Gene & Environment Interplay

lecture from week 6

what are the two varieties of gene-environment interplay?

gene-environment correlation & gene-environment interaction

gene-environment correlation

genetic control exposure to the environment

gene-environment interaction

genetic moderation of sensitivity to the environment (or environment moderation of relative importance of genes)

example of gene- environment correlation

people who inherit genetic predisposition for high creativity might also seek out an enriched environment

example of gene- environment interaction

children with genetic predisposition for lower creativity perform just as well as children with genetic predisposition fo higher creativity when living in enriched environment

rGE

gene environment correlation

passive gene-environment correlation

parental genes influence parental behaviors that play a role in determining the kind of rearing environment that they provide to their children

active gene- environment correlation

a person’s genes influence their own behaviors that play a role in determining how they shape and select their environments

evocative gene-environment correlation

a person’s genes influence their own behaviors that play a role in evoking different types of responses in other people

sources of rGE changes over lifetime

passive—matters more younger, whereas active—matters more with age

G by E (G x E)

gene - environment interaction

Compensation G by E

enriched environment (or absence of negative environment) compensates for/buffers against deleterious effects of geentic liability

Contextual Triggering/Diathesis-Stress G by E

phenotype seen only when environmental context and genetic liability are present

Social Control G by E

social restrictions prevent expression of genetic predisposition

Enhancement G by E

given a genetic predisposition (for positive behavior), some social contexts can lead to even better/improved outcomes

Diathesis

genetic vulnerability — is necessary but not sufficient for producing illness — needs environment to trigger it

Stress

environmental stressor / “trigger”

good practices for GxE research

large sample sizes, properly control for potential confounders, genome-wide approaches

analyses for causal inference

mendelian randomization

analyses for predicting genetic risk, cross-trait associations

polygenic risk scores

analyses for cross-trait associations

LD score regression for genetic correlations, genomic SEM

polygenic scores

take advantage of the polygenic nature of complex traits; create a cumulative genetic “score” for each person in a sample

how to create a polygenic score? (PGS)

need a large “discovery” GWAS, take effect sizes and p-values from the GWAS, weight effect size by the number of risk alleles a person has (0, 1, or 2) to create scores in independent “target” sample, add up across all the SNPS (each person will have one score, representing their relative genetic risk/liability for the trait in the original GWAS)

potential uses of PGS

1. as a “check” that a new GWAS is capturing the right measure

2. improve prediction of diseases

3. test for gene-environment interactions and correlations

potential issues with PGS for psychiatric disorders and behavioral traits

1. percent variance currently explained by PGS = small

2. GWAS-based risk models are population-specific (differences in SNP allele frequencies, LD patterns, and allelic effects between populations)

3. many other environmental factors also contribute to risk (age, lifestyle, socioeconomic status, stress & trauma)

mendelian randomization “randomized control trial”

assume that we are randomly assigned our genotypes at birth (mendelian randomization), which predict likelihood of our exposure; then any association between genotype and outcome can be viewed as a causal link between the exposure and the outcome

key assumptions of mendelian randomization

• genetic variant is associated with the risk factor

• genetic variant is not associated with confounders

• genetic variant influences the outcome only through the risk factor

horizontal pleiotropy

independent effect of exposure and outcome

vertical pleiotropy

exposure is effecting the outcome

what pleiotropy is tested in MR?

vertical pleiotropy (tested)

how are psychiatric disorders defined?

DSM5

Neurodevelopmental Disorders / Early Life Disorders

Intellectual Disabilities; Communication Disorders; Autism Spectrum Disorders; ADHD; Specific Learning Disorder; Motor Disorders

Attention Deficit Hyperactivity Disorder

characterized by:

1. Inattention Symptoms

2. Hyperactivity and Impulsivity Symptoms

ADHD subtypes — Predominantly Inattentive

6+ more inattentive symptoms but 5 or fewer hyperactive

ADHD subtypes — Predominately Hyperactive/Impulsive Presentation

6+ more hyperactive symptoms but 5 or less inattentive

Epidemiology of ADHD

• very persistent, continue to experience symptoms in adulthood

• relatively common — prevalence of 5-7% in children

• in childhood, male-to-female ration 4:1, in adults, ratio 1:1

Criterion to Diagnosis ADHD

• 6 symptoms in the past 6 months for under 17 years of age

• 5 symptoms in the past 6 months for adults

Environmental Risk Factors for ADHD

• exposure to toxins and pollutants, perinatal complications

• some “environmental” factors may be due to genetic confounding (smoking during pregnancy)

• strongest support for low birth weight, severe early life adversities

Heritability of ADHD

• 76% in children

• 40% in adults

• heritability similar across inattention and hyperactivity/impulsivity dimensions

• 50-80% of genetic factors shared across subtypes

Genetic Epidemiology of ADHD

• full siblings of someone or DZ twins — 8x higher risk

• MZ twins — 70x higher risk

Treatment for ADHD

• non-pharmacological treatments: behavioral parent training

• pharmacological treatments: stimulant & non-stimulant drugs, both of which increase extracellular levels of catecholamines (domaine, epinephrine)

• ritalin was most effective at treating ADHD in children, while amphetamines (adderall) were most effective at treating adults

Dopamine (DA) Hypothesis

dsyregulation in DA system impacts risk for ADHD

• malfunctioning of dopamine receptors, fewer dopamine receptors

• see higher density of DAT (Dopamine Transporters)

DA receptor

binds DA

DA Transporter

responsible for reuptake of DA out of synapse back into presynaptic terminals

Methylphenidate blocks DAT

increases synaptic DA, DA can bind to receptors

Default Mode Network (DMN)

connected system of brain regions that show activity when we are resting, daydreaming, etc.

DMN in ADHD

does not turn off (but methylphenidate reverses this effect)

Case-Control GWAS of ADHD

27 genomic risk loci, h²snp=14%

FOXP1 & FOXP2

known to play roles in synapse formation and the development of speech and learning

• poor language skills associated with higher inattention and hyperactivity

SORCS3

identified in GWAS and implicated in rare variants associated with ADHD

when are ADHD risk genes more highly espressed?

around early embryonic development — 19 weeks post conception

ADHD and traits in the population

ADHD represents the extreme end of a quantitative trait

Autism Spectrum Disorders

characterized by

• differences in social communication

• restricted or repetitive behaviors

heterogeneous and complex, with lots of variation in symptoms, language level, intellectual functioning, and co-occurring psychiatric and medical difficulties

Epidemiology of ASD

• 3% in population now

• increased prevalence may be due to multiple factors — broadening diagnostic criteria, increased awareness and recognition of autism spectrum-like traits

• male-to-female prevalence ration ranges from 3:1 to 4:1

female protective effect

need greater genetic “load” to reach threshold for diagnosis

• under diagnosis and misdiagnosis in women

environmental risk factors of ASD

• pre-and perinatal factors: parental age, birth complication, pre-term birth, gestational diabetes

• some factors may be confounded with genetics

Genetic Epidemiology of ASD

• twin studies estimate high h², ranging from 64-99%

• MZ cotwin has 88-92% risk

• relatives who do not have the disorder tend to have like traits (e.g. communication difficulties)

• trait dimensions show limited genetic overlap

common/polygenic risk for ASD

common risk genes for ASD are most highly expressed in prenatal developing cortical tissue

Copy Number Variants for ASD

• inherited CNVs are present in ~3% of individuals with ASD

• de novo CNVS are present in 4-10% of individuals with ASD

inherited CNVs can be present in unaffected siblings and parents — suggesting incomplete penetrance

Rare Mutations on ASD

• rare inherited variants tend to have smaller effect sizes than de novo rare mutations

• de novo mutations in protein-coding regions contributed to risk in 30% of ASD cases

Sex Differences in Genetic Risk for ASD

• evidence for “female protective effect”

• female cases show increased burden of rare risk variants and polygenic/common genetic risk

• females show higher distribution of genomic risk score (accounts for common and rare variants)

Multiple Genetic factors contribute to Parental Age Effects (ASD)

• de novo mutations increase with paternal age

• inherited rare variants associated with younger paternal age

• PGS for edu attainment associated with increased paternal and maternal age