"Genetics, diabetes and obesity
–
where
are
we?"
Torben Hansen Professor, MD, PhD
Hagedorn
Research Institute
Gentofte, Denmark
Winter School
2010‐2011
Individual
Health
& Nutrition
9 november 2010
Homo stupidus
Homo sapiens
1.
Genetics of increased BMI/obesity and
common forms of diabetes
2.
Missing heritability
3.
Our other genome
Features of obesity
Hypertension
High TG Low HDL
Insulin resistance
InflammationPro-coagulation
Type 2 diabetesand
vasculardisease
Triggers: OvereatingPhysical inactivity
What
about
genetics
of
obesity?What
about
genetics
of
obesity?
Photograph by Karen Kasmauski
©
National Geographic
2007
Studies Studies ofof
thethe
geneticsgenetics
ofof
complexcomplex
disordersdisorders: : thethe
developmentdevelopment ofof
technologytechnology
drives drives thethe
progressprogress
single variant
(100 SNPs; 103
genotypes)
detailed study of individual genes(102 SNPs; 105+
genotypes)
regional studies (104 SNPs; 108
genotypes )
genome‐wide association
(106
SNPs; 1010 genotypes)20052005
GenomeGenome‐‐widewide
association in 2007association in 2007
Studies Studies ofof
thethe
geneticsgenetics
ofof
complexcomplex
disordersdisorders: : thethe
developmentdevelopment ofof
technologytechnology
drives drives thethe
progressprogress
single variant
(100 SNPs; 103
genotypes)
detailed study of individual genes(102 SNPs; 105+
genotypes)
regional studies (104 SNPs; 108
genotypes )
genome‐wide association
(106
SNPs; 1010 genotypes)20052005
Whole exome
and whole
genome sequencing 2010
GenomeGenome‐‐widewide
association (GWA) in association (GWA) in shortshort
Type for ~/>500,000 SNPs
Obtain information about strength of associationgenome wide
(within limits of sample size, allele frequency, LD etc)
Validation in independent samples of what looks interesting
followed by combined analyses
By courtesy of http://www.broad.mit.edu/diabetes/
ExampleExample
ofof
a GWA a GWA withwith
389,000 389,000 SNPsSNPs
successfullysuccessfully genotypedgenotyped
Eighteen risk loci associate with obesity/increased BMI with genome-wide significance
Each of them are common but only increase risk of obesity with 8-33%
Odds ratio
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.00
Positional cloningHypothesis-free approach
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.00
Odds ratio
2007
MC4RPCSK1
2008
FTO
2009
TMEM18 SEC16BGNPDA2 KCTD15PTERNGR1
BAT2
BDNFFAIM2MAF ETV5
SH2B1NCP1
PRLMTCH2
Most of
obesity
risk
loci
are
related
to genes in the
hypothalamus
-> disturbed appetite regulation
MC4R
PTER NCR3/AIF1/ BAT2
PRLPCSK1
TMEM18
NEGR1SH2B1
MAF
FAIM2/BCDIN3D MTCH2
BDNF KCTD15
NPC1
FTO
Explain collectively about 1 % of the genetic component of obesity
Speliotes-EK et al. Nature Genetics. Online Oct. 10, 2010
Combined association analyses of 249,796 individuals show additional eighteen new loci associated with body mass index
Meta-analysis
of
GWAS identifies
13 new loci
associated
with
waist-hip
ratio
Nature GeneticsOnline
October
10, 2010
The clinician’s view!
Cumulative effect and predictive value of 32 loci
•
ARIC study, 8,120 individuals, population-based, whites
250
2627
2829
Mean
BM
I (k
g/
m2)
500
1000
1500
Nu
mb
er
of
ind
ivid
uals
(N
)
≤2122-23
24-2526-27
28-2930-31
32-3334-35
36-37≥38
Number of weighted risk alleles
0.17 k
g/m2 per
alle
le
435-551g p
er a
llele
2.71 kg/m2
6.9 - 8.8 kg
0
0
0.25
0.50
0.75
1.00
Sen
siti
vity
0.25 0.50 0.75 1.00
1-Specificity
AUC32 SNPs = 0.574
By By courtesycourtesy ofof Dr. Ruth Dr. Ruth LoosLoos
Speliotes et al., Nature Genetics. Online 10 October, 2010
GeneticsGenetics
ofof
obesityobesity
Allele frequency
(%)
Effect size
(OR)
0.1%
High
Low
1.1
1.5
3.0
>50
Modest
Inter-
mediate
Very rare
Rare Low frequency
Common0.5% 5%
Monogenic mutation
LEPLEPRMC4RPOMCPCSK1SIM1BDNFNTRK2
Deletion16p11.2, SH2B1
ABCB5EDIL3FOXP2NEGR1S1PR5ZPLD1Duplication
ARL15KIF2B
FANCL
LRRN6C
BDNF
MAP2K5
GNPDA2FLJ35779
GPRC5B
FTO
LRP1B
MTCH2
NEGR1 NPC1MTIF3
CTNNBL1
ETV5/SFRS10
MAFKCTD15
FAIM2
MC4RCADM2
PCSK1
TMEM160
PTBP2
PRL
PFKPPRKD1
NUDT3
NRXN3RPL27A
QPCTL
SH2B1
SEC16B
PTER RBJ
TMEM18
SLC39A8
TNNI3K
ZNF608
Biologic candidate approachPositional cloningHypothesis-free approach
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.00
SLC30A8CDKAL1
IGF2BP2
CDKN2AFTO
HHEX
WFS1TCF2
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.00
Risk of diabetes (Odds ratio)
37 risk loci associate with T2D with genome-wide significance
Each of them are common but only increase risk of type 2 diabetes with 5-35%Risk of diabetes (Odds ratio)
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
TCF7L2 KCNJ11PPARG
MTNR1B
THADANOTCH2
CAMK1DADAM30
JAZF1ADAMTS9
TSPAN8
KCNQ1
2010
DUSP9CENTD2
TP53INP1CHCHD9HNF1A
HMGA2ZFAND6
PRC1KLF14
PROX1
ZBED3
DGKBGCK
GCKR
ADCY5
BCL11A
KCNQ1IRS1
FTO
PPARGIRS1
ADAMTS9GCKRKLF14 CDKAL1
C2CD4A
Estimated
functions
of
variants at type 2 diabetes risk
loci
TCF7L2WFS1
KCNQ1
Undetermined:
Fasting
glucose
meta- analysis
discovers
G6PC2
GCK
MTNR1B
GCKR DGKBPROX1
ADCY5SLC2A2 GLIS3
ADRA2AMADD
FADS1
CRY2
FAM148BSLC30A8 TCF7L2
Five known FGP loci
Dupuis et al, Nat Genet online Jan 17th 2010
Two established T2D lociNine novel loci
slide 19 The Steno Advanced Update Course
TheThe
combinedcombined
impactimpact
ofof
16 16 riskrisk
variants variants onon
fastingfasting glucoseglucose
in a in a nonnon‐‐diabeticdiabetic
populationpopulation
number
of
risk
alleles
Fast
ing
gluc
ose
(mm
ol/l)
<12 12 13 14 15 16 17 18 19 20 21 22 >22
5.0
6.0
Num
bero
find
ivid
uals
020
060
080
0 Increase per allele: 0.04 mmol/l (p=10-17)
5.5400
slide 20
CommonCommon
variants variants explainexplain
a a minorminor
fractionfraction
ofof
thethe
variancevariance
in in metabolicmetabolic
phenotypesphenotypes
in in thethe
general populationgeneral population
Of total: 4‐6%
Of genetic: 8‐12%
Fasting
plasma glucoseHeritability
30‐40%
slide 21
GeneticsGenetics
ofof
nonnon‐‐autoimmuneautoimmune
diabetesdiabetes
PPARG
KCNJ11
PNDM genesKCNJ11ABCC8
INS
Other rare syndromes
MODY genesHNF1AHNF4A
GCKIPF1
HNF1BNEUROD1
CEL
TCF7L2
WFS1
IGF2BP2 CDKAL1
CDKN2A
FTO
SLC30A8
HHEX
HNF1B CDC123 JAZF1
TSPAN8
ADAMTS9
NOTCH2
THADA
KCNQ1
MTNR1B
IRS1
GCK
GCKR DGKB
PROX1 ADCY5
Allele frequency
(%)
Effect size
(OR)
0.1%
High
Low
1.1
1.5
3.0
>50
Modest
Inter-
mediate
Very rare
Rare Low frequency
Common0.5% 5%
and many more
Rare, penetrant
mutationsGood for individual prediction in rare cases
Too rare for population effect
Common SNPs
with low penetranceEffects too small for individual prediction
ExplosionExplosion
in in thethe
numbernumber
ofof
validatedvalidated
commoncommon variants for all variants for all diabetesdiabetes‐‐relatedrelated
metabolicmetabolic
traitstraits
Fasting
plasma glucose ~16 loci
Circulatinglipid
levels ~36 loci
Cardiovasculardisease ~18 loci
BMI orobesity ~18+ loci
Hypertensionor
BP ~14 loci
Unknown
causal
mutation and gene
Type 2 diabetes 37+ loci
Common
variants with
low
individual
impact
Identified
variants explain
5‐10 %
of
the
genetic
contribution
Collectively
the
variants have no
clinical
utility
for disease
discrimination
or
prediction
More More thanthan
90% 90% ofof
thethe
heritabilityheritability
ofof thesethese
disordersdisorders
is not is not explainedexplained
Additional common risk alleles?Additional common risk alleles?
GeneGene--environment interactions?environment interactions?
GeneGene--gene interactions?gene interactions?
Copy number variations?Copy number variations?
EpigeneticsEpigenetics??
TheThe ‘‘missing missing heritabilityheritability’’ ofof type 2 diabetes and type 2 diabetes and obesityobesity
More More thanthan
90% 90% ofof
thethe
heritabilityheritability
ofof thesethese
disordersdisorders
is not is not explainedexplained
Additional common risk alleles?Additional common risk alleles?
GeneGene--environment interactions?environment interactions?
GeneGene--gene interactions?gene interactions?
Copy number variations?Copy number variations?
EpigeneticsEpigenetics??
TheThe ‘‘missing missing heritabilityheritability’’ ofof type 2 diabetes and type 2 diabetes and obesityobesity
By By courtesycourtesy
ofof
Dr. Mark Dr. Mark McCarthyMcCarthy
TheThe
association association ofof
FTO FTO withwith
obesityobesity‐‐relatedrelated quantitativequantitative
traitstraits
in Inter99in Inter99
•
FTO associates
with
increased
BMI, body
weight
and waist circumference
•
No
association with
serum lipids
Andreasen CH, Diabetes 2008;57:95-101
The impact of the FTO rs9939609 genotype on BMI- levels is highly influenced by habitual physicial activity: Physical inactivity was associated with an increase of
1.95 BMI units in AA-allele carriers A population-based study of 5,722 middle-aged Danes
Physical
activity
was
self-reported
by questionnaire
Number
of
subjects
(TT/TA/AA)
Pint = 0.007
Andreasen CH, Diabetes 2008;57:95-101
APOLIPOPROTEIN A-V; APOA5
-1131T>C
• Carrier frequency
11 % among Caucasians
Association with fasting serum triglycerides in the general population
of 5,873 middle-aged Danes
T/T T/C C/C p
n(men/women) 5,152(2,563/2,589) 698(356/342) 23(10/13)
Age (years) 46 ± 8 46 ± 8 44 ± 7
BMI (kg/m2) 26.2 ± 4.5 26.4 ± 4.8 24.3 ± 4.6 0.7
Waist-to-hip ratio 0.86 ± 0.09 0.86 ± 0.09 0.84 ± 0.08 0.9
Fasting serum lipids (mmol/l)
Triglyceride 1.3 ± 1.1 1.7 ± 2.7 1.8 ± 0.9 110-15
Mean ± SD. Adjustment for age, sex, and BMI (where appropriate). Calculated by GLM assuming additivity.
The effect of APOA5 -1131T>C on fasting serum triglycerides is modulated by the level of hyperglycaemia
in studies of 5,873 middle-aged Danes
Fasting serum triglyceride (mmol/l)
Wild type TC + CC
p = 510-9
0
1
2
3
4 NGTIFGIGTScreen-detected T2DM
3880 438 609 221 561 57 73 28
The impact of APOA5 -1131T>C on fasting serum triglycerides is modulated by the degree of insulin resistance (Homa-IR)
in studies of 5,873 middle-aged Danes
Fasting serum triglyceride (mmol/l)
Wild type TC + CC
p = 510-9
0
1
2 < 44-55-66-77-88-9.59.5-11.511.5-14.514.5-20> 20
HOMA IR:equally sized subgroups
The impact of APOA5 -1131T>C on fasting serum triglyceride is modulated by smoking
Fasting serum triglyceride (mmol/l)
Wild type TC + CC
p = 0.0003
0
1
2
2.5
Daily and occasional smokersNon-smokers and previous smokers
0.5
1.5
The impact of APOA5 -1131T>C on fasting serum triglyceride is modulated by alcohol intake
Wild type TC + CC
Fasting serum triglyceride (mmol/l)
p = 0.005
0
4
6
None or moderate alcohol intakeHigh alcohol intake
2
Public health implications: Gene-Lifestyle and Gene-Metabolism
Interactions
Specific health behaviour recommendations seem appropriate based upon knowledge of the
FTO, LIPC and APOA5 genotypes
More More thanthan
90% 90% ofof
thethe
heritabilityheritability
ofof thesethese
disordersdisorders
is not is not explainedexplained
Additional common risk alleles?Additional common risk alleles?
GeneGene--environment interactions?environment interactions?
GeneGene--gene interactions?gene interactions?
Copy number variations?Copy number variations?
EpigeneticsEpigenetics??
TheThe ‘‘missing missing heritabilityheritability’’ ofof type 2 diabetes and type 2 diabetes and obesityobesity
TheThe
‘‘rare variant rare variant hypothesishypothesis’’
A A significantsignificant
proportion proportion ofof
thethe inheritanceinheritance
ofof
complexcomplex
disordersdisorders
is is
due to due to thethe
cumulativecumulative
load load ofof
rare rare variantsvariants
Rare Rare riskrisk allelesalleles ––nextnext frontierfrontier
Rare risk Rare risk allesalles??
Allele frequency
(%)
Effect size
(OR)
0.1%
High
Low
1.1
1.5
3.0
>50
Modest
Inter-
mediate
Very rare
Rare Low frequency
Common0.5% 5%
Monogenic mutation
LEPLEPRMC4RPOMCPCSK1SIM1BDNFNTRK2
Deletion16p11.2, SH2B1
ABCB5EDIL3FOXP2NEGR1S1PR5ZPLD1Duplication
ARL15KIF2B
FANCL
LRRN6C
BDNF
MAP2K5
GNPDA2FLJ35779
GPRC5B
FTO
LRP1B
MTCH2
NEGR1 NPC1MTIF3
CTNNBL1
ETV5/SFRS10
MAFKCTD15
FAIM2
MC4RCADM2
PCSK1
TMEM160
PTBP2
PRL
PFKPPRKD1
NUDT3
NRXN3RPL27A
QPCTL
SH2B1
SEC16B
PTER RBJ
TMEM18
SLC39A8
TNNI3K
ZNF608
??
WhereWhere
is is thethe
missing missing heretabilityheretability??
The Lundbeck
Foundation Initiative in genomics of type 2 diabetes and obesity
www.lucamp.org
•
Sequencing of the coding part of the human genome in 2,000 Danes
•
Gene-wise comparisons of ~
20,000
likely deleterious gene variants in 17,000 Danes
Exome
capturingGenome elements # Of elements Total length
Promoter 1,917 490,995
5-UTR 17,046 672,142
CDS-exons 209,095 32,584,741
3-UTR 6,524 401,002
Intron 55,372 4,349,839
•The “exome”
is defined to be the designed target region which would be hybridized with the NimbleGen
2.1M-
probe sequence- capturing array.
•
18,654 genes, ~34Mbp length in total
ProductionProduction
ofof
hugehuge
amountsamounts
ofof
data data throughthrough highhigh‐‐throughputthroughput
DNA DNA sequencingsequencing
in in ChinaChina
IndividualIndividual
sequencingsequencing
ofof
thethe
codingcoding
region region ofof
200 initial Danish 200 initial Danish human human genomesgenomes
coveringcovering
~20,000 genes ~20,000 genes discovereddiscovered
~53,000 ~53,000
novelnovel
variants (= not in variants (= not in availableavailable
databases)databases)
Li et al., Nat Genet, online 3 October 2010
b
The
bacteria
in our
distal
intestines
function
as
a metabolic organ
Nature 444, 21 December 2006
Is an imbalance in gut bacteria in part to blamefor metaboliccomplications
Is an imbalance in gut bacteria in part to blamefor metaboliccomplications
Dethlefsen L. et al. Nature 449: 811-818,2007
Human colon
bacteria≈
1.5 kg microbes
Bacteriodetes and Firmicutes account
for > 90%
Functions:
TrophicControl of epithelial cell proliferation
and differentiation
ProtectiveProtection against pathogens
MetabolicFermentation of non-digestible
polysaccharides and storage of the extracted energy in host fat depots. Contributes with about 10% of total daily energy infusion
Who
is who? The
microflora
of
human body
cavities
FirmicutesBacteriodetes
Obese mice have more Firmicutes
Obese mice have 50 % higher representation of gut Firmicutes division and proportionally less of the Bacteroidetes division than lean mice
This
trait
was
transmissible: When
germ-free
mice
were
given an 'obese
gut microflora' it
resulted
in an increase
in total body
fat compared
with
colonization with
a 'lean
gut microflora’
The Firmicutes have a higher capacity for fermentation of
non-digestible polysaccharides than
Bacteriodetes
Ley RE et al. PNAS 102:11070-5,2005 Turnbaugh
PJ et al., Nature 444:1027-31,2006
Features of obesity
Hypertension
High TG Low HDL
Insulin resistance
InflammationPro-coagulation
Type 2 diabetes,vasculardisease, cancer
Triggers: Genetic risk factorsOvereatingPhysical inactivityThe gut microbiome?
Massively parallel shotgun sequencing of genes from distal
gut bacteria of 86 Danes
and
38 from Spain
a culture-independent approachto identify the gut bacterial genes
Identification
of
3.3 million genes from distal
gut of 86 Danes
and 38 from Spain
•
Human gut gene set is at least
150 times larger
than
the
gene set of
the
human genome
•
> 99% of
gut genes are
bacterial
•
Each
individual
harbors
~ 160 gut bacterial
species with
~
530,000 bacterial
genes
•
64 out of
160 gut bacterial
species are
shared
by >90% of
individuals
•
Bacteriodetes and Firmicutes are
the
most abundant
Nature 464:59-65, 2010
Which
molecular
mediators are
involved
in the
metabolic
dysfunctions?
……bacterial
lipopolysaccharides?……bacterial
short
chain
fatty
acids?
…....bacterial
amino
acids? ……other
mediators
What
are
the
primary drivers?
….host genomics?
…..unhealthy
dietary
habits with preference
of
fat food?
Gut microbiome
based
interventions: perspectives
•
Coloscopic
replacement
with
a mixture of
cultured
or
fresh
’healthy’ bacteria
species?
•
Oral prebiotics?
•
Oral probiotics?
•
Irradications?
Individualizing
the
gut treatmentdependent on
host genotype profiles?
Gut Gut microbiomemicrobiome (genetic
diversity)Human Human genomegenome(genetic
variations)
Bacterial components and metabolites
Bacterial components and metabolites
DietDiet HealthHealth DiseaseDisease
InteractionInteraction
Is the presence of the Firmicutes imbalance determined by the individual make-up of the host nuclear genome?
Is the presence of the Firmicutes imbalance determined by the individual make-up of the host nuclear genome?
Next
steps
Oluf
Pedersen
Niels
Grarup
Thomas Sparsø
Michel Kristensen
Grete Lademann
Arne Lykke Nielsen Toby
Brunt
Thomas Sparsø
Anders Albrechtsen
Camilla Andreasen
Anette
Prior Gjesing
Trine Welløv
Boesgaard
Trine Nielsen Ehm
Anderson Kristoffer
Burgdorf
Karina Banasik
Annemette
Forman
Marianne Stendahl
Lise
Wantzin
Debbie Nadelman
Tina Lorentzen
Marie Vestmar
Malene Hornbak
Marie Harder
Nicolaj Krarup
Frida Emanuelsson
Lise Højsgaard Olsen Lise
Wegner
Signe Torekov
Lise
Wegner
Christian Rose
Dorit
Zobel
Kirstine
Stender-Petersen Johan Holmkvist
Lesli
Larsen
Meena
Hussain
Stine Anthonsen
Mette
Andersen Rahul
Kapur
Metagenomics of the Human Intestinal Tract
Institut National de la Recherche Agronomique, France
Hagedorn Research Institute, Denmark
University of Copenhagen, Denmark
Technical University of Denmark
Beijing Genomics Institute, China
Wellcome Trust Sanger Institute, UK
Danone Research, France
European Molecular Biology Laboratory, Germany
Genoscope – Centre national de séquençage, France
INRA Transfert S.A, France
Hospital Universitari Vall d´Hebron, Spain
Instituto Europeo di Oncologia, Italy
Pharma S.A, Spain
Wageningen Universitit, Netherlands
Thorkild I A SørensenThue SchwartzJens Juul Holst,Torben Jørgensen,Kim Overvad,Anne Tjønneland,Torsten LauritzenAnelli SandbaekSten Madsbad,Bernard ThorensSøren Brunak
Ulf Smith,C.Ronald Kahn,Marku Laakso,Hans Härring,Henrik Mortensen,Valgerdur Steinthorsdottir,Unnur Thorsteinsdottir,Kári Stefánsson
Hans-Henrik Parving,Peter Rossing,Lise Tarnow,V Mohan,Jan Lebl,Danniel Witte,Ole Schmitz,Ivan Brandslund,Cramer K Christensen
Hennning Beck-Nielsen,Kurt Højlund,Søren Urhammer,Flemming Pociot,Jørn Nerup,Thomas Mandrup-Poulsen,Toni Maasen,Leen M’T Hart,Fabrizio Barbetti,Giorgio Sesti,Matha Hribal
Juleen Zierath,Peter Damm,Lars Hansen,Sten Lund,Søren Lund,Katrine Almind,Henrik Vestergaard,Jesper Clausen,Shiro Maeda,Yusuke Nakamura,Hans Eiberg,Søren Brunak,Aage Vølund,Niels Møller,Hui Liwen
Jun Wang,Li Yuingrui,Tian Saturn,Jian Wang,Huanming Yang,Rasmus Nielsen,Robert Sladek,Johan Rung,David Meyre,Nabila Bouatia-Naji,Michael Benzinou,Unoki HiroyukiClifton BogardusLeslie Baier,M. Alan Permutt
Philippe Froguel,Mark McCarthy,Andrew Hattersley,Tim Frayling,Leif Groop,José Florez,Pål Njølstad,Arne Astrup,Johan Auwerx
Allan Vaag,Pernille Poulsen,Jørgen Wojtaszewski,Rasmus Riebel-Madsen,Kirstine Færch,Kasper Pilgaard