Developmental Plasticity – Sensitive Periods and Risk
of Obesity
Karen A Lillycrop
Institute of Developmental Sciences
Centre for Biological Sciences
University of Southampton
Genotype
E
P
I
G
E
N
O
M
E
Environment
Environment
Environment
Chronic disease susceptibility
(CVD, obesity, type II diabetes, cancer)
2
Early life environment can alter the epigenome
Establishment of
imprinting
Gametogenesis
Cell lineage specific
methylation
Erasure of parental
methylation
conception
birth
weaning
Epigenome
Prenatal
Nutrition/
Body
composition
Maternal stress
/childhood adversity
Endocrine disruptors
/pollution
Neonatal
Pubertal
Growth
maturation
aging
What is the evidence that early life environment can induce epigenetic
and phenotypic changes?
High fat
High fat
Protein restriction
Biochemical Journal 2010 427, 333-347 - Matthew J. Warner and Susan E. Ozanne
Maternal protein restriction induces the hypomethylation of the PPARα
Promoter in the liver of the offspring
Hypertension
Dyslipideamia
Insulin resistance
Maternal protein restriction
110
100
*
90
80
70
Control
PR
Maternal dietary group
Lillycrop et al., 2005,.2007
Β oxidation
PN34
700
1400
*
1200
1000
800
600
400
200
 hyroxybutyrate concentration (mol/l)
120
PPAR expression
PN34
mRNA concentration compared
to control (%)
DNA methylation compared to
control (%)
Hepatic PPAR methylation
PN34
600
500
400
300
200
100
0
0
con
Control
PR
PR
maternal dietary group
Maternal dietary group
5
150
100
*
*
**
*
**
50
E18
PN34
PN84
PR
on
C
PR
on
C
PR
on
0
C
% Methylation ( mean ±SD)
Maternal protein restriction induces a change in PPAR α methylation in
fetal liver
Maternal high fat feeding alters the methylation of the FADS2 gene
Female Liver 20:4n-6
% Total Fatty acids
40
7% Butter
21% Butter
7% Fish Oil
21% Fish oil
30
20
10
1
0
Expression (normalised Ct)
% methylation
2
40
20
sh
%
21
r2 = 0.56, P<0.0001
1
0
l
oi
il
h
Fi
s
21
%
Fi
7%
B
%
21
sh
ut
O
te
r
te
ut
B
7%
Fi
21
%
r
l
sh
O
7%
Fi
s
h
ut
B
%
21
oi
il
r
te
r
te
ut
B
3
7% Butter
21% Butter
2
7% Fish Oil
21% Fish oil
0
7%
Log DD Ct (mean SD)
CpG -394
7%80Butter
21% Butter
7%60Fish Oil
21% Fish oil
3
Fi
sh
Fi
7%
CpG -394
Female Hepatic FADS2 expression
4
oi
l
il
O
te
ut
%
21
7%
B
B
ut
te
r
r
0
40
50
60
70
CpG methylation
80
7
7
Over –feeding in early postnatal life leads to obesity and the hypermethylation of
the POMC promoter
Rat pups raised in small litters
(SL)
POMC methylation in the hypothalamus (PN21)
Rapid fat accumulation
Hyperglyceamia
Obesity
NL SL
Plagemann et al., 2009
Nutrition in early life can induce long term changes in DNA methylation
conception
birth
weaning
Growth
maturation
aging
sperm
GR, PPARa
IL-13R,
PPARa
High
Fat
Protein
restriction
Paternal diet
GR, PPAR, PEPCK
HNF4,ATR1
FADS2
Global
High
Protein
restriction
fat
restriction
Maternal diet
Epigenome
POMC
Long term changes
in gene expression
& metabolism
Over feeding
Altered disease risk
The Dutch Hunger Winter
A period of severe food shortage in the Netherlands in
1944.
Energy intakes dropped from 1800 to between 400 and
800 kcal per day (equivalent 100 - 200g pasta).
Individuals born to women exposed to famine during pregnancy
have an increased risk of CVD, T2D and obesity in later life
Alterations in DNA methylation has been shown in individuals from mothers who
Were exposed to famine during pregnancy compared to their non exposed siblings
IL10 GNASAS
IGF2
Lep
ABCA1 MEG3
Tobi et al., 2009
Developmental plasticity – Same genome, different phenotype
Early embryo
Adapted
Normal
Future Actual Postnatal
Environment
e.g. High Calorie Diet
Developmental Plastic
Stage
“prenatal environment”
Environmental Cue
e.g.
Poor Nutrition
Reduce energy demands
Increase capacity for fat storage
Less investment in muscle mass
High Disease Risk
MISMATCH
Response
Alternative
Developmental Path
Future Actual Postnatal
Environment
e.g. Poor Nutrition
Low Disease Risk
MATCH
11
Waddington 1942
birth
conception
GR,
PPARa
Growth
maturation
Epigenome
GR, PPAR, PEPCK
P16,
p21,HNF4,ATR1
Protein
restriction
weaning
FADS2
p16
Global
restriction
High
fat
POMC
Over feeding
FADS2
aging
Long term changes
in gene expression
& metabolism
Folic acid
Altered disease risk
Detect changes in methylation
Predict metabolic capacity & future
disease risk
Epigenetic marks as biomarkers?
In humans limited tissue availability?
Available tissues: Umbilical cord
Cord blood
Placenta
buccal cells
Blood
Methylation at the retinoid X receptor α (RXRA) promoter
at birth vs child’s fat mass
SWS children age 6 yrs
PAH children age 9 yrs
r= 0.32 P=0.009
n=64
r= 0..20 P=0.002
n=239
8
7
Child's fat mass
age and sex adjusted (kg)
Child's fat mass
age and sex adjusted (kg)
6
6
5
4
5
4
3
0
Umbilical cord RXRA methylation
Godfrey et al., 2011
3
8
0
8
-8
0
0
-8
-6
0
0
-6
-4
0
0
-4
Umbilical cord RXRA methylation
Values are means + SEM
Genome wide methylation array
Methyl DNA capture
Genomic DNA from SWS subjects sonicated and precipitated using
His-tagged MBD2b to enrich for methylated sequences
Hybridised to Human Promoter Array
60 mer probes spaced across promoter regions of 17,000 best
characterised transcripts covering -8kb to +2kb downstream of TSS
DATA analysis
Methylation levels of 100nt regions were estimated using the
Bayesian algorithm BATMAN
Differentially methylated Regions (DMRs –
100 nt) identified (Fishers Exact test) associated with
% fat mass age 6 yrs
93 DMRs identified, associated with changes in %
fat mass age 6 years
Confidential
Among the identified DMRs,
the top pathway enriched was DNA replication & repair
This pathway includes a DMR linked with cyclin-dependent kinase inhibitor 2A
(CDKN2A)
UCN
Igm
CD40
IDH1
MapK
GATA6
HLA-C
ENDRB
AKAP8
Akt
SAA1
CDKN2A
Jnk
P13K complex
CCND1
PCPB2
capsase
Hintone H3
RNA pol II
NFKb complex
E2f
cyclin A
APP
SIAH1
OAS1
SP1
26s proteosome
PCNA
Alpha tubulin
NGFR
ID1
FEN1
GAPDH
VMP1
GTPBP4
CDKN2A
●
Encodes for 2 cell cycle inhibitors p14ARF and P16INK4a , which play roles
in cell proliferation, differentiation and senescence
Lower umbilical cord CDKN2A methylation is associated with higher
child’s % fat mass (SWS children)
At 6 years
At 4 years
CpG1
CpG3
P=0.003
n=208
P=0.003
n-204
P=0.007
n=221
P=0.007
n=215
CpG4
CpG6
P=0.005
n=231
P=0.015
n=230
P=0.007
n=247
P=0.04
n=244
Lower CDKN2A methylation at birth associated with greater
ponderal index at age 18 months in 161 GUSTO infants,
reflecting higher adiposity
P=0.015
Perinatal CDKN2A
methylation in an
independent cohort from
Singapore with
adiposity data at age 18
months
Lillycrop & Godfrey et al, unpublished
Epigenetic biomarkers
Temporal stability of DNA methylation
Recent data suggests that DNA methyation can be dynamically
regulated
Methylation stability over time
The Early Bird Study
Longitudinal non-intervention study of healthy children between ages 5 to
16.
Annual measurements include physical activity, body composition and
insulin sensitivity and annual whole blood samples taken
Sirt1/PGC-1a (PPAR gamma co-activator (PGC))-1α
Sirt1
(energy sensor)
HNF4a
Energy
homeostasis
Glucose
metabolism
PGC1α
Muscle fibertype switching
Insulin
secretion
Mitochondrial
biogenesis
Adipogenesis
Methylation stability over time
PGC1a
Sirt1
HNF4a
80
CpG
CpG
-8 4 1
100
-9 4
-1 0 3
-1 1 7
-8 1 6
-5 1 5
-7 8 3
-6 1 7
40
-6 5 2
M e t h y la t i o n ( % )
M e t h y la t i o n ( % )
60
-1 2 1
90
-5 2 1
20
80
0
5 - 7
8 - 9
10
11
12
A g e (y e a rs )

13
14
-7 9
5 - 7
8 - 9
10
11
12
A g e (y e a rs )
Methylation not affected by the proportion of neutrophils or
lymphocytes (all P > 0.1; neutrophils r = -0.002 to 0.34,
lymphocytes r=0.001 to 0.23).
13
14
Methylation of CpG loci in PGC-1α was associated with adiposity
at 14 years
• Significant association
between the methylation of 4
of the 7 CpG’s at 5-7 years
and % body fat from 9-14
years (P<0.05).
For each 10% difference in
methylation at 5-7 years, % body
fat differed by 6.3% to 12.5%
(P<0.03) at age 14.

Does differential methylation of the CpG loci within PGC1a have
functional significance?
Methylation increases binding of the
pro-adipogenic HOXB9/PBX1 heterodimer.
Early life environment is critical for future health
In vitro
Culture?
Maternal/paternal
Nutrition
Maternal behavior
Nutrition
1.Early life environment is an important determinant of later obesity
risk
2.Early life environment can alter the epigenome and these epigenetic
changes induced in early life make a significant contribution to later
phenotype and disease susceptibility.
3.Detection of such altered epigenetic marks in early life may allow
the identification of individuals at increased risk of metabolic diease
in later life.
Pathways to obesity
Epigenetic signatures?
CDKN2A?
Are there modifiable
epigenetic marks?
mis-match
pathway?
Rapid catch up
growth
Epigenetic signatures?
PGC1a/Sirt1
obesity
?
Are there modifiable
epigenetic marks?
Greater fat mass,
less lean mass and
lower IQ at age 4
Poor educational attainment
Poor diet
Take less exercise
Obese
Ill-prepared for pregnancy.
Maternal
drivers?
Undernourished
Dietary restriction
IUGR
Minimal changes in diet and
health behaviours
Maternal
drivers?
Over nutrition
Maternal high energy diet
Maternal obesity
GDM
Southampton Developmental Epigenetics group
Graham Burdge
Keith Godfrey
Mark Hanson
New Zealand
Peter Gluckman
Teejal Bhatt
Nicki Irvine
Elisoe Cook
Leanie Grenfell
Becki Clarke
Paula Costello
Mark Burton
Danya Agfa
Jordan Price
Rob Murray
Emma Garratt
Sam Hoile
Charlie Simmons
Singapore
Joanna Holbrook
Walter Strunkel
Australia
RaeChi Huang
Leiden
Bas Zwaan
Plymouth
Terry Wilkin
Jo Hosking
Jon Pinkey