Transcript Searching for autism susceptibility genes - HGM2006

Searching for autism
susceptibility genes
Elena Maestrini
University of Bologna
International Molecular Genetic Study of Autism
Consortium
Helsinki June 2 2006
Autism
Neurodevelopmental disorder characterised
by impairments in 3 domains:
 Verbal and non-verbal communication
 Reciprocal social interaction
 Repetitive and stereotyped patterns of behaviours and interests
Onset before 3 years of age
Population prevalence of autism is ~10-20 per 10,000.
Male to female ratio of ~3:1.
75% of autistic people have mental retardation
~30% of cases have epilepsy
Belongs to the spectrum of Pervasive Developmental Disorders
(PDDs) which include Asperger syndrome, Atypical autism, Childhood
disintegrative disorder, PDD NOS. Prevalence of PDDs ~ 60/10,000
Autism is a complex disorder
1) TWIN studies
Monozygotic Twins (MZ)
Social
disorder
12%
Cognitive
disorder
4%
Social
and
cognitiv
e
No
disorder
92%
Autism
60%
Dizygotic Twins (DZ)
No
disorder
10%
Cognitive
disorder
10%
(Bailey et al, Psychol Med 25:63-67, 1995)
2) FAMILY studies
Sibling recurrence risk (~ 3%) for autism at least 30 times higher
than general population risk (~ 10/10,000).
A significant proportion of relatives are affected by the milder
phenotypes.
Approaches to identify susceptibility
genes
• LINKAGE studies using nonparametric methods (allele sharing
methods)
• ASSOCIATION studies
candidate genes
whole genome
• Chromosomal abnormalities
 Copy number variation
Non-parametric linkage studies of complex
disorders
More robust than model based approaches but ….
• They are insensitive
large numbers of sibpairs are required to detect a significant
increase in allele sharing
• They are imprecise
no assumption on number of genes and inheritance model can
not rely on recombination events for fine mapping
• Significance threshold
MLS > 3.6 genome-wide significance
often use simulations to estimate significance
Replication in an independent data-set
International Molecular Genetic Study of Autism
Consortium ( IMGSAC )
Team of clinicians and
researchers from 9
countries, coordinated by
Univ of Oxford (Prof A
Monaco, Prof A Bailey)
Collection of > 290
multiplex families
Inclusion criteria
• Autism Diagnostic Interview (ADI-R)
• Autism Diagnostic Observation Schedule (ADOS-G)
• IQ > 30
 Exclusion of other medical disorders
 Fragile X, tuberous sclerosis
 Cytogenetic abnormalities
IMGSAC GENOME SCREEN
Initial genome screen - 354 microsatellite markers in 83 sib-pairs, 11 extended families.
Typed additional markers under peaks of linkage in up to 268 ASP (307 ARP total)
3.5
3
152 ASP
(IMGSAC, 2001)
2.5
ASPEX MLS
D9S161 (2.12)
D16S497 (1.73 )
2
219 ASP
(Lamb et al, 2005)
1.5
1
0.5
0
1
2
3
4
5
6
7
8
9
10
Chromosome
11
12
13
14
15
16
17
18 19 20 21 22
X
Summary of genome scans
D^ M*
S^
A*
S^
C*
F*
I* B^
D^
I*
D^
F**
F**
7
6
5
4
3
A^
I^
9
8
11
10
2
C*
S^
I^M^
P^
A*
a* M^
C*
12
A*
A^
P*
1
I^
I^
13
14
15
16
17
P^M^
D^
S^
18
19
F*
D*
20
21
22
X
** MLS > 3.6
* MLS > 2.2
^ MLS > 1
I: IMGSAC (Lamb et al, 2005)
Further investigations of the linkage loci
1) Detailed linkage analysis to improve mapping
2) Candidate gene analysis
3) Association studies using high density SNPs
Reduce heterogeneity by sample stratification
based on different component traits of the autism phenotype
Phenotype
Language
Repetitive
behaviours and
sterotyped
patterns
Developmental
regression
Sex of
probands
NЎ of ARPs
Chr
region
Markers
Highest LOD scores
Reference
49 (PSD)
2q
D2S364/D2S335
NPL 3.32
Buxbaum et al (2001)
45 (PSD)
2q
D2S116
M MLS 2.86
Shao et al. (2002)
50 (PSD)
7q
13q
D7S1813
D13S800
M-HLOD 2.17
M-HLOD 2.54
Bradford et al (2001)
152 (QTL)
7q
D7S1799/D7S3058
QTL-Z 2.85 (WORD)
Alarcon et al (2002)
291 (QTL)
3q
17q
D3S3045/D3S1763 QTL-Z 3.10 (WORD)
D17S1290/D17S1301 QTL-Z 2.84 (WORD)
Alarcon et al (2005)
23 (IS)
15q
GABRB3
OSA-LOD 3.19
Shao et al (2003)
D1S1656
62 (OC)
1q
6q
19p
Buxbaum et al (2004)
D19S714
NPL 3.06
NPL 2.61
NPL 2.31
34
7q
21q
D7S483
D21S1437
NPL 3.7
NPL 3.0
Molloy et al (2005)
257 (170 MO,
145 FC)
17q
D17S1294/D17S798
M MLS 4.3 (MO)
Stone et al (2004)
109 (56 MO)
17q
D17S2180
M MLS 4.1 (MO)
Cantor et al (2005)
219 (145 MO,
74 FC)
7q
16p
15q
D7S480/D7S530
M MLS 2.55 (MO)
M MLS 2.48 (MO)
M MLS 2.62 (FC)
Lamb et al. (2005)
D6S1270
D16S407/D16S497
D16S407/D16S497
IMGSAC: effect of affected sibling sex on linkage
All (219 ASPs)
Lamb et al (2005) J Med Genet 42:132
3
3
chromosome 7
chromosome 16
2.5
2.5
Males p=0.075
MLS
Male pairs
(145 ASP)
Female containing
pairs (74 ASP)
Males p=0.026
2
2
1.5
1.5
1
1
0.5
0.5
0
0
0
50
100
150
200
250
0
20
40
60
80
100
120
140
160
3
chromosome 15
Non-males p=0.0011
3
chromosome 2
3
2.5
chromosome 17
2.5
2.5
2
2
MLS
2
1.5
1.5
1.5
1
1
1
0.5
0.5
0.5
0
0
0
0.5
1
1.5
2
2.5
3
0
20
0
0
20
40
60
80
100
120
140
• Multipoint MLS calculated using ASPEX sib_phase under additive model
• Significance assessed by permuting sibling sex 10,000x.
40
60
80
100
120
140
160
IMGSAC: Parent of origin linkage modelling
Chromosome 7, all ASP
Chromosome 9, all ASP
2.5
2.5
2
2
All (219 ASPs)
Paternal linkage
Maternal linkage
1.5
1.5
1
1
Lamb et al (2005)
0.5
0.5
0
50
100
150
200
250
0
40
80
120
160
• Possible parent of origin specific effects
• Involvement of an imprinted gene(s), & 2 loci underlying linkage
to chr 7q ?
Sex-specific linkage in chromosome 17q
Sex-stratified genome scan in 257 AGRE families
All
267 AGRE + 73 Vanderbilt families
MO
FC
Sutcliffe et al.
Am J Hum Genet 77:265 (2005)
Stone et al. Am J Hum Genet 75:1117 (2005)
Cantor et al. Am J Hum Genet 76:1050 (2005)
Serotonin transporter locus (SLC6A4)
SLC6A4/ 5HTT - 17q11.2
High blood platelet serotonin levels consistently detected in subgroups of
autistic individuals and their relatives
 Serotonin-reuptake inhibitors ameliorate some symptoms

Several association studies focused on two functional polymorphisms
 insertion/deletion polymorphism in the promoter (HTTLPR)
 HTT-VNTR in intron 2
Inconsistent results or modest association with S allele
Rare non-synonymous variants (Sutcliffe et al 2005)
Variants in SLC6A4 may have a small effect on serotonin blood levels

• ITGB3 identified as a QTL locus for blood serotonin levels in the Hutterites
•
population (Weiss et al. 2004, 2005). This effect is seen primarily in males
ITGB3 is localized on chrom 17 ~ 20 cM distal to 5HTT.
• Possible association of a Leu/Pro variant in ITGB3 with autism susceptibility,
with different effects in males and females (Weiss et al. 2006)
IMGSAC Candidate gene studies
2q24.2-q32.2
GENE
FUNCTION
RPRM
KCNJ3
NR4A2
TBR1
GAD1
DLX1
DLX2
RAPGEF4
CHN1
ATF2
HOXD1
KIAA1604
UBE2E3
NEUROD1
FRZB
NCKAP1
GULP
INPP-1
NAB1
apoptosis
~40 Mb
neuronal cells excitability
transcriptional regulation
brain development
neurotransmitter metabolism
brain development
brain development
neuronal signal transduction
neuronal signal transduction
transcriptional regulation
brain development
possible role in brain development and function
~ 190 genes
Resequencing or
DHPLC screening of
coding and regulatory
regions in 32 - 48 affected
individuals from autism
families that are
contributing to the linkage
peak
ubiquitination
brain development
brain development
apoptosis
apoptosis
phosphoinositides metabolism
transcriptional regulation
Test common variants for
association with autism by
case-control and/or TDT
studies in the whole
IMGSAC family sample.
IMGSAC Candidate gene studies
Chromosome 7
NCAM
RELN
LAMB1
LRNN1
PTPRZ1
CUTL1
DLX5 & DLX6
FOXP2
CHRM2
COPG2
SRPK2
SYPL
GRM8
CPA1 & CPA5
MEST
EN2
Neuronal cell adhesion molecule
Reelin
Laminin beta-1 chain precursor
Leucine-rich repeat protein, neuronal 1
Protein tyrosine phosphatase receptortype Z, polypeptide 1
Cut-like 1
Distal-less homeobox genes 5 & 6
Forkhead box P2
Cholinergic receptor, muscarinic 2
Coatamer protein complex, subunit
gamma 2
Serine/Arginine rich protein kinase 2
Synaptophysin-like protein
Glutamate receptor, metabotropic 8
Carboxypeptidase isoform 1 & 5
Mesoderm specific transcript homolog
Engrailed 2
Chromosome 16
TBX6
UBN1
A2BP1
ABAT
CREBBP
GRIN2A
KIAA1243
BFAR
EMP2
SSTR5
T-box 6
Ubinuclein 1
Ataxin 2-binding protein
4-Aminobutyrate aminotransferase
CREB binding protein
Glutamate receptor, ionotropic
KIAA1243 protein
Bifunctional apoptosis regulator
Epithelial membrane protein 2
Somatostatin receptor 5
• Rare missense variants found in RAPGEF4 (chr2) and RELN (chr7)
• Some evidence of association with common variants in ABAT and
GRIN2 (chr16)
Autism candidate gene studies
 Over 150 candidate genes studied in the last 10 years
 No clear association with autism
Position of genes and frequency of publications in candidate gene/association studies over the last decade
14
12
SLC6A4
Number of studies
10
HOXA1
8
GABRB3
RELN
FOXP2
6
HOXB1
UBE3A
FMR1
EN2
SLC25A12
4
MECP2
2
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
20 21 22
23
Chromosome




Heterogeneity and clinical complexity of autism
Different diagnostic and inclusion criteria used in different studies
Too small sample sizes
No comprehensive analysis of variation (only 1- few SNPs/gene)
HapMap
Public database of common human variation: > 3 millions SNPs
genotyped in 269 DNA samples from 4 populations
The block-like structure of LD
HOTSPOTS
• A large part of the genome falls into segments of strong LD, known
as “haplotype blocks”, separated by segments of low LD
• Within a block, variants are strongly correlated to each other and a
small number of distinct allele combinations (haplotypes) account for
most of the genetic variation in a population
Tagging SNPs
Select subset of SNPs which adequately summarises genetic
variability within the gene
TAG SNPs
 May not be common to different populations
 May depend on definition method employed
• Gabriel
et al. (block based selection of tag SNPs)
• r2 (htSNPs selected so that all SNPs are highly correlated
(r2>0.8) to at least 1 SNP in the tag set)
High density SNP association study for the
investigation of autism loci on chrom 7q and 2q
2.5
3
2
2.5
2
ASPEX MLS
1.5
1
1.5
1
0.5
0.5
0
0
0
0.5
1
1.5
2
2.5
2q24.2-q32.2
~ 40 Mb
3
0
50
100
150
200
250
Position (cM)
7q21.3q33
40 Mb
Power calculations carried out to determine the optimal selection of SNPs and
samples.
 126 parent-child trios from autistic multiplex families selected for IBD sharing
 200 gender-matched controls
Test statistics: HHRR (family based approach); Case-Control
Prof Anthony Monaco
Strategy overview
Download all HapMap SNP genotypes in chromosome 2 candidate region
HapMap release 13
(phase 1) CEU
18,389 HapMap SNPs
1 SNP/3.4kb
Define blocks of strong LD
891 LD Blocks
Average SNPs/block = 8
Identify
htSNPs
Average N of htSNPs / Block = 3.3
Align blocks with genes
Number of Genes = 183
Genotype intragenic block htSNPs in selected families
Number of Haplotype Blocks Required = 419
Number of htSNPs to Genotype = ~1480
G/G
A/C
A/T
G/C T/T A/C T/C
Test htSNPs for association to autism
Genotyped 1536 SNPs on
both chr 2 & chr 7 in 576
samples.
Results
Genotyped 1536 SNPs on both chr 2 & chr 7 in 576 samples on Illumina Platform
• 97.7% SNPs successfully typed (35/ 1536 excluded)
• Sample genotyping success ~ 99%
• 99.79% genotyping efficiency after removing failed SNPs/samples
Genotype calls for rs2368352
1,60
1,40
No rma li zed R
1,20
AA
1
AB
BB
0,80
NA11881
0,60
NA12006
0,40
0,20
0
247
0
171
0,20
0,40
56
0,60
Normalized Theta
0,80
1
Statistical analysis
Stratification analysis (STRUCTURE)
No evidence for population substructure on chromosome 2 or 7 between autism,
control and HapMap CEPH samples.
Case-control analyses.
• Single-locus logistic regression allowing for additive and dominance effects,
adjusting for gender main effect.
• Block-based haplotype analysis using GENEBPM algorithm (Morris A, 2005)
with dominance, adjusting for (i) gender main effect and (ii) gender main
effect and interaction.
Family-based analyses.
• Single locus TDT.
• Block-based haplotype analysis using TRANSMIT.
• Block-based haplotype analysis using GENEBPM algorithm, allowing for
dominance and parent of origin effects, comparing probands with internal
controls.
Assign prior probability of 0.01 to each block to overcome multiple testing issue.
Takes account of linkage signal and allows for underlying LD and heterogeneity (equivalent to
expectation that each region will contain at least 2 associated blocks).
Chromosome 7
Probands v/s unrelated controls
1.0
Experiment-wise posterior probability of association
IMMP2L
0.8
Strong evidence for association (90%)
NM015328
0.2
0.4
0.6
PTPRZ1/2
0.0
Posterior probability
Positive evidence for association (75%)
0
200
400
600
Block
800
Chromosome 7
Probands v/s internal controls (family-based analysis)
Experiment-wise posterior probability of association
1.0
FBXL13
LHFPL3
0.6
MUC3A/B
WNT16
0.2
0.4
CUTL1
0.0
Posterior probability
0.8
IMMP2L
0
200
400
600
Block
800
Parent of origin effects
FBXL13 (Block 188)
Individuals at greatest risk when inheriting causal variant from mother alone.
Posterior mean (SD) of the parent of origin effect of the causal variant: -3.02 (0.50).
IMMP2L (Block 376)
Individuals at greater risk when inheriting causal variant from father.
Posterior mean (SD) of the parent of origin effect of the causal variant: 1.25 (0.80).
LHFPL3 (Block 220)
Individuals at greatest risk when inheriting causal variant from father alone.
Posterior mean (SD) of the parent of origin effect of the causal variant: 0.72 (1.38).
Chromosome 2
Probands v/s unrelated controls
1.0
Experiment-wise posterior probability of association
TAI2HUMAN
0.6
ZNF533
NM018981
0.2
0.4
OSBPL6
0.0
Posterior probability
0.8
NOSTRIN
0
200
400
Block
600
800
Chromosome 2
Probands v/s internal controls (family-based analysis)
0.8
1.0
Experiment-wise posterior probability of association
UPP2
ZNF533
0.2
0.4
0.6
NM024770
0.0
Posterior probability
NOSTRIN
0
200
400
Block
600
800
Results summary
NOSTRIN (nitric oxide synthase trafficker) and ZNF533 (zinc finger protein
533) genes on chromosome 2, and IMMP2L (IMP2 inner mitochondrial
membrane protease-like) gene on chromosome 7 give positive results in
both case-control and family based analysis.
Strong evidence of differential risk according to the parental origin of the
causative variant for IMMP2L, FBLX13 (F-box and leucine-rich repeat
protein 13) and LHFPL3 (lipoma HMGIC fusion partner-like 3 ) genes on
chromosome 7.
This effect is not seen on chromosome 2.
What is known about these genes?
NOSTRIN: encoding nitric oxide synthase trafficker.
Nitric oxide (NO) is a potent mediator in biologic processes such as neurotransmission,
inflammatory response, and vascular homeostasis. Binds eNOS and triggers translocation
of eNOS to vescicle like subcellular structures, leading to inhibition of NO release
IMMP2L: inner mitochondrial membrane peptidase-like.
Implicated in Gilles de la Tourette Syndrome, a complex neuropsychiatric disorder
showing phenotypic overlap with autism spectrum disorder.
LRRN3 (leucine rich repeat neuronal 3) gene is transcribed in the opposite orientation
within an intron of IMMP2L
FBXL13: F-box and leucine-rich repeat protein 13.
Substrate-recognition component of the SCF (SKP1-CUL1-F- box protein)-type E3
ubiquitin ligase complex. LRRC17 leucine rich repeat containing 17 gene is transcribed in
the opposite orientation within an intron of FBXL13

Sequenced the entire coding sequence and putative regulatory regions of
NOSTRIN, ZNF533, IMMP2L and LRRN3 in individuals with the most significant
risk haplotype.
No novel coding variants identified.
Replication!
• Test top 5% of associated haplotype blocks from each
chromosomal region in a new independent sample (~200
trios + 200 controls) using the same SNPs (420 htSNPs)
• Use higher density Phase II HapMap data to refine the
haplotypic structure in the 4 top genes (NOSTRIN,
ZNF533, IMMP2L and FBXL13)
Perspectives

Large scale, high throughput analysis of
genome variation

Better characterization of the phenotype
•
•

component traits
study of milder phenotypes in relatives
International collaborations
NAAR AUTISM GENOME PROJECT: Analysis of >1000
multiplex autism families (Europe, USA, Canada)
Acknowledgements
University of Bologna
Elena Bacchelli
Francesca Blasi
Claudio Toma
Simona Carone
Prof Giovanni Romeo
Department of Biology
Medical Genetics Laboratory
S.Orsola-Malpighi Hospital
University of Oxford
Wellcome Trust Centre for Human Genetics
Janine Lamb
Gabrielle Barnby
Nuala Sykes
Andrew Morris
Prof Anthony Monaco
Department of Psychiatry
Prof Anthony Bailey
International
Molecular
Genetic
Study of
Autism
Consortium
Funding
• The Wellcome Trust
• The Nancy Lurie Marks
Family Foundation
• UK Medical Research
Council
• NAAR
• Telethon Italy
• European Commission