ASD: Genetic Heterogeneity

Rare mutations (copy number variants), de novo
1. More common in Simplex vs Multiplex families
  1. Different types of genetic contributions for multiplex vs simplex families?
2. Only account for a small percentage of cases
  1. 10% (Geschwind, 2011)
3. Apparently not as heritable as common variants
  1. But, they have variable penetrance and expressivity (see points below)
  2. May not be "causal" (5-10% of ASD, 1-2% of general population
  3. Perhaps just a risk factor (susceptibility vs causal)
4. Therefore, we don't expect un-affected relatives with BAP traits to have these mutations
5. But how is it that in families where there IS a de novo mutation, family members who don't carry that mutated gene still can display BAP traits?
6. BUT, one study has shown that unaffected parents can transmit this
  1. Sanders et al., 2011 (see Geschwind, 2011)
7. What are the mechanisms that account for these mutations?
8. Do these rare mutations account for just ONE phenotype, or multiple? (e.g., think Mendalian genetics)
  1. See Gesschwind, 2011
9. Also known as "Mendelian Mutations"?
10. Not mutually exclusive from "common variants"
  1. Can individuals have a combination of common and rare mutations that perhaps contribute to different aspects of the ASD phenotype?
Common Polymorphisms (common variants)
1. Dominant genetic model of autism
2. Multigenic inheritance of common polymorphisms contributing to ASD risk
  1. One common variant not enough -- multiple interactions that result in a phenotype?
3. Multiplex families (inheritance, heritability-- look at pedigrees)
  1. Vs. rare mutations, which occur more often in simplex families
4. Not causal of ASD (unlike rare mutations)
5. Expressed more subtly
6. To what aspect of the ASD phenotype are they related?
  1. Tricky: ASD phenotype itself is very heterogeneous
7. Chromosomal regions suggestive of linkage (see Eapen, 2011)
  1. Meta-analysis confirming 7q22–32 and showing sugges- tive evidence for regions 10p12–q11.1 and 17p11.2–q12
  2. 2q21–33, 3q25–27, 3p25, 4q32, 6q14–21, 7q22, 7q31–36, 11p12–13, 17q11–21
Methodological Limitations
1. Small sample sizes
  1. Power? Just how many individuals do we need? In the 1000s?
  2. 10,000 or more individuals needed? (Abrahams 2010)
  3. But despite larger sample sizes, still no replicability (Bill, 2009)
2. These lead to inconsistent results across studies/poor replicability
  1. But more consistent: chromosome 7q, 17q
3. Greater degree of locus heterogeneity than suspected?
4. Lack of standardised assessment methods for diagnosis, cognitive functioning and adaptability in ASD (Freitag 2007)
5. Inclusion of subjects with ASD features, but no clear ASD diagnosis (Freitag 2007)
6. Diagnoses difficulty -- Autism proper, AS or PDD-NOS? And how about with or without ID?
7. Let's say if we find a gene that's consistent in families--how do we know WHICH endophenotype it's linked with?
  1. As in, perhaps we're only testing certain aspects of communication/language/behaviour but not others? What have we missed out on?
8. When considering heritability (family transmission), how to determine who is affected and who is not?
  1. May incorrectly assign those with BAP traits to "unaffected" status (Piven, 2001)
9. Consider comorbidity in ASD--have the individuals been screened for other genetic disorders? Who has ID and who has none?
10. PDD diagnosis other than autism? As in, Asperger's, PDD-nos included as well?
  1. "proband heterogeneity"
11. Controls with a genetic problem? What inferences can we make?
  1. How does this translate to family members too?
12. How to define an "abnormality", when manintaining the endophenotypic approach to studying genes in ASD?
  1. Bailey et al, 1998
MAIN IDEA
1. No specific gene accounts for entire ASD syndrome
2. Multiple genes with multiple interactions
  1. Locus heterogeneity (variations in same gene) vs allelic heterogeneity (variations in multiple, different genes)? (Eapen, 2011)
  2. Gene interactions, plus gene and environment
  3. Just how many interacting genes are we dealing with?
  4. 104 genes, 44 genomic loci (Eapen, 2011)
3. ASD: a collection of rare disorders resulting from several genetic defects resulting in the same phenotype? (Eapen, 2011)
4. Idiopathic cases: no unifying pathology
  1. ASD is hard to study because it lacks any clear unifying pathology at the molecular, cellular, systems level (e.g., compared to alzheimer's)
5. Must also account for "ENVIRONMENTAL"', non-genetic factors
  1. E.g.: study twins raised in separate homes--but no study as yet
6. Use family/twin studies, molecular genetic studies, single gene disorder studies to examine which genes contribute to ASD
7. **When we say that the etiology of ASD is known for between 10-20% of cases, what do we mean? (Geschwind 2011)
8. The heterogeneity of ASD is reflected in the "spectrum" of AD itself
9. "COMPLEX GENETICS" vs "MENDELIAN GENETICS"
10. single cause of ASD only accounts for 1-2% of cases (the rest are idiopathic)
  1. These single cause cases are more homogenous, the idiopathic ones are heterogeneous
11. "In neurodevelopmental disorders, the timing, location and degree to which gene expression is disrupted dictate the emergent phenotype" (Geschwind & Levitt, 2007)
12. Comorbid with other psychiatric disorders
  1. But what's from different vs same mechanisms? For example, the genes for ASD converge on same parts of the brain, so features of ASD but also anxiety/depression?
Large genetic studies
1. Autism Genetic Resource Exchange (AGRE)--Geschwind (www.agre.org)
2. Autism Genome Project (AGP)--Szatmari et al, 2007
3. "Homozygosity Mapping Collaborative for Autsim" (HMCA), Morrow et al., 2008
Single-gene Syndromes
1. Fragile X, Rett Syndrome, tuberous sclerosis, Joubert Syndrome
2. These individuals show features of ASD at frequencies higher than expected
  1. These disorders are associated with an increased risk of ASD
3. Assess whether individuals with ASD also have these disorders (specifically Fragile X, when doing genetic study)
4. However, note that ASD individuals with these genetic disorders make up a minority of cases
Risk-factor model
1. Each gene contributes a risk to ASD
2. Only after a threshold is reached, then the person displays ASD phenotypes
3. Clustering of subthreshold risk alleles may account for BAP (Frietag 2009)
4. Accounts for the notion of "endophenotypes"
Methods for finding genes
1. Candidate gene vs genome-wide approaches?
2. Whole Genome Scan Results
3. Quantitative Trait Loci
4. Linkage Studies
  1. Mapping genes in families
  2. Alleles close together on the same chromosome to be transmitted together during meiosis
  3. 1) Full genome screens (consider all chromosomes)
  4. 2) Looking at certain chromosomal area of interest
    1. Hard to replicate across studies? But chromosome 7q and 17q more consistent
  5. See Freitag 2007 for a review (Table 2 for phenotypes explored in these linkage studies)
  6. Do linkage studies in small families first, then see if results can be replicated in larger families
Twin/Family Studies
1. MZ: 36-96%, DZ: 0-30% (Freitag, 2007)
  1. MZ: DZ ratio = >4:1
2. >90% heritability rates
3. *MZ concordance of <100% points to weak environmental influence
4. 2-6% recurrence risk in siblings (Spence 2004)
5. Prevalence 50-100x greater in siblings vs general population (Freitag, 2007)
6. For twin and family studies, must be specific about which ASD features the concordance/recurrence estimates are based on
  1. E.g.: social + communication deficits seem to co-occur more frequently than repetitive/stereotyped behaviour. Does this imply that different genes are involved?
7. Twin studies on Asperger's and PDD-NOS--> non conducted to date
8. "Findings from twin and family studies concerning a more broadly defined autism phenotype in nonautistic relatives of autistic individuals provide one potential complementary strategy for finding genes underlying this complex phenotype." (Piven, 2001)
  1. Confirmed by finding aggregation of social/communication deficits and repetitive/stereotyped behaviours in these families
9. Consider "infants" who show signs of ASD (Gerdts, 2011)
  1. Are they merely BAP traits, or early signs of ASD?
  2. Do longitudinal study to characterise this
Male:Female Ratio
1. 4:1, why more males than females?
2. But no evidence for X-linked loci
3. BAP rates for male vs female don't differ (Freitag, 2007)
4. But in many other studies, they show more BAP in males than females (Gerdts & Bernier, 2011)
Genes-Brain
1. "Convergence of pathways"
2. ASD genes have implications for neural development/function
3. Converge on brain regions that serve language/communication, social interaction, behavioural flexibility
4. Example: CNTNAP2
  1. Anterior regions for joint attention (precursor to language)
5. Frontal structures
6. Temporal
7. Amygdala
  1. Fewer neurons in amygdala? (Amaral, 2007)
  2. But amygdala also enlarged? (Abrahams, 2010)
8. Cerebellum
9. Increased brain volume (white and gray matter)
  1. Must consider what stage of development (perhaps enlarged in some regions in children, but decrease growth rate in adulthood?) - Amaral, 2008
  2. Macrocephaly in ~20% of ASD kids (Geschwind & Levitt, 2007)
    1. Macrocephaly: head circumference
  3. Increased neuron densities, size, etc. many different factors that can contribute to increased volume
  4. What does macrocephaly mean behaviourally/cognitively? (Gerdts 2011)
10. Errors in neuronal migration (most common finding)
  1. Temporal, frontal lobes
11. Synaptopgenesis/pruning/dendritic development
12. Cortical minicolumns, reduced thickness
13. "Disconnection Syndrome" (Geschwind)
  1. Not necessarily a disruption of previously connected regions (Geschwind & Levitt, 2007)
  2. Rather, a failure of normal development (that can have many dif causes)
  3. Higher-order, multimodal areas
  4. But must also consider subcortical involvement
  5. Disconnection between DLPFC and ACC regions for joint attention (impt precursor for language/social behaviour)
  6. Differential gene expressions, contributing to differences in disruption to frontal vs temporal lobes?
  7. "In other words, which systems are function- ally disconnected, and how severe and widespread the disconnectionis, willcontribute to theultimatephenotype." (Geschwind & Levitt, 2007)
Intellectual Disability
1. Individuals with ASD and ID -- genes cause the intellectual impairment
2. So, to find ASD specific genes, study ASD cases without ID? (Eapen, 2011)
3. Prevalence of ID in ASD (65%), Gerdts 2011
4. Is there a continuum of ASD on its own vs ASD occuring with ID? (Eapen, 2011)
5. "Although the syndromes often overlap, the autisms are not synonymous with global intellectual disability or mental retardation" (Geschwind & Levitt, 2007)
6. Verbal-performance IQ discrepancy: genetic? (Bailey et al, 1998)
7. If mental retardation in ASD is genetic, and if family members who display similar BAP traits don't have mental retardation, then those genes are not transmitted? (Bailey et al., 1998)
8. Relatives of those with ID also at higher risk of developing ID? (Bailey et al., 1998)
  1. NO--early reports say yes, but generally relatives have normal IQ (Gerdts, 2011)