Autism: Connecting Genes to Brain to Cognition

Daniel H Geschwind, MD PhDDaniel H Geschwind, MD PhD

Departments of Neurology, Psychiatry and Departments of Neurology, Psychiatry and Human GeneticsHuman Genetics

UCLA School of MedicineUCLA School of Medicine

HUGO May 2007

Cognitive Function

Cerebral Structure

Gross Anatomy

Microscopic

MolecularChemical

         

DevelopmeDevelopmentnt

Genes Environment

Autism Genes

Rare Variation(mutation)

Laterality or Human BrainPatterning (circuit development).

Common Variation •Syndrome of ASD•Endophenotypes

Evolutionary/Cross Species Comparison (HumanCognitive Specializations)

Clinical Investigation

Basic Investigation

What is Autism?

• Autism is a neuropsychiatric syndrome first described in 1943.

• characterized by deficits in 3 domains:– Communication/

Language– Social interactions– Restrictive, repetitive

behavior

• Onset prior to 3 years of age.

• 4:1 ratio of male to female

• Effects 1 in 200 children today.

• More common than other childhood disorders, ie childhood cancer, DM, Down’s

• Normal life expectancy.

• 10-25 % will develop epilepsy

• Autistic behaviors shared by many other disorders, ie: TS, Fragile X, untreated PKU and others.

Idiopathic Autismis the most heritable of

neuropsychiatric syndromes

• Sibling recurrence risk is 6-10%

• This translates to a 25-65 fold increase over the general population risk (depending on what incidence rate you use).

• Twin studies: – MZ 60-70% concordance for strict autism

– 90% for autism spectrum

– Published DZ concordance 0-6%

– Heritability = 0.7 to 0.9

Autisms - Known Genetic Causes

About 10% of individuals with ASD have identifiable genetic disorders:

• Chromosome abnormalities: ~5-10% (1-2% (dup)15q)

• Fragile X syndrome: 1-2% (15-30% of children with FRAX 15-30% of children with FRAX have autistic features)have autistic features)

• Rett syndrome: <1%

• Tuberous Sclerosis: <1%

• Angelman syndrome: when dx’d with AS, ~50% have autism

• Very Rare mutations

Create a Large, Open Resource

Autism GeneticAutism GeneticResource ExchangeResource Exchange

• An open resource shared An open resource shared with the scientific with the scientific communitycommunity

• More than 750 families.More than 750 families.

• 10k SNP Genome Scan and 10k SNP Genome Scan and fine mapping datafine mapping data

• Phenotype data: Phenotype data: – ADI-R, ADOSADI-R, ADOS

– basic cognitive and language basic cognitive and language testingtesting

– physical/neuro examsphysical/neuro exams

– medical historiesmedical histories

• Karyotyping/molecular cytogenetics

Autism Full Genome Scans 1998 - 2003

IMGSAC (1998) 99 familiesIMGSAC (1998) 99 families

CLSA, Barrette et al. (1999) 75 familiesCLSA, Barrette et al. (1999) 75 families

Risch et al. (1999) 139 familiesRisch et al. (1999) 139 families

Phillipe et al. (1999) 51 familiesPhillipe et al. (1999) 51 families

IMGSAC (2001) 152 familiesIMGSAC (2001) 152 families

AGRE, Liu et al. (2001) 110 familiesAGRE, Liu et al. (2001) 110 families

Shao et al. (2002) 90 familiesShao et al. (2002) 90 families

Auronen et al. (2002) 38 familiesAuronen et al. (2002) 38 families

AGRE Yonan et al. (2003) 345 familiesAGRE Yonan et al. (2003) 345 families

Autism Genetics Collaborative (AGC)Duke (Margaret Pericak-Vance, Michael Cuccaro, John Gilbert); Mt. Sinai School of Medicine (Joseph Buxbaum, Jeremy Silverman, Christopher Smith); Paris Autism Research International Sibpair Study (Catalina Betancur, Thomas Bourgeron, Marion Leboyer); Stanford University (Joachim Hallmayer); University of Iowa (Veronica Vieland, Thomas Wassink); University of North Carolina (Joseph Piven); University of Toronto/Hospital for Sick Children - McMaster University (Steve Scherer, Peter Szatmari, Andrew Paterson); Vanderbilt University (James Sutcliffe, Jonathan Haines)

Autism Genetics Resource Exchange (AGRE)University of California - Los Angeles (Dan Geschwind, Stan Nelson, Rita Cantor, Maricela Alarcon, Kenneth Lange, Sarah Spence); Emory University (David Ledbetter, Christa Lese-Martin); University of Chicago (Conrad Gilliam)

Collaborative Programs of Excellence in Autism (CPEA) University of Washington (Gerard Schellenberg, Geraldine Dawson, Ellen Wijsman); University of Utah (William McMahon, Hilary Coon); University of Pittsburgh (Bernie Devlin)

International Molecular Genetic Study of Autism Consortium (IMGSAC)Oxford University (Anthony Monaco, Anthony Bailey, Janine Lamb); University of Bologna (Elena Maestrini); Deutsches Krebsforschungszentrum, Heidelberg (Annemarie Poustka, Sabine Klauk); University of Illinois – Chicago (Ed Cook); University of Michigan (Catherine Lord)

Autism Speaks, CIHR, CAN, Genome Canada, HHMI, Hospital for Sick Children Foundation, INSERM, MRC, NICHD, NIDCD, NIMH, NINDS, NLM Family Foundation, Swedish National Medical Council, Wellcome Trust, EU

Autism Genome Project (AGP)Autism Genome Project (AGP)

AGP Linkage analysis

N = 522

N = 731

N = 1168

# multiplex

11p12-13

Nature Genetics 2007

AGP Copy Number Variants

1 2

2p16: 2 affected siblingsNRX1

1q21: three familiesPreviously implicated in MR

17

17p12: three families SM, CMT

22

22q11.2: two familiesInterpretation complicated

17 de novo CNVs (14 found in both ASP)

18 CNVs overlap ASD-related rearrangements

Numerous overlapping/recurrent CNVs

Families with transmission of maternal 15q gains

DeNovo CNV found in:

• 3% familial cases and

• 10% of simplex cases

• 1% controls (all dups)

Geschwind and Levitt, 2007

The Autisms

Coexistence with other neurodevelopmental

disorders

AUTISM

epilepsy in 30%

Rare disorders

10%

Language

OCD

Social Behavior

“Endophenotypes”

ASD Endophenotypes• Language Delay (Age at speech milestones) *

– Buxbaum et al. 2002 (chr 2) Alarcon et al. 2002; 2005 (Chrom 7q35, Chrom 3p)

• Social Cognition (Responsiveness; Duvall et al. 2005)

– Chrom 11p (Z = 3.4), Chrom 17q (3.1), Chrom 5p (2.8)

• Non-verbal communication (Chen et al. 2005)

– Chrom 16p (OSA-Lod 3.8), Chrom 8q (OSA-Lod 3.4)

• Sex (Male vs. Female)* (Stone et al. 2004; Cantor et al. 2005)

• Head Circumference*

• Seizures/Regression

• NeuroImaging or Electrophysiology

• Gene Expression?

.Solutions to complex genetics: endophenotypes

StratificationBy Sex

Language delay

Endophenotype: Social Behavior (Social Responsiveness

Scale*)

SRS SCORE

Non Autistic Sibs of Autistic Probands _ AGRE25

20

15

10

5

0

Std. Dev = 39.47

Mean = 56.8

N = 61.00

SRS SCORE

180.0170.0

160.0150.0

140.0130.0

120.0110.0

100.090.0

80.070.0

60.050.0

40.030.0

20.010.0

0.0

Clinical Control Sibs _ WASH U 25

20

15

10

5

0

Std. Dev = 27.92

Mean = 28.9

N = 46.00

The SRS is an established quantitative measure of autistic social impairment which generates a singular severity score for autistic social impairment (higher score equates with greater severity).

(Parent-teacher correlation =.76) *Constantino et al., 2000, 2004

Social Cognition

• SRS --parent and teacher report

• Samples– 105 sibships (i.e.

families, ¼ of AGRE sample)

– > 200 sibpairs (i.e. combinations of sibs)

– Used all pairs of phenotyped sibs – unweighted

Molecular Endophenotype: Can peripheral gene expression

identify etiological subtypes of ASD?

• 1) Can gene expression profiles be used to differentiate autism subjects into etiological categories.

– Can we do this for subjects with “simple” causes of autism FMR1-FM or dup/idic 15q

• 2) Are any of the shared “candidate” pathways dysregulated in idiopathic autism ?

• 3) Mutations emphasize role of gene regulation in related neuro-developmental disorders

– A2BP1/FOX1 deNovo deletion (Leese- Martin et al. 2007)--neuronal specific splicing factor.

– FMR1 (RNA binding/processing).

– Rett syndrome (MeCP2).

Geschwind 2001 Nature Reviews NeuroscienceVol 2 (6): 437 Macmillan Magazines Ltd

(Microarray Analysis of ASD

with known genetic causes)

Gene Expression can distinguish ASD subjects by etiology

(dup)15, FMR1, Control

Gene Expression also overlaps significantly

Nishimura et al. in press

Identification of Autism Susceptibility Genes

• Find Chromosomal Region by linkage analysis.• Search through genic regions to find gene (s)

using association analysis.

Common VariantHigh Density SNP analysis Replication

Rare VariantRe-sequence candidate genes,Cytogenetic abnormalities Causal Disease Variant

Chr 7 Association with Age at First Word (2758 SNPs)

Basepair Position

P

eulav

0.1

0.01

0.001

138Mb 140Mb 142Mb 144Mb 146Mb 148Mb 150Mb 152Mb

CNTNAP2

KIAA1549

PRKAG2TIF1

ATP6V0A4OR9N1P ZNF398

FLJ42291

145500000.000145600000.000

145700000.000145800000.000

145900000.000146000000.000

CHROM.POS

100.0

5

6

7

8

9

383-3 85K (JT877)383-3 390K (JT2425)

Not deletedDeletedUnknown (all homozygous)

SNP Genotyping

390K

85K

145500 145600 145700 146000145800 145900

Chromosome 7 Location (kb)

0.7

0.5

0.6

0.8

1.0

0.9

Pro

be

Ra

tio

CNTNAP2 deletion in AU383

•21 SNPs exhibited the same non-Mendelian pattern of inheritance, ranging from 145,663,096 – 146,826,809 base pairs, within the CNTNP gene in autistic individual.

CASPR2 is enriched in human fetal

frontal cortex and basal ganglia

Brett Abrahams, PhD

Its expression appears to mark a cortical-striatal-thalamo (and limbic) circuitry known to be involved in response selection and inhibition and language learning,

Is Autism a Developmental Disconnection Syndrome?

• Emerging evidence for genes potentially involved in “connectivity”– Common variation

• GABARB3 (Multiple)

• Met (Cambell et al. 2006)

• CNTNAP2 (Alarcon et al. unpublished)

– Rare mutations• Neuroligin 3 and 4 (Soderstrom et al. 2003)

• Shank3 (Durand et al. 2006)

• CNTNAP2 (Strauss et al. 2006)

Neurexin 1,3 Neuregulins, CNTNAP2 (dNeurexin 4) and others are involved in neuronal migration, pathfinding pathways and synaptogenesis.

•Deficits are related to functioning of higher order association areas: language, social behavior, and mental flexibility all require rapid, extensive integration of higher order information via 2nd and 3rd order association areas.

•Circuits developed in primates

Summary

While ASD risk is largely genetic, its etiology is multi-factorial and very heterogeneous.

Rare genetic variants contribute to at least 10% of cases, and are likely to explain more (20%?).

Understanding the contribution of common variation and its interaction with rare variants is the next step.

We have had some success, but we are learning that the standard approaches alone are relatively weak.

Other methods: Microarray, Bio-informatics, Pathway Analysis?

WE NEED TO DROP DSM related measurements and study traits that may be related to underlying pathophysiology or biology.

Hope for a new understanding: “Developmental Disconnection?”

AcknowledgementsUniversity of Chicago (microsatellite genotyping)Conrad GilliamJJ LiuAmanda Yonan

Emory University (cytogenetics)Christa Leese-MartinDavid Ledbetter

CSHLJonathan SebatMike Wigler

Mattew StatePat Levitt

UCLA•Maricela Alarcon: Language QTL, •Jackie Duvall/Jen Stone: Pathway, •Sarah Spence: Phenotyping, •Rita Cantor Group (New QTLs)•Stan Nelson Group (SNPs)Jennifer Stone (Male risk factors)Yuhei Nishimura (microarray)

Washington University•John Constantino •Richard Todd (SRS)

•AGP

AGRE Clara Lonjonchere PhDNancy Hart webmasterCure Autism Now Foundation

Funding Sources: NIMH (R01 and STAART;

DHG)CAN (DHG), Autism SpeaksM.I.N.D. Institute (MA, SS)