Tuomo Rankinen, PhD

1
Heritability of Chronic Diseases
Tuomo Rankinen, PhD Pennington Biomedical
Research Center Human Genomics Laboratory Baton
Rouge, LA
2
(No Transcript)
3
(No Transcript)
4
Familial Aggregation

• When relatives exhibit greater trait similarity
  than do pairs of unrelated individuals
• Function of shared genes, shared environments,
  and/or some combination of both
• Research strategies
• Family history approach
• Family study approach
• Case studies w/ population comparison
• Within vs. between family variation

5
Familial Risk of DiseaseLambda coefficient
?R KR / K

• KR prevalence of disease / trait among
  relatives of an affected proband
• K prevalence of disease / trait in general
  population

6
Lambda coefficient for an autosomal dominant
trait with variable levels of penetrance and
attributable fraction, for different disease
frequencies
7
Changes in MZ Twins with Overfeeding and Negative
Energy Balance
Overfeeding Negative Energy Balance
Changes in body weight (kg)
Changes in body weight (kg)
Changes in body weight (kg) Changes in body
weight (kg)
Bouchard et al. Obes Res, 1994
Bouchard et al NEJM, 1990
8
Familial aggregation in the HERITAGE Family Study
9
The HERITAGE Family StudySBP50 training
responses in families
SBP50 training response
Family Rank
10
(No Transcript)
11
Gene effects

• Polygenic many genes each of small effect,
  measurable influence on trait only in aggregate
  (additive)
• Major a single gene that has a measurable
  effect on a trait
• Oligogenic several genes, each having
  measurable effects on a trait
• Pleiotropic a gene that effects multiple traits
• Other Effects include environmental, GxG and GxE
  interactions/correlations

12
Heritability

• Proportion of phenotypic variation directly
  attributable to genetic differences among
  individuals
• VG genetic, VC common environment, VE
  unique environment
• Variance components derived from trait
  correlations among family members
• Research strategies
• Twin studies
• Family studies
• Nuclear vs. extended families (incl. Adoptions)
• Heritability estimate is always population and
  environment specific!!

VP VG VC VE
13
VP VG VC VE
G
C
E
c
h
e
P
Genetic heritability h2 VG / VP Cultural
heritability c2 VC / VP
14
Components of genetic variance
VG VA VD VI

• VA Additive genetic variance
• Phenotypic variation which is transmissible from
  parents to offspring
• VD Dominance variance
• Variation due to nonlinear interaction between
  the transmissible alleles at a single locus
• Contributes only to covariance between relatives
  who can share a genotype IBD (e.g., full sibs and
  twins)
• VI Epistatic variance
• Variation due to all nonlinear interactions among
  alleles at different loci

15
Expected covariances for pairs of relatives
16
Heritability estimatesTwin studies

• MZ twins 100 genes in common
• DZ twins 50 genes in common
• Non-twin sibs 50 genes in common
• If rMZ gt rDZ, evidence for genetic effect
• If rDZ gt rMZ, evidence for environment effect
• If rDZ gt rsibs, evidence for environment effect
• Heritability twice the difference between MZ and
  DZ intraclass correlations

h2 2rMZ rDZ
17
Genetic Liability of Type 1 Diabetes Among
Finnish TwinsHyttinen et al. Diabetes
2003521052-55

• 22,650 twin pairs (born 1958-86) from the
  national twin register
• Diabetes status from the National Hospital
  Discharge Register (1970-98), the Finnish
  Diabetes Register (1965-98), and the Central Drug
  Register (1964-)
• 228 twin pairs (247 cases) with type 1 diabetes
• 44 monozygotic pairs
• 93 same-sex dizygotic pairs
• 90 opposite-sex dizygotic pairs
• Structural equation model
• VP VA VD VC VE

18
Concordance rates for type 1 diabetesin Finnish
twins
VA 88 VE 12
From Hyttinen et al. Diabetes 2003521052-1055
19
Familial CorrelationsNuclear family
Spouse rFM Parent-Offspring rFC rMC Sibli
ng rCC
t2 (VG VC) / VP generalized
heritability twice average P-O and Sib correlation
20
Heritability estimatesNuclear families

• Only spouses provide information about
  non-genetic covariance!
• Heritability estimate includes both genetic and
  shared environmental sources of variation

21
(No Transcript)
22
Segregation Analysis

• Relies on non-normality in data
• Admixture of distributions
• Assume trait determined (in part) by single
  biallelic gene
• e.g. gene A may take one of two allelic forms, A
  and a

23
Distribution
Frequency
Standardized Trait Value
24
Segregation Model
VP Vmg (VGC) VE
Gene frequency p f(A) q f(a), Where (pq)
1 Genotypic frequency p2 f(AA) 2pq f(Aa) q2
f(aa)
? Genotypic means
Mode of transmission (d) Codominant Recessive
mAAmAa Dominant mAamaa
Mendelian transmission tAA 1 tAa ½ taa 0
Displacement t maa mAA
25
(No Transcript)
26
Research Strategies to Study the Genetic Basis
of Complex Multifactorial Traits

• Genome-wide scans
• Linkage analyses
• Microsatellites
• Family data
• Hypothesis generation

• Candidate genes
• Association studies
• Biallelic markers (SNPs)
• Independent subjects or family data
• Hypothesis testing

27
Positional Cloning in Post-Human Genome Era
Modified from Boerwinkle et al. Circulation 2000
28
Genetic Linkage

• Location of genetic loci sufficiently close
  together on a chromosome that they do not
  segregate independently
• No or very little recombination between the loci
  (? ltlt 0.5)
• Property of the relative position of loci, not
  their alleles
• The observed co-segregation within a given family
  can involve any allele at the marker locus

29
Genetic Linkage

• Crossing-over a reciprocal event in a pair of
  homologous chromosomes, takes place when the
  chromosomes have already duplicated at the
  beginning of the first meiotic division
• Recombination fraction (?) the probability that
  an odd number of crossover events will take place
  between two loci the proportion of recombinant
  gametes

30
From www.accessexcellence.org
31
Modified from Ott 1999
32
Types of Linkage Analysis

• Model-based linkage analysis
• model for the trait segregation required
• Mendelian segregation
• Recessive / Dominant
• Gene frequency
• Penetrance
• Model-free linkage analysis
• genetic model for the trait not needed
• allele sharing I.B.D. between relative pairs
• excess allele sharing between affected sibs
• phenotypic vs. genotypic resemblance between
  relative pairs

33
Proportion of Alleles Shared IBD

• The genetic similarilty of the sib pair is
  measured as the proportion of alleles they share
  identical-by-descent
• fj1 and fj2 probabilities that the jth sib
  pair share 1 and 2 alleles IBD, respectively

34
Allele Sharing Identity-by-Descent (IBD)
1 / 2
3 / 4
Parental genotypes
Number of alleles shared IBD
Proportion of alleles shared IBD
0
0
1
0.5
2
1
35
Model-Free Linkage for Quantitative
Traits(original Haseman-Elston)
36
Dilemma
Gene-Environment Interactions?
37
ADRB2 gene polymorphism, BMI and physical activity
Body mass index
Waist circumference
Meirhaeghe et al. Lancet 1999
38
Physical activity, breast cancer and BRCA1 and
BRCA2 mutations
King et al. Science 2003
39
Coronary heart disease risk and LIPC genotype
Hokanson et al. Am J Epidemiol 2003
40
Physical activity, coronary heart disease risk
and LIPC genotype
Hokanson et al. Am J Epidemiol 2003
41
Are chronic diseases / complex traits due to
genetic or environmental factors?
YES!
42
Recommended Reading

• Khoury MJ, Beaty TR, Cohen BH. Fundamentals of
  Genetic Epidemiology. Oxford University Press,
  New York 1993
• Rao DC, Province MA (eds.) Genetic Dissection of
  Complex Traits. Academic Press (Elsevier Science)
  2001
• Ott J. Analysis of Human Genetic Linkage. Johns
  Hopkins University Press 1999