Genetic Approaches to Thinking, Moving and Feeling

Richard B. Lipton, M.DProfessor and Vice Chair, Neurology

Professor of Epidemiology and Population Health

Albert Einstein College of Medicine

Why consider genetic approaches?

• Genetic factors play a crucial role in disease that affect thinking, feeling and moving

• Genetic factors are relevant to potential shared mechanisms: vascular disease, inflammation, recovery from injury

• Method for parsing the heterogeneity• Environmental risk factors interact• Identify molecular targets and treatments with

pleiotropic effects

The best of times

• Human Genome Project-Identify the 30,000 genes in the human genome-Sequenced 3 billion base pairs(2003)-<2% of the genome codes for protein-3 million loci with single base pair differences

• HapMap Project-Systematic genotyping of 3 million SNPs

• Available of high throughput technologies

The genome, the transcriptome and the proteome

• The genome is static• The transcriptome is dynamic and tissue and cell

specific. SNPs in a coding region may alter protein structure; SNPs in a promoter region may alter mRNA levels. Non-coding RNA can activate arrays of related genes.

• The proteome reflects post-transcriptional processes including RNA spicing, post-translational modification

Mendelian vs. Complex Diseases and Traits

1. Mendelian (single gene) a. Autosomal dominant (Huntington’s disease) b. Autosomal recessive (Cystic fibrosis) c. X-linked

2. Complex diseases and traits are multifactorial a. Oligogenic or polygenic susceptibility genes

(No one gene is necessary or sufficient) b. Locus heterogeneity (different genes

can cause same trait) c. Pleiotropic effects d. Environmental factors contribute

Genes EnvironmentGenes

+Environment

Causes of Complex Cognitive, Emotional and Motor Traits

Defining the phenotypes

• Define the phenotype of interest-Diseases-heterogeneity is the rule-Single domain phenotypes: declarative memory, depression, walking speed -Multiple domain phenotypes: executive function and walking speed-Covariance shared by traits

-Use biological markers or endophenotypes• Once genes are identified contrast carriers and non-

carriers to refine the phenotypes• Challenge: pleiotropy and polygenicity

Can a single gene influence thinking, feeling and moving

• Yes for Huntington’s, TDP-43• Relevant genes might influence transmitters, ion channels,

recovery from insults, plasticity, dendritic complexity, etc.• A gene may influence an area of the brain that has distal

secondary affects• A gene may influence several areas of the brain each of which

influences a distinct cognitive, emotional or motor process• A gene may influence several brain regions or

processes each of which affects more than one domain (Kovas and Plomin, Trends in Cog Sci, 2006)

Approach to Complex Diseases

Assess the influence of genetic factors (heritability)-Family studies-Twin studies

Identify specific genetic factors-Linkage analysis-Allele sharing methods-Case-control studies

Determining the Genetic Component

Familial aggregation-Measure family relative risk.-Suggests but does not prove genetic mechanism -Look for aggregation of one domain (thinking) in relatives of probands selected for another domain (moving)

Twin studies-If concordance rate for MZ twins > DZ twins some genetic influence

-Look for aggregation of one domain in MZ and DZ twins with issues in another domain

Heritability: the proportion of the variance in a disease or trait accounted for by genetic factors

Higher Concordance Rates for MZ vs. DZ Twins Shows a Significant Genetic Component

Trait MZ DZ

Schizophrenia 46 14

Insulin-dependent diabetes mellitus 30 6

PD < age 50* 100 17

PD ≥ age 50* 11 11

Survival** to 90 (females) 16.8 9.0

Concordance

*Tanner, 1999

** Hjelmborg et al, 2006

Identification of Genetic Factorsin Complex Diseases

• Linkage analysis: Identify extended families where disease or an endophenotype appears Mendelian (usually early-onset)

• Allele sharing methods, genome wide screening in affected sibling pairs (Identity by descent)

• Association studies in human populations

Identification of Genetic Factorsin Complex Diseases: Linkage Analysis

Identify large families where the trait appears Mendelian (Are there high density families with relevant phenotypes?)

Look for genes or DNA segments which segregate with the disease or trait.

Genes and DNA segments close to each other on a chromosome tend to be inherited together.

Look for coinheritance of polymorphic markers and disease.

Common Useful Genetic Markers

• Simple Sequence RepeatsTens of thousands in genome

Typically di-, tri- or tetra- nucleotide repeats

(GT)n

unique flanking DNA sequence

AT

• Single Nucleotide Polymorphisms (SNPs)Millions distributed throughout genomeSingle base pair substitutions

unique flanking DNA sequence

Association Studies

Comparison of allele frequencies between unrelated affected and unaffected individuals.

Disease-marker association exists when alleles at the marker locus occur with different relative frequencies in affected and unaffected individuals.

Most important: Use unaffected individuals from the same population!

Affected cases

Unaffectedcontrols

Case – Control

Association Study Approaches

Candidate gene search – functional variants (if possible) in gene with biological relevance:

Genome-wide scan – Dense set of markers

throughout genome:

•Single marker association

•Define common haplotypes–Assess haplotypes for association

Gene Identification in Complex Traits

using Candidate Gene Approaches

1. Select candidate genes based on biology and the availability of functional SNPs or SNP haplotypes

2. Select candidates for thinking, feeling and moving from the genes-Expressed in brain-Related to a specific domain-Identified based on biology (brain recovery-ApOE, inflammation-CRP, vascular risk genes-lipid related, intracellular signal transduction genes, longevity, energy metabolism)

3. Options limited by current hypotheses

What is the Significance of a Population Association Between a Disease and a Particular Allele (Genetic Variant)?

1) Allele is directly involved in the pathogenesis of the disease

2) The result is a false positive due to statistical error

3) The result is a false positive due to inadequate matching of cases and controls (population stratification)

4) Linkage disequilibrium

Whole Genome Association (WGA):

• In WGA, high density chip arrays containing hundreds of thousands of SNPs are used to screen the entire genome on a single array.

• Using cases and controls, WGA association results in the generation of thousands of genotypes. Identify SNPs and SNP haplotypes associated with disease.

• Then need to determine the disease causing SNPs among these, either coding sequence changes or possible promoter/enhancer etc variations.

Affymetrix Genotyping Technology

250 ng Genomic DNA

RE Digestion

Adaptor Ligation

Nsp NspNsp

Fragmentationand Labeling

PCR: One Primer Amplification

Complexity Reduction

Hyb & Wash

AA BB AB

250,000 Genotypes

Genetic approaches to thinking, feeling and moving: Centenarian studies

• Only ~1/10,000 individuals is 100 years old

• Exceptional longevity occurs with greater frequency in the siblings and offspring of Centenarians

• Longevity genes may contribute to successful cognitive, motor and emotional aging

• LonGenity PPG PI: Nir Barzilai focus on Ashkenazi Jews as a founder population

98

95

92

Barzilai et al, PLoS Biology 2006

104 90 years before

A Major Barrier to Genetic Studies in Centenarians

What is the appropriate control group?

A Major Barrier to Genetic Studies in Centenarians

What is the appropriate control group?

• Age mates of centenarians?

A Major Barrier to Genetic Studies in Centenarians

What is the appropriate control group?

• Age mates of centenarians?

• Alternative

– Study centenarians their offspring and ages mates of their offspring

• Hypothesis: Longevity gene frequency

Centenarian > Offspring > Controls

0

5

10

15

20

25

30

35

40

DM (%) MI (%) Stroke (%)

Pre

vale

nce

in

po

pu

lati

on

HTN (%)

Offspring of Centenarians are Less Likely to Have Age-Related Diseases

JAGS 2004; 52:274

P

O

C

**

****

**

CntnrnOffspringControl

** p<0.01

Modeling Changes in the Frequency of a Genotype as a Function of Age

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

60 65 70 75 80 85 90 95 100

Age

Gen

oty

pic

Fre

qu

ency

Longevity genes

Aging or “killing” genes

Genes not contributing to life-span

0

5

10

15

20

25

30

35

65 75 85 95 105

Age (Year)

Fav

ora

ble

gen

oty

pe

in

po

pu

lati

on

(%

)

CETP VV

APOC3 CC

ADIPOQ del/del

Favorable Longevity-Associated Genotypes in Unrelated 65-108 Year-Old Ashkenazi Individuals

Of the three longevity genes identified to date

All are associated with large lipoprotein particle size

The favorable form of CETP is associated with high HDL levels and large lipoprotein particle sizes and with successful cognitive aging

The favorable form of adiponectin is associated with succesful motor aging

0

10

20

30

40

50

60

70

80

CETP(ug/mL)

ADIPOQ

VV I/V

La

rge

LD

L

(%)

(ug/mL)

-- A/-

*

APOC3(mg/dL)

CC C/A

*

*p<0.05

“Longevity Genotypes” are associated with HDL and LDL particle size

*

CETP VV Genotype and Cognitive Function

*p<0.01

CE

TP

VV

fre

qu

ency

(%

)

0

20

30

40

10

MMSE<25 MMSE ≥ 25

*

Centenarians

*p<0.049

0

10

15

20

5

Dementia (n=31)

Non-demented (n=129)

*

EAS

Barzilai et al, Neurology 2007

Is Size of Lipoproteins Associated with Cognitive Function?

20

30

40

50

% large HDL size

*

*P<0.003

9.1

9.2

9.3

9.4

9.5

9.6

HDL size (nm)

MMSE>25

MMSE<25*

ADIPOQ: A Longevity Genotype with a Successful Motor Aging Phenotype

• The ADIPOQ del/del genotype is associated with longevity and a reduced risk of insulin resistance.

• The less favorable forms of ADIPOQ genotype has links with insulin resistance and metabolic syndrome, pathways which may influence motor function.

• Verghese et al. therefore examined the relationship between ADIPOQ del/del and gait performance in 322 subjects (mean age 78, 27% AJ, 63% women) who received quantitative gait measures.

Relationship between genotype and measures of gait

• In linear regression analysis, adjusted for age and sex, the favorable form of the ADIPOQ gene (del/del) was associated with better performance on stance, swing, and double support phases.

• These variables generally reflect balance and rhythm (Verghese et al. JNNP,2007).

Gait Variables Del/Del(n = 72)

Ins/Ins(n = 80)

Velocity, cm/sec 92.2 ± 25.5 93.3 ± 24.3

Cadence, steps/ min 101.7 ± 12.4 102.0 ± 12.1

Stride length, cm 108.4 ± 19.5 108.5 ± 21.2

Stance, % 62.6 ± 2.9* 63.8 ± 3.8

Swing, % 37.4 ± 2.9* 36.2 ± 3.8

Double support time, % 25.0 ± 5.2* 27.6 ± 8.8

Stride length variability, SD 4.3 ± 2.3 4.7 ± 3.3

*p<0.05 (linear regression adjusted for age & sex)

322 subjects (27% AJ, 63% women) Mean age 77.8y

Adiponectin del/del genotype is associated with better balance and rhythm on gait

Summary

• Define the phenotypes of interest with care considering the spatiotemporal expression

• Consider family aggregation and twin studies to look at distribution within families of domains of interest.

• Consider searching for genes that account for the covariance among traits

Summary

• Bank DNA (and other tissue?)• Consider candidate gene and WGAS in

cross sectional and longitudinal studies (LonGenity focuses on AJs, EAS on the Bronx population)

• Begin studies in midlife or early adult life to reduce influence of phenocopies

• Use genes to refine phenotypes

Many thanks

Genetic Research Methods: Advantages and Disadvantages

Advantages

• Identify single genes with large effects

• Survey the whole genome

Disadvantages

• Must be family-based

• Limited power for complex disorders

• Low resolution

Study the

Linkage

• More power for complex diseases

• Large samples for genes with small effect

Association

• Customize choices • Find selected associationsCandidate genes (100-1000 SNPs)

• Survey the genome • Expensive

• Bioinformatic challenge

• False positives

Whole genome (105-106 SNPs)

Pair Wise Concordance in Survival to Age 90+ Among Swedish, Danish and Finnish Twins

MZ

Concordant Total Pairs Pairs Concordance

Male 90+ 30 394 7.6%

Female 90+ 93 554 16.8%

Total 90+ 123 948 12.9% DZ

Concordant Total Pairs Pairs Concordance

Male 90+ 28 646 4%

Female 90+ 103 1096 9%

Total 90+ 131 1724 7.5%

Concordance = C/C+D

Hjelmborg et al, Human Genetics, 2006

Can Plasma HDL Levels Predict Longevity?

*p<0.0001 vs. Others

HD

L (

mg/

dl)

Females

0

40

50

60

70

80 SpouseOffspringProband

n=157

*

n=122 n=147

• These results support an association between specific genes and motor function.

8.6

8.8

9.0

9.2

9.4

9.6

9.8

60 65 70 75 80 85 90 95 100<

Age

HD

L P

arti

cle

Siz

e (n

m)

LD

L P

arti

cle

Siz

e (n

m)

Control

Offspring

Probands

20.4

60 65 70 75 80 85 90 95 100<

Age

20.6

20.8

21.0

21.2

21.4

21.6

Lipoprotein particle size as function of age

Heritability (h) of lipoprotein particle size 0.4-0.7

Barzilai et al JAMA 290:2030, 2003

Are lipoprotein sizes associated with protection from age-related diseases?(in offspring of centenarians)

20

30

40

50

60

% large HDL size

*

*P<0.003

20

30

40

50

60

70

% large LDL size

HealthyHTNCVD

*

Barzilai et al JAMA 290:2030, 2003

Lipoprotein and their size in healthy or subjects withthe Metabolic Syndrome (MS)

20

30

40

50

60

70

% large HDL size %large LDL size

Healthy

MS

*P<0.001

*

*

Barzilai et al JAMA 290:2030, 2003

0

5

10

15

20

25

30

1 2 3

HDL tertile groups

Min

imen

tal s

ocre

*

*

*

Average Mini-Mental Score of Tertile HDL Groups*

*p<0.04

J. Gerontol 57A, M712, 2002

(HDL 37±2 mg/dl)

(HDL 51±2 mg/dl)

(HDL 75±2 mg/dl)

0

2

4

6

8

10

12

14

16

CETP(ug/mL)

ADIPOQ

VV I/V

Le

vels

*

(ug/mL)

-- A/-

*

APOC3(mg/dL)

CC C/A

*

*p<0.05

Are “Longevity Genotypes” Associated with Clinically-Significant Phenotypes?

Cross Sectional HDL Levels (Data from the Framingham study)

Age

HDL (mg/dl)

50

50

Genetic Research Methods: Advantages and Disadvantages

Advantages

• Identify single genes with large effects

• Survey the whole genome

Disadvantages

• Must be family-based

• Limited power for complex disorders

• Low resolution

Study the

Linkage

• More power for complex diseases

• Large samples for genes with small effect

Association

Genetic Research Methods: Advantages and Disadvantages

Advantages

• Identify single genes with large effects

• Survey the whole genome

Disadvantages

• Must be family-based

• Limited power for complex disorders

• Low resolution

Study the

Linkage

Levels of analysis

• Genetics- study of single genes and their effects. Focus on the disease.

• Genomics-study of the functions and interactions many genes, their transcription and regulation

• Proteomics-Study of proteins• Metabolomics-Study of small molecules

(physiologic indicators) in plasma, tissues or cells including peptides, lipids, carbs, drugs