Cell, Vol. 50, 909-915,
September
11, 1987, Copyright
0 1987 by Cell Press
Functional Redundancy and Structural Polymorphism
in the Large Subunit of RNA Polymerase II
Michael Nonet, Doug Sweetser, and Richard
Whitehead Institute for Biomedical Research
Nine Cambridge Center
Cambridge, Massachusetts 02142
and Department of Biology
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
A. Young
Summary
The RNA polymerase
II large subunit contains tandem
copies of the sequence Pro Thr Ser Pro Ser Tyr Ser at
its carboxyl terminus, the number of which varies from
26 in yeast to 52 in mice. Our results indicate that the
heptapeptide
repeat sequence is unique and essential
to RNA polymerase
II. We have determined
that a portion of the heptapeptide
repeat domain lsessential
for
viability by constructing
and analyzing unidirectional
deletions of the carboxy-terminal
coding sequence in
yeast. Cells containing
an RNA polymerase
II large
subunit
with less than 10 complete
heptapeptide
repeats are inviable, those containing
lo-12 complete
repeats are conditionally
viable, and those with 13 or
more complete repeats are unconditionally
viable. The
inviable deletion mutants studied here have truncated
RNA polymerase
subunits that are stable, but functionally deficient. Finally, the number of repeat units
is polymorphic
in wild-type
yeast strains. These results have implications
for the function of this unusual
sequence in transcription.
Introduction
Eukaryotic nuclear RNA polymerases I, II, and Ill synthesize ribosomal precursor RNA, pre-mRNA, and small stable RNAs, respectively (Lewis and Burgess, 1982; Sentenac, 1985). These enzymes share the ability to bind
DNA, initiate an RNA chain de novo, elongate the nacent
RNA, and terminate and release the RNA product. Two
features of the transcription reaction differ significantly for
the three polymerases: the recognition of different promoters and the complexity of the DNA transcribed. RNA
polymerase II transcribes a much greater portion of genomic DNA sequences and responds to a greater diversity of gene expression signals than do the other two enzymes.
The three RNA polymerases are complex enzymes that
share certain structural features (Paule, 1981; Sentenac,
1985). They are composed of 9-14 polypeptides, among
which are two very large (135-220 kd) and many relatively
smaller (less than 50 kd) proteins. The two large subunits
of RNA polymerase II share some sequence homology
and antigenic determinants with the corresponding
subunits of RNA polymerases I and III. The two large subunits
also share sequence homology with the two large subunits of prokaryotic RNA polymerases (Allison et al., 1985;
Sweetser et al., 1967). Of the smaller polypeptides, three
are generally shared by all three eukaryotic RNA polymerases, two additional proteins are shared by RNA polymerases I and III, and the remaining polypeptides appear
to be unique components of the enzymes.
The largest subunit of the eukaryotic RNA polymerase
II has an unusual component that is absent from its counterparts in RNA polymerases I and Ill and in prokaryotic
RNA polymerases. The carboxyl terminus of this protein
consists of 26 (in yeast) to 52 (in mice) tandem copies of
a highly conserved 7 amino acid sequence, followed by a
number of unrelated residues (Allison et al., 1985; Corden
et al., 1985). The function of the heptapeptide repeat is not
known. The 7 amino acid sequence itself is unusual, consisting of the consensus Pro Thr Ser Pro Ser Tyr Ser. The
structure formed by this repeat is unclear, but the high proline content would prevent the formation of alpha-helical
structure. To obtain clues to the function of the heptapeptide repeat domain, we have examined the features of the
repeat that are essential for cell viability.
Results
and Discussion
Construction
of a Series of Deletion Mutants
A series of unidirectional deletions was generated in DNA
encoding the carboxyl terminus of the RNA polymerase
large subunit (RP67) by using the strategy diagrammed in
Figure 1. Each of the deletions constructed in this manner
replaced a 180 bp region 3’of the R/%37 coding region with
a BamHl linker such that all of the deletions contain identical sequences 3’ of the BamHl linker. While the 180 bp
region contains the natural RPB7 polyadenylation
site
(Nonet et al., 1987) replacement of this region alone with
a BamHl linker did not affect the growth of the cell in acontrol experiment (see below).
Two independent approaches were taken to investigate
the effects of the RPB7 deletions on the viability of yeast
cells. First, plasmids containing various deletion8 were
transformed into E. coli and clonally purified. Each of the
plasmids was mapped with BamHl and other restriction
endonucleases
to determine the approximate
position
of the deletion endpoint. Deletion endpoints occurred
throughout DNA encoding the amino acid repeat and in
coding sequences upstream of the repeat. Yeast cells
were then transformed separately with approximately 100
different purified plasmids containing
RPB7 carboxyterminal deletions to determine the in vivo phenotype conferred by the mutations, as described below.
The second approach used to study the effect of RPB7
carboxy-terminal
deletions on cell growth was to screen
directly for mutants that exhibited a conditional lethal
phenotype. Yeast cells were transformed directly with the
library of deletions constructed in vitro, cells that were
temperature-sensitive
and cold-sensitive were isolated as
described below, and the plasmid of interest was isolated
and further characterized.
Cell
910
1
Cut with SnaBl and Nsil. trim with
exonuclea~es
Ill and VII.
m
D-i lE”P
VblAC
W-{
rpt.1‘G
1
Add linker, religate, and transform
Isolate individual clones.
E. coli
Transform
yeast with individual clones,
select for plasmid encoded LEUZ marker.
1
1
Replica plate to 5.FOA plates
to select for cells which have
lost URA3 plasmid.
Final strain contains
the partially deleted
of the RPBl gene
Figure 1. Construction
the CarboxyTerminal
of Unidirectional
Repeat of RPE7
Deletions
in DNA
only
copy
Encoding
Plasmid pRP114 was digested with SnaBl and Nsil and subsequently
with exonuclease
Ill to unidirectionally
delete sequences
from the
carboxy-terminal
repeat coding region (shaded in black) of RPB7. After
blunting the ends with exonuclease
VII, a BamHl linker was added to
mark the deletion endpoint. Individual clones (pRP114A) were isolated
in E. coli, and plasmid DNA was prepared. The yoati strain used, 226,
has a chromosomal
replacement
of RPBI with MS3 and a URA3, RPEl
centromere
plasmid (pRP112) to supply wild-type RPBl function. The
plasmid pRP112 was replaced with a mutant version of RP87 carried
by plasmid pRP114A by transforming
226, selecting for Leu+ cells,
and replica-plating
to 5FOA plates to select against the URA3 gene.
The resulting strains, if viable, contain only the mutant pRP114A plasmid to supply RPEV function.
Viability of Carboxy-Terminal
Deletions
The phenotypes produced by the deletion mutants were
investigated by using a plasmid shuffle technique (Boeke
et al., 1987) in the S. cerevisiae strain 228 (Figure 1). The
complete chromosomal copy of the RF%1 gene has been
replaced with a HIS3 gene in this strain, and a centromere
plasmid that contains the RPBl and U&I3 genes complements the RNA polymerase II defect. To assay the effect
of various deletions on cell viability in vivo, the Leustrain 226 was transformed with centromere plasmids
containing LEUP and the RPBl deletions and Leu+ cells
were replica-plated
to media containing 5-fluoro-erotic
acid (5-FOA) (Figure 2). This media permits the growth of
only those cells that have lost the plasmid carrying the
wild-type copies of RPBl and lJRA3. Thus, each of the
5-FOA-resistant cells is left with the centromere plasmid
containing LEUP and the R/W
deletion mutation. The
ability of Leu+ cells to grow in the presence of 5-FOA indicates that the R/W
deletion can complement the complete chromosomal deletion of RfBl. In contrast, the inability of Leu+ cells to grow on 5-FOA medium indicates
that the R&31 deletion cannot complement the R/W
chromosomal deletion.
The carboxy-terminal
deletion mutants fall into three
categories. In the absence of a wild-type copy of the RPBl
gene, cells harboring some RPBl carboxy-terminal
deletions are completely viable, others are conditional for viability, and a third set are inviable. Plasmids with deletions
that produced viable cells were designated pV, those that
produced conditionally viable cells were designated PC,
and those that produced nonviable cells were designated
pN. Strains containing these plasmids are designated V,
C, or N (N strains, unlike V and C strains, carry in addition
to the pN plasmid a complementing
pRP112 plasmid).
Based on the position of the BamHI linker, there was no
obvious relationship between the extent of the deletion
and the viability of the carboxy-terminal
deletion mutants.
However, all of the deletions that produced conditional
phenotypes had endpoints that mapped to a small region
in the middle of the repeat.
Sequence Analysis of Catboxy-Terminal
Endpoints
To obtain more precise information about the nature of
each deletion, the exact deletion endpoint was determined by sequence analysis (Figure 3). Since the sequences 3’ to the linker insertion site are identical in all
of the clones, an oligonucleotide
primer could be used to
determine the sequence of the deletion endpoints in each
of the plasmids. Forty-eight unique deletion endpoints
were elucidated, 38 of which occurred in the repeat sequences. Multiple isolates were obtained for some of the
mutations. Since the sequence of one repeat unit differs
from adjacent repeats by an average of 6 of 21 nucleotides, it was possible to determine the precise position of
each deletion endpoint. It is worth noting that the appreciable variation of the heptapeptide coding sequence
indicates that the repeat protein, rather than the repeat
DNA, is responsible for the function under investigation.
The results summarized in Figure 3 do reveal a relationship between the viability of the carboxy-terminal
deletion
mutants and the amount of amino acid repeat that is
deleted when the influence of the nonrepeat “tail” is taken
into account. The RPBl carboxyl deletion mutants end in
one of 3 amino acid sequences depending on the reading
frame at the nucleotide adjacent to the BamHl linker (Figure
4). Deletion mutants whose nucleotide endpoints occur in
frames A and C add short carboxy-terminal
tails of 10 and
3 amino acids, respectively, to the remaining portion of the
heptapeptide
repeat. Mutants whose endpoints end in
frame B have a long carboxy-terminal
tail of 39 amino
RNA Polymerase
911
II Heptapeptide
Repeat
5-FOA, -LEU
-LEU, -URA
12OC
Figure
Strain
amino
Strain
Plates
in the
2. Analysis
of the Viability
3o”c
of Deletion
38’C
Mutants
226 was transformed
with pRP114 plasmids containing RPB7 deletions, and single Leu+ colonies were patched onto plates containing
all
acids except leucine (DO-LEU). (A) The cell patches were replica-plated
to plates containing DO-LEU, DO-LEU-URA,
and DO-LEU plus 5-FOA.
226 transformants
that grew on DO-LEU plus 5-FOA were clonally purified and tested for cold-sensitivity
(B) and temperature-sensitivity
(C).
were incubated for 24 hr at 36% (A), for 9 days at 12% (B), or for 36 hr at 30°C or 38% (C). The plasmid and strain nomenclature
is described
text.
acids added to the heptapeptide repeat. of the 48 unique
deletion endpoints examined, 8 different deletions were
isolated that end with the long carboxy-terminal
tail and
are inviable, independent of the heptapeptide repeat deletion size. In three additional
cases, deletion mutants that
should end in frame B are viable. Here the deletion endpoint nucleotides TA create a termination codon with the
BamHl linker sequence GGATCC so that translation of
RPB7 mRNA ends at the linker. These results indicate that
the ultimate nonrepeat carboxy-terminal
amino acids of
this subunit are not essential for viability, although the 39
amino acid tail has a deleterious effect.
There is a striking correlation between cell viability and
the length of the RP67 heptapeptide repeat domain that
appears when the RPB7 deletion mutants ending with the
long carboxy-terminal
tail encoded in frame B are eliminated from the analysis. The 9 deletion mutants that are
nonviable at all temperatures all have endpoints in DNA
encoding the first 10 heptapeptide repeats. All of the remaining deletion mutants are viable and appear in DNA
encoding 11 or more heptapeptide repeats. Thus, it appears that the RPB7 protein must contain at least 10 complete heptapeptide repeats to produce RNA polymerase II
in a form essential for cell viability.
A most interesting correlation exists between the length
of the heptapeptide repeat and conditional viability. Five
deletion mutants exhibit temperature-sensitive,
cold-sensitive, or both conditional phenotypes. All 5 mutants have
deletion endpoints that occur in DNA encoding heptapeptide repeats 11, 12, and 13. This cluster defines the transition region between deletion endpoints that produce inviable cells and those that produce viable cells. The
conditional
phenotypes of these mutants indicate that
their RNA polymerase
or simply structurally
tures.
II enzyme is functionally inefficient
unstable at the extreme tempera-
Effect of Deletions on Cell Growth Rate
To determine whether deletion length has an effect on
growth rate, we assayed the doubling time of selected
strains carrying different deletions (Table 1). Carboxyterminal deletions that left intact at least 13 complete heptapeptide coding sequences have doubling times that are
not significantly different from the wild-type strain. In contrast, two conditionally lethal deletion mutants, C23 and
C3, have doubling times that are approximately 10% and
35% slower than wild-type strains, respectively. Although
growth rates were not determined for all of the deletion
mutants, the colony sizes of all of the unassayed mutant
Table
1. Growth
Rate of RPB? Deletion
Doubling
Mutants
Time
Strain
YPD
SC
222
227
VI4
v20
V8
v2
v3
v7
v5
C23
c3
86
87
89
90
89
90
88
87
93
94
120
105
102
105
106
107
105
103
104
109
114
146
Doubling
described
times are presented
in minutes.
in Experimental
Procedures.
YPD and SC media
are
Cell
912
(11 p T
(2) v s
NSIC
(3) p T
N12-A
(4 p T
(6)
p
171
p
(8)
(9)
P
N1I.A
P
(10)
P
(111
p
(12)
p
(13)
P
T
T
T
T
T
T
T
T
T
114)
P
T
WI
P
T
(5)
P
(16)
P
(17)
P
118)
P
w
P
T
T
T
T
(201
p
T
(21)
p
T
T
T
p
(23) p
WI
(24)
P
(25)
p
(26)
p
T
T
T
(27)
p
G
S
S
S
S
s
to44
s
s
“ii?
s
s
s
s
Cl-A
s
F
P G
F
P
G F
W-A N31-8
P
T
Y
wi3*
P
A
Y
P
s
Y
NW-A
P
s
Y
P
s
Y
P
s
Y
P
s
Y
P
s
Y
NIS-A
P
s CY,
P
s
Y
cat
P s
Y
P
s
G
NW.6
S
NlCB
S
S
s
s
s
s
s
s
“‘;“’
s
s
s
s
3. Summary
of RP67
RPBl repeat
S
s
s
s
P
P
VISA
P
S
s
S
S
P
P
P
P
Linker
DNA 3’of RPEl
-f-----Y-
. . . ..‘NNNNNNGGATCCTTTAAA......
AU
FRAMES
BU
c-
6
FRAME A -GSFKLRPSNF
FRAME B
FRAME C
(IO)
-XDPLNYVRPTSKLCMPlPCGGGAQRPQPTrGLLPNAGVA
-XIL (3)
Figure 4. The Three Possible Translation Frames That Encode
Acid Sequences
beyond the Heptapeptide
Repeat
(391
Amino
(A) The structure of RPBl deletion endpoints consists of the remaining
repeat coding region, the hexanucleotide
BamHl linker, and sequences 3’ of the RPBI coding region that begin with the 3’ thymine
of the Nsil site described in Figure 1. The reading frame at the junction
between the linker and the RfB7 deletion endpoint is defined as A, B,
or C.
(6) The amino acids that constitute each tail sequence
in frames A,
B, and C are listed.
A
B
C
D
Y
s
Y
s
Y
“2-A
s
Y
“1.6’
N Y
s
Y
G Y
A “v”
s
s
s
N2-0
S
s
S
y
PKQDEQKHNENENSR
Figure
A
Carboxy-Terminal
Deletion
Figure 5. Western
Terminal Deletion
Blot Analysis
Mutants
of Cell Lysates
of Selected
Carboxy-
Crude extracts were subjected to SDS-PAGE,
transferred
to nitrocellulose, and probed with antibodies as described
in Experimental
Procedures. (A) Wild-type strain 227. (B) Mixture of strains 227 and C3. (C)
Strain C3. (D) Strain Nil.
Endpoints
The amino acid sequence
of the RPBI heptapeptide
repeat is shown
in single letter code. The repeat number is given to the left of each repeat. The deletion endpoint of each mutant is shown by placing the
mutant designation
under the first amino acid altered by the deletion
endpoint. RPB7 carboxy-terminal
deletion endpoints that produce viable, conditionally
viable, and nonviable cells are designated V, C, and
N, respectively.
The mutant designation
is followed by a frame desig
nation (A, 8, or C) that indicates the amino acid tail added to the heptapeptide repeat (see Figure 4). Exceptions
to these frame designations are frame A* (this clone did not contain a linker and encodes the
7 amino acid tail KLRPSNF); frame B’ (these clones contain no tail because the repeat sequence-linker
junction produces
a termination
codon); and frame C’ (this clone contains 2 linker units encoding the
sequence
WIWIL).
strains indicate that they do not have significant growth
defects. These results suggest that an RNA polymerase
II large subunit carboxyl terminus consisting of 13 or more
complete heptapeptide
repeats is functionally equivalent
to that of wild-type cells. The wild-type repeat number
probably reflects optimization
of function through evolution.
Effect of Deletions on RPBl Protein
To provide a more direct demonstration that the DNAdeletion mutants produce RPBl protein that is truncated and
to investigate the stability of these truncated proteins, lysates of two cells harboring plasmids with large deletions
and a lysate of a wild-type cell were studied by Western
analysis. One of the plasmids produces viable and one
produces inviable cells in the plasmid shuffle assay.
Crude protein extracts were subjected to SDS-PAGE on
7.5% gels, electroblotted to nitrocellulose, and incubated
with anti-RNAP
II antibodies. The results demonstrate
that the RPB7 proteins present in the deletion strains are
truncated versions of the wild-type protein (Figure 5).
Similar results were obtained with 5 other deletion mutants. This analysis demonstrates that these truncated
proteins accumulate, but are functionally defective.
Polymorphism
in the Carboxy-Terminal
Repeat of Yeast
Sequence analysis of the carboxy-terminal deletion clones
revealed that DNA encoding the RF’67 heptapeptide re-
RNA Polymerase
913
II Heptapeptide
Repeat
CCCAACATACTCTCCTACCTCTCCAGCGTACTCACCAACATCACCATCGT
100
ACTCACCAACATCACCATCGTACTCGCCAACATCACCATC
ACATCACCATCGTATTCACCAACGTCACCATCATATTCGCCmCGTCACC
200
ATCATATTCGCCAACGTCGCCATCGTATTCTCCAACGTCACCATCGTATT
CGCCAACGTCGCCTTCCTACTCTCCCACGTCGCCAAGCTACAGCCCTACG
300
(C) 1026
(D) 907
TTACAGCCCAACGTCACCAAGTTACAGCCCAACGTCTCCAGCCTATTCCC
r*,
q.
400
CAACATCACCAAGTTATAGTCCTACATCGCCTTCATACTCTCCAACATCA
CCATCCTATTCCCCAACATCACCTTCTTACTCTCCCACCTCTCCAAACTA
--f
+
(E)
683
+
(F)
515
+
c
-
c*Iw
500
TAGCCCTACTTCACCTTCTTACTCCCCAACATCTCCATCTCCAGGCTACAGCCCAG
c
1078
+
072
+
603
-
600
GATCTCCTGCATATTCTCCAAAGCAAGACGAACRRAAGCAT
GAAAATTCCAGATGATATAGTATATCATCCTTACGTATTTGACGTTATTA
CATTATATATAGTTTCTAAATAATRTTTCTATTTCTAGTTTATTTTTGTATCATM
700
TAAAAACGTATACCAAATATACCATTATTTTTCATAGCATTATGGTAGGG
ATAGGGAATCAAGTAACTAATTTATATCCGCAGAGCATTGGGAAAACCAA
800
ATGGTAGACTCTTAACTCTGACCTTTTTAGCAATTAAGCTCTTGAAGATA
.._-.
1
CGGCGCTAGTAAATGCATTTAAATTACGTCCGTCCAACTTCTAAGCTTCA
RPM
900
TCAAAAGTGTTACCGTCCGG
Figure 6. DNA Sequence
Repeat in Strain S268C
of the Region
Encoding
the Heptapeptide
The sequence
is numbered
from the beginning of the heptapeptide
coding region. The large underlined sequence
is a 42 nucleotide duplication of adjacent repeat sequences
that exists in S288C relative to an
A364A derivative
(Allison et al., 1965). Two single nucleotide differences between the two strains are also underlined. The A indicates the
position of a 21 nucleotide deletion that occurs in S288C relative to the
A364A derivative.
peat in S. cerevisiae strain S288C differs from that prevrousty published for an A364A derivative (Allison et al.,
1985). The two strains differ at 4 sites in the heptapeptide
coding sequence (Figure 6). A 42 nucleotide duplication
adds exactly 2 repeat units and a 21 nucleotide deletion
removes 1 repeat unit, producing a net addition of 1 repeat
unit to the S288C RR37 carboxyl terminus relative to the
A364A derivative. In addition, this DNA segment differs by
2 single nucleotide substitutions that do not alter the
amino acid sequence of the carboxyl terminus in the two
strains.
The polymorphism
observed between this portion of
RPBl
in S228C and the A364A derivative suggested that
yeast strains might generally differ in the number of DNA
repeat units encoding the heptapeptide. To investigate this
possibility, Mspl restriction digests of genomic DNA from
a variety of Saccharomyces strains was probed with RPB7
DNA (Figure 7). The results demonstrate that 2 of the 6
RPB7 DNA fragments are polymorphic
in length. Of these,
the darker band encodes the repeat region of the RPt31
gene. The darker intensity of this band is due to more efficient hybridization of repeat sequences since the effective
concentration of repeat sequences is higher than a standard single copy genomic sequence. Two distinct repeat
sequences are found among these research strains. The
Figure 7. Restriction
Length Polymorphisms
among RPBl
tide Coding Sequences
in Various S. cerevisiae Strains
Heptapep
(Upper panel) Genomic DNA from strain X2180-2 (an S286C diploid
from which the RtW gene was cloned) (Young and Davis, 1983) from
wild-type strain DE1827 (an S286C derivative),
and from canonical
S. cerevisiae
strains S288C, A364A, and FL100 (Mortimer
and Johnston, 1986) was digested with the restriction endonuclease
Mspl, subjected to electrophoresis
on a 3% agarose TAE gel, transferred
to
nitrocellulose,
and probed with the labeled 7.1 kb R/W
Hindll DNA
fragment from strain X2180-2 as illustrated in the lower panel. Mspl
digestion creates a single DNA fragment containing all but 10 bp of the
heptapeptide
coding sequence.
In addition, all other Mspl DNA fragments containing R/W
DNA are electrophoretically
distinguishable
and of small enough size to visualize
relatively small changes
in
length. The positions of molecular weight markers are given on the
right, and the molecular weights of the Mspl fragments
of the RfBl
gene (marked A through F) are shown on the left.
(Lower panel) A partial restriction
map of the RPBl gene of strain
X2180-2. Mspl restriction
sites are marked above the line; Hindlll restriction sites, below the line. The Mspl fragments
present on the autoradiograph
in (A) are marked A through F. The RPBl mRNA transcript and RPB7 coding region (with the repeat region shaded) are also
schematically
depicted.
S288C strain, strains used in this laboratory, and the
S. bayanus strain all contain DNA encoding 27 heptapeptide repeats. In contrast, RP67 DNA in strains A364A,
FLlOO, and S. carlbergensis
appears to encode 26 heptapeptide repeats. Finally, since hybridization was done at
low stringency, the Southern analysis demonstrates that
the heptapeptide coding sequence occurs only in RPBl
DNA, and thus not in DNA encoding subunits of the other
two classes of RNA polymerase, I and Ill.
The other RPB7 DNA fragment that is polymorphic in
length contains the RPB7 promoter. This polymorphism
may account for functional differences in the ability of specific RPB7 DNA fragments to complement null mutations
in RPB7 in the S288C and A364A derivative strains (Allison et al., 1985; Nonet et al., 1987). The results presented
Cdl
914
Table
2. Yeast
Strains
Name
Genotype
Alias
Source
222
223
226
227
S268C
A364A
FL100
X21 80-2
MATa ura552 his3A200 /eu2-3 /eu2-7 12
MATa ura552 his3A200 leu2-3 leu2-112 rpbld 196::HIS3
MATa ura352 his3A200 /eu2-3 /eu2-712 rpbld 187::HIS3
MATa ura552 his3A200 /eu2-3 leu2-7 72 rpblAl87::HIS3
MA Ta ma/-gal2
MATa adel ade2 gall Ural his7 lys2 tyrl
MA Ta
MATaIMATa
ma/-/ma/ga12/ga/2
DBY1627
N121
N247
N249
DBY34
DBY24
DBY613
Y4
D. Botstein
This lab
This lab
This lab
D. Botstein
D. Botstein
D. Botstein
This lab
here indicate that the RPBl gene varies both in promoter
and in heptapeptide coding sequences in the S288C and
A384A strains.
Models of Heptapeptide
Repeat Function
What is the function of the RPB7 heptapeptide repeat sequence? RNA polymerase II transcribes a much greater
portion of genomic DNA sequences and responds to a
greater diversity of gene expression signals than RNA
polymerases I and III. The heptapeptide
repeat domain
occurs only in the large subunit of RNA polymerase II in
eukaryotes; it is not found in RNA polymerases I or III; nor
is it found in prokaryotic RNA polymerase (Ovchinnikov et
al., 1982). The heptapeptide repeat domain does not appear to play an essential role in DNA binding or RNAcatalysis since a form of the RNA polymerase II enzyme that
lacks the carboxy-terminal
repeat (Allison et al., 1985) is
capable of transcribing DNAsequences
nonselectively in
vitro (Dezelee et al., 1978). Considering these facts, we
propose four models for repeat function that are consistent with the results presented here.
The heptapeptide might interact with some Pans-acting
transcription factors. Reiteration of a binding site would increase the concentration of that site and might reduce the
amount of any one specific transcription factor necessary
for function. Alternatively, the carboxyl terminus could act
as a “cow-catcher,” facilitating removal of histones and
DNA binding proteins that might otherwise inhibit elongation. For example, it might act as a transient site for DNA
binding proteins that are temporarily removed during transcription. The repeat domain could provide a site for
general transcription control. As a site for factor binding,
phosphorylation,
or proteolysis, it could function to control
RNA polymerase activation or inactivation. Finally, the
heptapeptide repeat domain could perform a function entirely unrelated to catalysis. For instance, it might furnish
a means to localize the enzyme within the nucleus.
The RNA polymerase II large subunit deletion mutants
that exhibit a conditional phenotype all have deletion endpoints that cluster in the middle of the yeast repeat sequence, in the region that defines the boundary between
deletions that produce inviable cells and those that produce viable cells. Extragenic suppressors of these conditional mutants may allow us to infer the function of the
carboxy-terminal
repeat by revealing gene products that
interact with it.
$RPi 121
[pRPl12]
$RP114]
Experimental
Pmcedures
Yeast Strains and Media
Yeast strains are listed in Table 2. The rpblA796::H/S3
allele consists
of a replacement
of the 1 kb Bglll fragment containing RPB7 coding
sequences
with a 1.7 kb BamHl fragment containing the HIS3 gene.
The rpblAl87:rH/S3
allele consists
of a replacement
of the 5.9 kb
BstEll-SnaBI
fragment containing the entire RPEl coding sequence
with a 1.3 kb BamHI-BssHII
fragment containing the HIS3gene.
Yeast
RNA polymerase
genetic nomenclature
is described
in Nonet et al.
(1987). YPD medium is 2% yeast extract, 1% Bacto-peptone
(Difco
Laboratories),
and 2% glucose. Synthetic complete (SC) media consists of 0.3% yeast nitrogen base without amino acids minus ammonium sulfate (Difco Laboratories),
1% ammonium
sulfate, and 4 g/l of
the following mixture: 4 g of leucine, 2 g of each of the 19 other standard amino acids, 2 g of inositol, 0.5 g of adenine, 0.5 g of uracil, and
0.2 g of p-aminobenzoic
acid. Dropout medium minus uracil (DO-URA)
consists of SC lacking uracil. Other dropout media similarly consisted
of SC lacking the appropriate
base or amino acid(s). Plates were 2%
agar. Plates containing 5-FOA were made as described by Boeke et al.
(1984).
Plasmids
YCp50 is a URA3 centromere
plasmid described by Kuo and Campbell
(1983). The plasmid pSB32, a gift from J. Trueheart (Whitehead
Institute), is a LEU2 centromere
plasmid derived from YCp50 by replacing
the URA3 gene on YCp50 with the LEUP gene. Plasmid pRP112 consists of the RPB7 gene on the 7.5 kb Aatll-Sall fragment from pRP19
(Nonet et al., 1987) inserted into the Aatll-Sal1 sites of YCpSO. Plasmid
pRP114 consists of the RP67 gene on the 7.5 kb Aatll-Sphl
fragment
from pRP19 inserted into Aatll-Sphl
of pSB32.
Construction
of Deletions
in RPBl DNA
Unidirectional
deletions starting from the SnaBl site of pRP114 were
constructed
according
to Yanisch-Perron
et al. (1985) as illustrated in
Figure 1. The LEUP centromere
shuttle plasmid pRP114, which contains the entire RPB7 gene, contains a unique exonuclease
Ill-sensitive restriction
site (SnaBl) located 20 bp after the termination
codon
of the RPB7 gene and a unique exonuclease
Ill-resistant
restriction
site (Nsil) 180 bp farther downstream.
The positions of these sites in
pRP114 provided
a means of producing
deletions
unidirectionally
through DNA encoding the carboxy-terminal
repeat with exonuclease
III. Plasmid pRP114 was digested with SnaBl and Nsil, then treated for
various times with exonuclease
Ill, digested with exonuclease
VII to
create flush ends, and ligated to BamHl linkers. The BamHl linker (lnternational Biotechnologies,
Inc.) was added to mark the deletion endpoint, and the deletion clones were transformed
into HBlOl. DNA endonuclease
digests and ligation reactions were carried out according
to Maniatis et al. (1982).
Viability
of RPBI Deletion Mutants
Ninety-one cloned plasmids with various RPB7 carboxy-terminal
deletions were assayed for their effect on cell viability by the plasmid shuffle strategy depicted in Figure 1 and described
in the text. Thirty-five
RPB7 deletion mutants were viableafter
removal of the wild-type RPB7
gene, and 56 failed to grow in the absence of the wild-type allele. Some
RNA Polymerase
915
II Heptapeptide
Repeat
of the plasmids that were isolated from E. coli failed to transform yeast
cells, even in repeated experiments.
Because the LEUP genes contained on these plasmids were fully functional, as assayed by their ability to complement
the IeuB mutant of HBlOl, and because larger deletions failed to transform
yeast more often than short deletions,
it
appears likely that some of the RPB7 deletion mutants behave as dominant lethals in yeast.
Screen for Condltional
Mutants in Deletion Library
Yeast strain 223 was transformed
directly with the library of plasmids
containing
carboxy-terminal
deletions,
and 550 Leu+ Ura- transformants that were viable at 24% were obtained. These cells were tested
for conditional
viability by replica-plating
to 12oC, 24°C 36%
and
38%. Ten of these cells failed to grow at 36% and 36% and are
temperature-sensitive
(ts). Twenty-nine
of these cells failed to grow
at 12% and are cold-sensitive
(cs). Some cells exhibited both temperature-sensitive
and cold-sensitive
phenotypes.
Although the screen
for conditional growth was carried out in strain 223, the plasmids of interest were transferred
to strain 226 for all other experiments.
The plasmids pV21, 22, and 26 were temperature-sensitive
in the 223 background, whereas they were fully viable in the strain 226, in which the
complete RPBl gene is deleted. All other plasmids produced identical
phenotypes
in the two strains.
DNA Manipulations
Restriction
analysis, gel electrophoresis,
and Southern blot analysis
were performed
essentially
as described
by Davis et al. (1960) and
Maniatis et al. (1982). Transformations
of yeast were accomplished
by
a lithium acetate method (Kuo and Campbell,
1983). Genomic yeast
DNA was isolated as described
by Boeke et al. (1985). Yeast centromere plasmids were isolated from yeast by transforming
E. coli with
yeast genomic DNA. Plasmid DNA sequencing
was done as described
by Chen and Seeburg (1965) using an oligonucleotide
with the seauence 5’CCGCAAGGAATGGTGCATCG-3’.
Western
Analysis
Crude protein extracts were prepared from yeast cells grown to ODsoo
= 0.5 as described
by Ohashi et al. (1982). Approximately
20 ug of total
protein was electrophoresed
onto a 7.5% SDS-PAGE
gel with a low
N,N’-bisacrylamide
cross-linker
concentration
as described
by Dreyfuss et al. (1984). Electrotransfer
of protein to nitrocellulose
was performed essentially
as described
by Towbin et al. (1979) except that
20 mM NasHPOdNaHsP04
(pH 8.8) was used as the transfer buffer.
Rabbit antiserum directed against the purified large RNA polymerase
II subunit was a gift of J.-M. Buhler and A. Sentenac (CNRS, Paris).
The ProtoBlot immunoscreening
system (Promega Biotec) was used
as the secondary
antibody detection system.
Acknowledgments
We are grateful to C. Scafe for valuable discussion
and criticism and
C. Carpenter for preparation
of this manuscript.
M Nonet is a National
Science Foundation
predoctoral
fellow. This work was supported
by a
grant (GM34365)
from the National Institutes of Health.
The costs of publication of this article were defrayed in part by the
payment
of page charges.
This article must therefore
be hereby
marked “advertisement”
in accordance
with 16 U.S.C. Section 1734
solely to indicate this fact.
Received
References
Allison, L. A., Moyle, M., Shales, M., and Ingles, C. J. (1985). Extensive
homology
among the largest subunits of eukaryotic
and prokaryotic
RNA polymerases.
Cell 42, 599-610.
Boeke, J., LaCroute. F., and Fink, G. (1964). Apositiveselection
for mutants lacking orotidine-5’-phosphate
decarboxylase
activity in yeast:
5-fluoro-orofic
acid resistance.
Mol. Gen. Genet. 797 345-346.
Boeke, J. D., Garfinkel,
D. J., Styles, C. A., and Fink, G. R. (1985). Ty
elements transpose
through an RNA intermediate.
Cell 40, 491-500.
J., Natsoulis,
agent in yeast molecular
genetics.
G., and Fink, G. R. (1987). IFluoro-
Meth. En-
Chen, E., and Seeburg, P (1985). Supercoil sequencing:
a fast and
simple method for sequencing
plasmid DNA. DNA 4, 165-170.
Corden, J. L., Cadena, D. L., Aheran, J. M., Jr., and Dahmus, M. E.
(1985). A unique structure
at the carboxyl
terminus
of the largest
subunit of eukaryotic
RNA polymerase
II. Proc. Natl. Acad. Sci. USA
82, 7934-7936.
Davis, R. W., Botstein,
Genetics (Cold Spring
ratory).
D., and Roth, J. R. (1960). Advanced
Bacterial
Harbor, New York: Cold Spring Harbor Labo-
Dezelee, S., Wyers, F., Sentenac, A., and Fromageot,
F! (1976). Two
forms of RNA polymerase
B in yeast: proteolytic conversion
in vitro of
enzyme B, into Brr. Eur. J. Biochem. 65, 543-552.
Dreyfuss,
G., Adam, S., and Choi, Y. (1984). Physical changes in cytoplasmic messenger
ribonucleoproteins
in cells treated with inhibitors
of mRNA transcription.
Mol. Cell. Biol. 4, 415-423.
Kuo, C-L.,
and Campbell,
J. (1963). Cloning of Saccharomyces
cerevisiae
DNA replication genes: isolation of the CDC6 gene and two
genes that compensate
for the cdc6-1 mutation. Mol. Cell. Biol. 3,
1730-1737
Lewis, M. K., and Burgess, R. R. (1982). Eukaryotic
RNA polymerases.
In The Enzymes,
Vol. XV, P Boyer, ed. (New York: Academic
Press),
pp. 109-153.
Maniatis, T, Fritsch, E. F., and Sambrook, J. (1962). MolecularCloning:
A Laboratory Manual (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory).
Mortimer, R. K., and Johnston,
J. R. (1986). Geneology
of principal
strains of the yeast genetic stock center. Genetics
713, 35-43.
Nonet, M., Scafe, C., Sexton, J., and Young, Ft. A. (1987). Eukaryotic
polymerase
conditional
mutant that rapidly ceases mRNA synthesis.
Mol. Cell. Viol. 7; 1602-1611.
Ohashi, A., Gibson, J., Gregor, I., and Schatz, G. (1962). Import of proteins into mitochondria:
the precursor
of cytochrome
c, is processed
in two steps, one of them heme-dependent.
J. Biol. Chem. 257,
13042-13047.
Ovchinnikov,
Y. A., Monastyrskaya.
G. S., Gubanov.
V. V.. Guryev,
S. O., Salomatina,
I. S., Shuvaea, T M., Lipkin, V. M., and Sverdlov,
E. D. (1982). The primary structure of E. coli RNA polymerase.
Nucleotide sequence
of the rpoC gene and amino acid sequence
of the 8’
subunit. Nucl. Acids Res. 10, 4035-4044.
Paule, M. R. (1981). Comparative
subunit composition
of the eukaryotic
nuclear RNA polymerases.
Trends Biochem. Sci. 6, 128-131.
Sentenac, A. (1965). Eukaryotic
78, 31-91.
RNA polymerases.
Crit. Rev. Biochem.
Sweetser, D., Nonet, M., and Young, R. A. (1967). Prokaryotic
and eukaryotic
RNA polymerases
have homologous
core subunits,
Proc.
Natl. Acad. Sci. USA 84, 1192-1196.
Towbin, H., Staehelin, T., and Gordon, J. (1979). Electrophoretic
transfer of proteins from polyacrylamide
gels to nitrocellulose
sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76,4350-4354.
Yanisch-Perron,
C., Vieira, J., and Messing, J. (1985). Improved Ml3
phage cloning vectors and host strains: nucleotide sequence
of the
M13mp16 and pUC19 vectors. Gene 33, 103-119.
Young, R., and Davis, R. (1983). Yeast RNA polymerase
tion with antibody probes. Science 22, 776-762.
May 29, 1967
Boeke, J. D., Trueheart,
erotic-acid
as a selective
zymol., in press.
II genes:
isola-