M. Staubs A. Jensen E. Scrace G. Waterfield M. Cham bon P: Cloning of the human oestrogen receptor. cDNA.
Proc Natl Acad Sci USA 82: 7889-7893. 1985
Kum ar V. Green S. Stack G. Berry M. Jin J-R. Chambon P:
Functional domains of the human oestrogen receptor. Cell
51: 941-951.1987
U m esono K. Evans RM: Determinates of target gene spec­
ificity for steroid thyroid hormone receptors. Cell 57:11391146.1989
Koh EH . Ro J. Wildsick DM. Hortobagyi GN. Buck GN:
Analysis of the oestrogen receptor gene structure of h u ­
man breast cancer. Anti Cancer Res 9:1841-1846.1989
Watts CKW. Handel ML. King RJB. Sutherland RL: O es­
trogen receptor gene structure and function in breast can­
cer. J Steroid Biochem Molec Biol 41: 529-536.1992
Hill SM. Fuqua SAW. Chamness GC. Greene G L.
M cG uire WL: Estrogen receptor expression in hum an
breast cancer associated with an estrogen receptor gene re ­
striction fragment length polymorphism. Cancer Res 49:
145-148.1989
Yaich L. Dupont WD. Cavener DR. Pari FF: Analysis of
the PVU II restriction fragment-iength polymorphism and
exon structure of the estrogen receptor gene in breast can­
cer and peripheral blood. Cancer Res 52: 77-83.1992
A ndersen TI. Heimdel KR. Skrede M. Tveit K. Berg K.
B orresen A-L: Oestrogen receptor (ESR) polymorphisms
and breast cancer susceptibility. Hum G enet 94: 665-670.
1994
Piva R. Rimondi AP. Hanan S. Moestu I. Alvisi A. K um ar
VL. del Senno L: Different methvlation of oestrogen re ­
ceptor DNA in human breast carcinomas with and w ithout
oestrogen receptor. Br J Cancer 61: 270-275. 1990
Lapidus RG. Fergusson AT. Ottaviano YL. Pari FF. Smith
HS. Weitzman SA. Baylin SA. Issa J-PJ. Davidson NE: M e­
thvlation of oestrogen and progesterone receptor gene 5
CpG islands correlates with lack of oestrogen and proges­
terone receptor gene expression in breast tumors. Clin
C ancer Res 2: 305-310. 1996
O ttaviano YL. Issa J-P. Pari FF. Smith HS. Baylin SB. D a ­
vidson NE: Methvlation of the estrogen receptor gene
CpG island marks loss of estrogen receptor expression in
hum an breast cancer cells. Cancer Res o4; 2o52-2oco.
1994
Fersusson AT. Larridus RG. Bavlin SB. Davidson NE: Demethylation of the estrogen receptor gene in estrogen re ­
ceptor-negative breast cancer cells can reactivate estrogen
receptor gene expression. Cancer Res 55: 2279-2283. 1995
109. Sluyser M: Steroid/thyroid receptor-like proteins with on­
cogenic potential: a review. Cancer Res 50: 451-458,1990
110 . Garcia T. Lehrer S. Bloomer WD. Schachter B: A variant
estrogen receptor messenger ribonucleic acid is associated
with reduced levels of estrogen binding in human mam­
mary tumors. Mol Endocrinol 2: 785-791.1988
111. Schmutzler RK. Sanchez M. Lehrer S. Chaparro C. Phillips
C. Rabin J. Schachter B: Incidence of estrogen receptor
polymorphism in breast cancer patients. Breast Cancer
Res Treat 19:111-117.1991
112 . Scott GK. Kushner P. Vigne J-L. Benz CC: Truncated
forms of DNA-binding estrogen receptors in human breast
cancer. J Clin Invest 38: 700-706.1991
113. Fuqua SAW. Allred DC. Elledge RM. Kreig SL. Benedix
MG. Nawaz Z. O'Malley BW. Greene GL. McGuire WL:
The ER-positive/PgR negative breast cancer phenotype is
not associated with mutations within the DNA binding do­
main. Breast Cancer Res Treat 26:191-202,1993
114. Murphy LC. Dotzlaw H: Variant estrogen receptor mRNA
species detected in human breast cancer biopsy samples.
Mol Endocrinol 3: 637-693.1989
115. Dotzlaw H. Ackhalaf M. Murphy LC: Characterization of
estrogen receptor variant in mRNAs from human breast
cancers. Mol Endocrinol 3: 687-693.1989
116. Murphy LC. Hilsenbeck SG. Dotzlaw H. Fuqua SAW: Re­
lationship of clone -i estrogen receptor variant messenger
RNA expression to know prognostic variables in human
breast cancer. Clin Cancer Res 1:155-159.1995
117. Fuqua SAW. Faiette NF. McGuire WL: Sensitive detection
of estrogen receptor mRNA by polymerase chain reaction
assay. J Natl Cancer Inst 82: 858-361.1990
118. Fuqua SAW. Fitzgerald SD. Chamness GC. Tandon AK.
McDonnell DP. Nawaz Z. O'Malley BW. McGuire WL:
Variant human breast tumor estrogen receptor with consti­
tutive transcriptional activity. Cancer Res 51:105-109.1991
119. Fuqua SAW. Fitzgerald SD. Allred DC. Elledge RM. Na­
waz Z. McDonneil DP. O'Malev BW. Greene GL.
McGuire WL: Inhibition of estrogen receptor action by a
naturally occurring variant in human breast tumors. Can­
cer Res 52: -1-83—186.1992
120. Karwik PS. Kuikarni S. Liu X-P. Budd GT. Bukowski RM:
Estrogen receptor mutations in tamoxifen-resistant breast
cancer. Cancer Res 54: 349-353. 1994
121 . Daffala AAI. Johnston SRD. Smith IE. Detre S. King N.
Dowsett M: Exon 5 deletion variant estrogen receptor
messenger RNA expression in relation to tamoxifen resist­
ance and progesterone receptor pS2 status in human
breast cancer. Cancer Res 55: 288-293. 1995
Roodi N. Baiiev LR. Yuo N-Y. Verner CS. Yee CJ. Dupont
LD. Pari FF: Estrogen receptor gene analysis in estrogen
receptor-positive and receptor-negative primary breast
cancer. J Natl Cancer Inst 87: 4 -if> _ i5 i. 1995
12: Rujakariar R. Waiker RA: Pathological and biological fea­
tures of mammographicaily detected invasive breast carci­
nomas. Br J Cancer ”1: 150-154. 1^95
124. Fuqua SAW. Hill SM. Chamness GC. Benedix MG. Green
24
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
13".
138.
GL. O ’Malley BW. McGuire WL: Progesterone receptor
gene restriction fragment length polymorphisms in human
breast tum ours. J Natl Cancer Inst 83: 1157-1160.1991
W alker R A . Varlev JM: The molecular pathology of hu­
man breast cancer. Cancer Surv 16: 31-57,1993
Masiakowski P. Breathnach R. Bloch J. G annon T. Krust
A. C ham bon P: Cloning of cDNA sequences of hormone
regulated genes from the MCF-7 human breast cancer cell
line. EM B O J 10: 7895-7903.1982
May FEB. Westley BR: Cloning of estrogen-regulated
m essenger RNA sequences from human breast cancer
cells. C ancer Res 46: 603-6040.1986
Skilton R A . Luqmani YA. McClelland RA. Coombs RC:
C haracterisation of a messenger RNA selectively ex­
pressed in human breast cancer. Br J C ancer 60:168-175.
1989
H enry JA . Nicholson S. Hennessy C. Lenard TWJ. May
FEB. Westlev BR: Expression of oestrogen regulated pN R
- 2 m R N A in human breast cancer: relation to oestrogen
receptor m R N A levels and response to endocrine therapy.
Br J C ancer 50: 3832-3837.1989
K oerner FC. Goldberg D. Edgerton SM. Schwartz LH: pS2
protein and steroid hormone receptors in invasive breast
carcinomas. Int J Cancer 52: 183-188.1992
W alker R A . Rajakariar R. Dookeran KA: Comparison of
pS2 and cathepsin D in primary breast carcinoma. The
Breast 4: 137-142.1995
Henry JA . Piggott NH. Mailick VK. Nicholson S, Famdon
JR . Westley BR. May FEB: pN R-2,’pS2 immunohisto­
chemical staining in breast cancer: correlation with prog­
nostic factors and endocrine response. Br J Cancer 62:615—
622. 1991
C arr M. May FEB. Lennard TWJ. Westley BR: D eterm i­
nation of oestrogen responsiveness of breast cancer by
com petitive reverse transcription - polymerase chain reac­
tion. Br J Cancer 72: l-C -1 4 3 4 ,1995
Poulson R. Wright NA: Trefoil peptides: a newly recog­
nised family of epitheiiai mucin-associated molecules. Am
J Physiol 265: 205-213.1993
N unez A-M . Berry M. Imier J-L. Chambon P: The 5' flank­
ing region of the pS2 gene contains a complex enhancer
reaion responsive to oestrogens. epidermal growth factor,
a tum our prom oter (TPA). the c-H-ras oncoprotein and
the c-jun protein. EMBO J 8: 323-829. 1989
D ookeran KA. Rye PD. Dearing SJ. W alker RA: Expres­
sion of the pS2 peptide in primary breast carcinomas: com­
parison of membrane ana cytoplasmic staining patterns. J
Pathol 171: 123-129. 1993
Westlev BR. Rochefort H: A secreted glycoprotein in­
duced bv oestrogen :n human breast cancer cell lines. Cell
20: 553-362.1980
Thorpe SM. Rochefort H. Garcia M. Freiss G. Christensen
J. Khalaf S. Paoluccii F. Parr B. Rasmussen BB. Rose C:
Association between high concentrations of 52 k cathepsin
D and poor prognosis in primary breast cancer. Cancer Res
4 9 : o008—
6014. 1989
139. Tandon AC. Clarke GM. Chamness GC. Chirgwin JM.
McGuire WL: Cathepsin D and prognosis in breast cancer.
N Engl J Med 322: 297-302.1990
140. Duffy MJ. Brouillet J-P. Reilly D. McDermott E. O’Higgins N. Fennelly J. Maudelonde T. Rochefort H: Cathepsin
D concentration in breast cancer cytosols: correlation with
biochemical, histological and clinical findings. Clin Chem
37:101-104.1991
141. Garcia M. Derocq P. Pujol P. Rochefort H: Overexpression
of transfected cathepsin D in transformed cells increased
their malignant phenotype and metastatic potency. Onco­
gene 5:1809-1814.1990
142. Walker RA. Denlev H. Dookeran KA: Cathepsin D in
breast carcinomas - the role of the stromal cell component.
Oncology Rep 1: 227-231.1994
143. Thor A, Benz C. Moore D. Goldman E. Edgerton S,
Landry J. Schwartz L. Mavall B. Hickey E. Weber
LA: Stress response protein (srp-27) determination in
primary human breast carcinomas: Clinical histologic and
prognostic correlations. J Natl Cancer Inst 83: 170-178.
1991
144. Manning DL; Robertson JFR. Ellis IO. Elston CW.
McClelland RA. Gee JMW. Jones RJ. Green CD. Cannon
P. Blarney RW. Nicholson Rl: Oestrogen regulated genes in
breast cancer: association of pL IV l with lymph node in­
volvement. Eur J Cancer 30A: 675-678.1994
145. Downward J. Yarden Y. Mayes E. Scarce G. Tottv N.
Stockwell P. Ullrich A. Schiessinger J. Waterfield MD:
Close similarity of epidermal growth factor receptor and
v-erbB oncogene protein sequences. Nature 307: 521-527,
1984
146. Sainsbury JRC. Needham GC. Malcolm A. Famdon JR.
Harris AL: Epidermal growth factor receptor status as pre­
dictor of recurrence and of death from breast cancer. Lan­
cet 1:1398-1402.1987
147. Klijn JGM. Bems PMJJ. Schmitz PIM. Foekens JA: The
clinical significance of epidermal growth factor receptor in
breast cancer: a review of 5232 patients. Endocr Rev 13:
3-17.1992
148. N icholson S. Haicrow P. Fam don JR. Sainsbury JRC.
C ham bers P. Harris AL: Expression of epidermal growth
factor receptors associated with lack of response to endo­
crine therapy in recurrent breast cancer. Lancet i: 182-185,
1989
149. Nicholson Rl. McClelland RA. Gee JMW. Manning DL.
C annon P. Robertson JFR. Ellis IO. Blarney RW: Epider­
mal growth factor recexpression in breast cancer: associ­
ation with response to endocrine therapy. Br Cancer Res
Treat 29: 117-125. 1994
150. H arris AL. Nicholson 5: Epidermal growth factor recep­
tors in hum an breast cancer. Cancer Treat Res 40: 93-118.
1988
151. Chrysogelos SA. Yaraen Rl. Lauber AH. Murphv JM:
M echanisms of EGF receptor regulation in breast cancer
cells. Br Cancer Res Treat 31: 22"-23o. 19^4
152. R oberts AB. Thompson NL. Heme V. Flanders K. Sporn
25
153.
154.
155.
MB: Transforming growth factor beta,: possible roles in
carcinogenesis. Br J Cancer 57: 594-600. 1988
K nabbe C. Lippman ME. Wakefield LM. Flanders KC.
Kasid AL. Dervnck R. Dickson RB: Evidence that trans­
form ing growth factor-P is a hormonally regulated negative
grow th factor in human breast cancer cells. Cell 48: 417428.1987
A rrick BA. Lopez AR. Elfman F. E bner R. Damskv CH.
D erynck R: Altered metabolic and adhesive properties
and increased tumorigenicitv associated with increased ex­
pression of transforming growth factor B,. J Ceil Biol 118:
715-726.1992
D utta A, MacLennan K. Flanders KC. Sacks NPM. Smith
I. M cKinna A. Dowsett M. Wakefield LM. Spom MB.
Baum M. Colletta AA: Induction of transforming growth
factor (3, in human breast cancer
following tamoxi­
fen treatm ent. Cancer Res 52: 4261-4264.1992
Benson JR . W'akefield LM. Baum M. Colletta AA: Syn­
thesis and secretion of transforming growth factor beta iso­
forms by primary cultures of human breast tum our fibro­
blasts
and their modulation by tamoxifen. B r J C an­
cer 74: 352-358.1996
M acCallum J. Keen JC. B artlett JMS. Thom pson AM .
Dixon JM . Miller WR: Changes in expression of trans­
form ing growth factor beta m RN A isoforms in patients
undergoing tamoxifen therapy. Br J C ancer 74: 47*1—178.
1996
W alker RA. Dearing SJ: Transforming growth factor beta,
in ductal carcinoma
and invasive carcinomas of the
breast. E ur J Cancer 28: 641-644.1992
G orsch SM. Moli VA. Stukel TA. Gold LI. A rrick BA: Immunohistochemical staining for transforming growth fac­
tor p, associates with disease progression in human breast
cancer. Cancer Res 52: 6949-6952.1992
D alai BI. Keown PA. Greenberg H: Im munocvtochemical
localization of secreted transforming growth factor P, to
the advancing edges of primary tumours and to lymph
node metastases of human mammary carcinoma. A m J
Pathoi 143: 381-3S9.1993
W alker RA. Dearing SJ. Gallacher B: Relationship of
transform ing growth factor beta, to extracellular matrix
and strom al infiltrates in invasive breast carcinoma. Br J
C ancer 69: 1160-1165.1994
invivo
156.
167.
168.
169.
170.
171.
invitro
157.
158.
172.
173.
in-situ
159.
160.
161.
162.
W alk er R A . G allacher B: D eterm ination of transform ing
g row th factor beta. m RNA expression in breast carcino­
m as by
hybridization. J Pathol 177: 123-127.1995
K em ler R: Classical cadherins. Sem in Ceil Biol 3: 149-155.
174.
175.
176.
situ
177.
178.
179.
insitu
163.
• 1992
K em b ier R: From cadherins to catenins: cytoplasm ic p ro ­
tein in teractio n s and regulation of cell adhesion. T rends
G e n e t 9: 317-321. i9*93
165. V lem inckx K. Vakaet L. M areel MM. Fiers W. Van Roy F:
G en etic m anipulation of E -cadhenn expression by e p ith e ­
lial tu m o u r cells reveals an invasion suppressor role. C ell
6n: 107-119. 1991
166. M oil R. M itze M. Frixen VH. B irchm eier W: D ifferential
164.
loss of E-cadherin in infiltrating ductal and lobular breast
carcinomas. Am J Pathol 143: 1731-1742. 1993
Oka H. Shiozaki H. Kobayashi K. Inove M. Tahara H. Kobavashi T. Takatsuka Y. Matsuyoshi N. Himo S. Takeichi
M. Mori T: Expression of E-cadherin cell adhesion mole­
cules in human breast cancer tissues and its relationship to
metastasis. Cancer Res 53:1696-1701,1993
Lipponen P. Saarelainen E. Ji H. Aaltomaa S. Syrjanen K:
Expression of E-cadherin (E-Cd) as related to other prog­
nostic factors and survival in breast cancer. J Pathol 174:
101-109.1994
Jones JL. Rovall JE. Walker RA: E-cadherin relates to
EGFR expression and lymph node metastasis in primary
breast carcinoma. Br J Cancer 74:1237-1241.1996
Lindblom A. Rotstein S. Skoog L. Nordenskjold M, Larsson C: Deletions on chromosome 16 primary breast carci­
nomas are associated with development of distant metasta­
ses. Cancer Res 53: 3707-3711.1993
Rasbridge SA. Gillett CE. Sampson SA. Walsh FS. Millis
RR: Epithelial (E-) and piacental (P-) cadherin cell adhe­
sion molecule expression in breast carcinoma. J Pathol 169:
245-250.1993
Kanai Y. Oda T. Tsuda H. Ochiai A. Hirohashi S: Point
mutation of the E-cadherin gene in invasive lobular carci­
noma of the breast. Jpn J Cancer Res 85:1035-1039.1994
Berx G. Cleton-Jansen A-M. Nollet F. De Leeuw JF. Van
der Vijver MJ. Comeiisse C. Van Roy F: E-cadherin in a
tumour/invasion suppressor gene mutated in human lob­
ular breast cancers. EMBO J 14: 6107-6115,1995
Kashiwaba M. Tamura G. Suzuki Y. Maesawa C. Ogasawara S. Sakata K. Sotadate R: Epithelial-cadherin gene is
not mutated in ductai carcinomas of the breast. Jpn J Can­
cer Res 86:1054-1059.1995
Hynes RO: Integrins. versatility, modulation and signalling
in cell adhesion. Ceil 69:11-25.1992
Z utter MM. Krigman HR. Santoro SA: Altered integrin
expression in adenocarcinoma of the breast. Analysis by
hybridization. Am J Pathol 142: 1439-1448.1993
Z utter MM. Mazouiian G. Santoro SA: Decreased expres­
sion of integrin adhesive protein receptors in adenocarci­
noma of the breast. Am J Pathol 137: 863-870.1990
Pignatelli M. Hanby AN. Stamp GWH: Low expression of
J},. ou and a , subunits of VLA integrins in malignant mam­
mary tumours. J Pathoi 165: 25-52.1994
Natali PG. Nicotra MR. Botti C. Mottoiese M. Bigothi A.
Segatto O: Changes in expression of apfij integrin hetero­
dimer in primary and metastatic breast cancer. Br J Cancer
66:318-322.1992
Jones JL. Cntchley DR. Walker RA: Alteration of stromal
protein and integrin expression in breast - a marker of premalignant change ’ J Pathol 167: 399—406. 1942
Pignatelli M. Cardiilo MR. Hanby A. Stamp GWH: Inte­
grins and their accessor, adhesion molecules in mammary
carcinomas: loss of polarization in poorly differentiated tu­
mours. Hum Pathol 25: 1159-1166. 1992
Blasi F. Vassalli JD. Dano K: L'rokinase-tvpe plasminogen
180.
181.
182.
in
26
153.
154.
155.
186.
187.
188.
189.
190.
191.
192.
activate: proenzyme, receptor and inhibitors. J Cell Biol
104:801-804.19S7
Duffy MJ. Reilly D. O'Sullivan C. O'Higgins N, Fennelly
JJ. A ndreasen P: Urokinase-plasminogen activator, a new
and independent prognostic m arker in breast cancer. Can­
cer Res 50: 6S27-6S29.1990
Janicke F, Schmitt M. Pache L. Ulm K. Harbeck N. Hoefler
H. Graet'f H: Urokinase (uPA) and its inhibitor PAI-1 are
strong independent prognostic factors in node negative
breast cancers. Br Cancer Res Treat 24:195-208,1993
G rondhahl-H ansen J. Christensen IJ. Rosenquist C. Brun­
ner N. M ouridsen HT. Dano K. Buchet-Toft M: High levels
o f urokinase-type plasminogen activation and its inhibitor
PAI-1 in cytosolic extracts of breast carcinomas are associ­
ated with poor prognosis. Cancer Res 53: 2573-2521.1993
Bouchet C, Spvratos F. Martin PM, Hacene K. Gentile A.
O globine J: Prognostic value of urokinase-tvpe plasmino­
gen activator and its inhibitor PAI-l and PAI-2 in breast
carcinomas. Br J Cancer 69: 398-405.1994
C onstantini V. Sidoni A. Deveglia R. Cazzoto OA. Bellezza G. Ferri I. Bucciareili E. Nenci GG: Combined over­
expression of urokinase, urokinase receptor, and plasmi­
nogen activator inhibitor-1 is associated with breast cancer
progression. Cancer “ : 1079-1088. 1996
Pvke C. G raem N. Raifkiaer N. H oever-Hansen G. Brun­
ner N. D ano K: Receptor for urokinase in present in tu­
mour-associated macrophages in ductal breast carcinoma.
Cancer Res 53: 1911-1915.1993
Bianchi E. Cohen RL. Thor AT. Todd III RF. Mizukami IF.
Lawrence DA. Ljung BM. Shuman MA. Smith HS: The
urokinase receptor is expressed in invasive breast cancer
but not in normal breast tissue. Cancer Res 54: 841-866.
1994
W oessner JF: Matrix metalloproteinases and their inhib­
itors in connective tissue remodelling. FASEB J 5: 21452154.1991
Powell WC. Matrisian LM: Complex roles of matrix metalloprotetnases in tumour progression. In: G unthert U.
Schleg PM. Birchmeier W teds) A ttem pts to U nderstand
M etastasis Formation. I. Metastasis R elated Molecules.
Springer-Verlag. Heideiburg. 1996. pp 1-21
C a w s t o n T E : P ro tein inhibitors o f m e t a l l o p r o t e i n a s e s . In:
B a r n e t t A J . Salvese.n G teds) P r o t e in a s e I n h ib ito r s. E l s e ­
vier . A m s t e r d a m . L986. pp 5S9-6L0
193.
L’ria J A . F e r r a n d o A A. Velasco G. F reije JM P . L o p e z - O t i n
C: S t r u c t u r e an d expression in b r e a st t u m o u r s o f h u m a n
T I M P - 5 . a n e w m e m b e r of the m e t a i l o p r o t e i n a s e in h i b i t o r
famiiv. C a n c e r Res 54: 2091-2094. 1994
194.
M o n t e a u u d o C. Merino MJ. San JJ. L io tta L A . Stetler-S tev e n s o n W G : I m m u n o n is to c h e m ic a l d i s t r i b u t i o n o f typ e IV
c o l l o u e n a s e in norm al, oenign a n d m a l i g n a n t b r e a s t tissue.
195.
A m J P a t h o l 136: 55 5-5 Q2. 1990
D a v ie s B. M iles D W H a p p e r t i e l d LC. N a y l o r MS. B o b r o w
L G . R u b e n s RD. Baikwiil FR: Activity o f t y pe IV collagen as es in benuzn and maiiunant b r e a s t di sease. Br J C a n c e r
07: 1126-1131. 1993
196. D 'Errico A. Gerbisa S. Liotta LA. Castronovo V. StetlerStevenson WG. Grigioni WF: Augmentation of type IV
collagenase. laminin receptor and Ki 67 proliferation anti­
gen associated with human colon, gastric and breast carci­
noma progression. Mod Pathol 4: 239-246. 1991
197. Iwata H. Kobavashi S. Iwase H. Masaoka A. Fujimoto N.
Okada Y: Production of matrix metalloproteinases and tis­
sue inhibitors of metalloproteinases in human breast carci­
nomas. Jpn J Cancer Res 87: 602-611.1996
198. Basset P. Bellocq JP. Wolf C. Stoll I. Hutin P. Limacher JM.
Podhajcer OL. Chenard MP. Rio MC. Chambon P: A novel
metailoproteinase gene specifically expressed in stromal
cells of breast carcinomas. Nature 348: 699-704,1990
199. Kawami H. Yoshida K. Ohsaki A. Kuroi K. Nishivama M.
Toge T: StromeIvsin-2 mRNA expression and malignancy:
comparison with clinicopathological features and type IV
collagenase mRNA expression in breast tumors. Antican­
cer Res 13: 2319-2324.1993
200. Hahnel E. Harvey JM. Joyce R. Robbins RD. Sterrett GF,
Hahnel R: Stromelysin-3 expression in breast cancer biop­
sies: clinicopathological correlations. Int J Cancer 55: 771—
774,1993
201. Engel G. Heseemeyer K. Auer G, Backdahl M. Eriksson E,
Linder S: Correlation between stromelysin-3 level and out­
come of human breast cancer. Int J Cancer 58: 1-7.1994
202. Onisto M. Riccio MP. Scannapieco P. Caenozzo C. Griggio
L. Spina VI. Stetler-Stevenson WG. Garibsa S: Gelatinase
A/TIMP-2 imbalance in lymph node positive breast cancer
as measured by RT-PCR. Int J Cancer 63: 621-626.1995
203. Kossakowska AG. Huchcroft SA. Urbanski SJ. Edwards
DR: Comparative analysis of the expression patterns of
metalloproteinases and their inhibitors in breast neoplasia,
sporadic colorectal neoplasia, pulmonary carcinomas and
malignant non-Hodgkin's lymphomas in humans. Br J
Cancer 73: 1401-1408.1996
2 0 4 . Poulsom R. Hanby A.M. Pignatelli M. Jeffrey RE. Longcroft JM. Rogers L. Stamp GWH: Expression of gelatinase
A and TIMP-2 mRNAs in desmoplastic fibroblasts in both
mammary carcinomas and basal carcinomas of the skin. J
Clin Pathol 46: 429-436. 1993
205. Soini Y. Hurskainen T. Hoyntya M. Oikarinen A. AutioHarmainen H: 72 kD and 92 kD type collagenase. type IV
collagen and laminin mRNAs in breast cancer: a study by
hybridisation. J Histochem Cytochem 42: 945-951.
1994
206. Emonard H. Remack A. Boel A. Grimaud JA. Stetler-Ste­
venson WG. Foidart JM: Tumour cell surface associated
binding site for the Mr “2.000 type IV collagenase. Cancer
Res 52: 5845-5848.1992
20~. Sato H. Takino T. Okado Y. Cjo J. Shinagawa A. Yamamo­
to E. Seiki M: A matrix metailoproteinase expressed on the
surface of invasive tumour ceils. Nature 3”0: oi-o5. 1994
208. Will H. Hinzmann B: cDNA sequence and mRNA tissue
distribution of a novel human matrix metailoproteinase
with a potential transmembrane segment. Eur J Biochem
231: 602-608.1665
insitu,
27
209. Sato H . Seiki M: Membrane-type matrix m etalloproteinas­
es (M T-MMPs) in tumour metastasis. J Biochem 119: 209215.1996
210. Takino T. Sato H. Shinagawa A, Seiki M: Identification o f
the second membrane type matrix m etailoproteinase (MTM M P2) gene from a human placenta cD NA library: MTM M P form a unique membrane-type subclass in the M M P
family. J Biol Chem 270: 23013-23020,1995
211. N om ura H . Sato H, Seiki M. Mai M, O kada Y: Expression
of m em brane-type matrix metailoproteinase in hum an
gastric carcinomas. Cancer Res 55: 3263-3266,1995
212. Tokuraku M. Sato H. Murakami S. Okoda Y, Watanabe
Y.Seiki M: Activation of the precursor of gelatinase
A/72 kDa type IV collagenase/MMP-2 in lung carcinomas
correlates with the expression of membrane-tvpe matrix
metailoproteinase (MT-MMP) and with lymph node me­
tastasis. Int J Cancer 64: 355-359,1995
Addressforoffprint:R.A. Walker. Breast Cancer Research Unit.
Clinical Sciences. Glenfield Hospital, Groby Road, Leicester,
LE3 9QP, UK