Molecular Psychiatry (2007) 12, 376–384 & 2007 Nature Publishing Group All rights reserved 1359-4184/07 $30.00 www.nature.com/mp ORIGINAL ARTICLE Dissecting the locus heterogeneity of autism: significant linkage to 12q14 DQ Ma1, ML Cuccaro1, JM Jaworski1, CS Haynes1, DA Stephan2, J Parod2, RK Abramson3, HH Wright3, JR Gilbert1, JL Haines4 and MA Pericak-Vance1 1Center for Human Genetics, Duke University Medical Center, Durham, NC, USA; 2Translational Genomics Research Center, Phoenix, AZ, USA; 3Department of Neuropsychiatry and Behavioral Science, School of Medicine, University of South Carolina, Columbia, SC, USA and 4Center for Human Genetics Research, Vanderbilt University Medical Center, Nashville, TN, USA

Autism is a common neurodevelopmental disorder with a significant genetic component and locus heterogeneity. To date, 12 microsatellite genome screens have been performed using various data sets of sib-pair families (parents and affected children) resulting in numerous regions of potential linkage across the genome. However, no universal region or consistent candidate from these regions has emerged. The use of large, extended pedigrees is a recognized powerful approach to identify significant linkage results, as these families potentially contain more potential linkage information than sib-pair families. A genome-wide linkage analysis was performed on 26 extended autism families (65 affected, 184 total individuals). Each family had two to four affected individuals comprised of either avuncular or cousin pairs. For analysis, we used a high-density single-nucleotide polymorphism genotyp- ing assay, the Affymetrix GeneChip Human Mapping 10K array. Two-point analysis gave peak heterogeneity limit of detection (HLOD) of 2.82 at rs2877739 on chromosome 14q. Suggestive linkage evidence (HLOD > 2) from a two-point analysis was also found on 1q, 2q, 5q, 6p,11q and 12q. Chromosome 12q was the only region showing significant linkage evidence by multipoint analysis with a peak HLOD = 3.02 at rs1445442. In addition, this linkage evidence was enhanced significantly in the families with only male affected (multipoint HLOD = 4.51), suggesting a significant gender-specific effect in the etiology of autism. Chromosome-wide haplotype analyses on localized the potential autism gene to a 4 cM region shared among the affected individuals across linked families. This novel linkage peak on chromosome 12q further supports the hypothesis of substantial locus heterogeneity in autism. Molecular Psychiatry (2007) 12, 376–384. doi:10.1038/sj.mp.4001927; published online 19 December 2006 Keywords: genome screen; autism; single-nucleotide polymorphisms; extended families

Introduction Autism is one of the most heritable complex genetic disorders in psychiatry. A strong genetic component in Autism (OMIM 209850 (Online Mendelian Inheri- autism is indicated by an increased concordance rate in tance of Man, http://www.ncbi.nlm.nih.gov/entrez/ monozygotic (60% for narrow (diagnosed as autism), dispomim.cgi?id = 209850)) is a neurodevelopmental and 91% for broader phenotypes (diagnosed as autism disorder characterized by three areas of abnormality: spectrum disorder (ASD)) versus dizygotic twins (0% impairment in social interaction, impairment in com- for narrow and 10% for broader phenotypes),2,3 and a munication and restricted and repetitive patterns of 75-fold greater risk to siblings of idiopathic cases, interest or behavior. Developmental abnormalities are compared to the general population prevalence.4 apparent in the first 3 years of life and the characteristic Estimates of the number of involved in autism impairments persist into adulthood. With improved range from 2–10,5,6 to 15 or more,7 to 100 loci.8 detection and recognition of autism, resulting from a More than 12 genome-wide screens have been broadening of the diagnostic concept and systematic performed in autism.7,9–19 Results from these screens population approaches, a recent prevalence study indicate potential susceptibility genes spread across reports that autistic disorder affects as many as one in the entire genome. Although several promising 300 children in a US metropolitan area.1 regions have been indicated (e.g., on chromosomes 7, 15 and 17), no universally accepted gene has Correspondence: Dr MA Pericak-Vance, Center for Human emerged. In addition, hundreds of association studies Genetics, Duke University Medical Center, 595 LaSalle St, have been conducted on more than 130 autism Blg.7540, Box 3445, Durham, NC 27710, USA. candidate genes, based on location in a linkage peak E-mail: [email protected] Received 2 August 2006; revised 19 September 2006; accepted 24 and/or their potential biological function. A recent September 2006; published online 19 December 2006 review of published autism studies reveals at least Genome screen in extended autism families DQ Ma et al 377 one study indicating a positive finding for virtually Boards (IRBs) at the participating sites approved every chromosome.20 the study protocol. Families were ascertained using Collectively, these studies yield convincing evi- clinical referrals and through active recruitment in dence for multigenic inheritance with locus and various lay organizations. After a full description of allelic heterogeneity as well as epitasis playing a role the study was given to the families, written informed in autism etiology. Heterogeneity can significantly consent was obtained from parents and from children reduce the power of replicating and detecting who were able to give informed consent. For the linkages,21 which could partly explain the unsuccess- current study, a total of 27 extended ASD families ful search to date for autism genes. One approach to were available and, within these families, there were addressing genetic heterogeneity is to identify homo- 68 individuals with autism. These highly selective geneous subsets of the data using phenotypic para- pedigrees were chosen from a larger data set of 315 meters. The identification of phenotypic subtypes has multiplex Caucasian families and are noteworthy in been applied in autism studies and appears to have their extended family structure with both avuncular been successful in refining linkage results.22–25 and cousin pairs occurring across multiple genera- The use of highly selected extended families is tions (Supplementary Figure 1). another approach to improve the power of detecting All family history interviews are standardized to linkage in the presence of genetic heterogeneity. create a multigeneration pedigree of first, second and Terwilliger and Goring26 point out that ascertaining third-degree relatives. Identification of individuals with extended families potentially increases the homoge- clear or even questionable affection status lead to neity of the data set. In addition, these families further investigation in the form of a personal interview contain significantly more linkage information. All and an examination. This information was used to published genome screens with the exception of one15 create large pedigrees with extensive information about restricted their analysis to an affected sib-pair design. medical, developmental, behavioral and related dis- Although this represents a powerful approach, there orders. By this approach, we ascertained a substantial are also potential drawbacks in the use of cousin pairs number of extended families in which ASDs were and hence bi-linearity in linkage studies, particularly present. Four of the families contained parents who in psychiatric disorders, because of the possibility qualified as having an ASD, on the basis of a clinical of assortative mating. Hodge et al.27 explored these diagnosis of Asperger disorder (ASP). Unaffected status difficulties using simulation studies and found that in siblings was based on the absence of obvious the loss in linkage information was small especially indicators of an ASD, as determined by systematic when parental phase was known. questioning during the family history interview. The resulting 12 genome screens also used micro- The collaborative autism team from the Duke CHG satellite markers, with an approximate density of one and the WS Hall Psychiatric Institute contributed 17 marker per 10 cM. The feasibility of using of high- families, and nine families were ascertained by the density single-nucleotide polymorphisms (SNPs) for Vanderbilt CHGR. Probands for the study consisted of genome-wide linkage analysis and the substantially individuals between the ages of 3 and 21 years who greater linkage information they provide in complex were clinically diagnosed with autism using Diag- diseases has been affirmed.28–32 Results indicate that low nostic and Statistical Manual (DSM)-IV criteria. A genotyping error rates and map construction generate consistent set of diagnostic criteria was applied to all reliable and robust results. Moreover, the increased families. For this study, the presumptive clinical information content identified linkage signals that were diagnosis of autism was confirmed based on clinical not detected in low-resolution microsatellite scans. evaluation using DSM-IV diagnostic criteria, the In this study, a total of 26 extended families, Autism Diagnostic Interview, Revised (ADI-R) and including both avuncular and cousin pairs, were medical records. The ADI-R33,34 is a validated, semi- selected for genome-wide linkage analysis. We hy- structured diagnostic interview, which yields a pothesize that these extended families are a powerful diagnostic algorithm based on the DSM-IV criteria data set for linkage study,26 and that use of a high- for autism. All participants met current diagnostic density SNP genotyping assay, the Affymetrix criteria for autism and were included only if they had GeneChip Human Mapping 10K array, ensures that a minimal developmental level of 18 months on the genome-wide, sufficiently informative markers will Vineland Adaptive Behavior Scale Score,35 or an IQ have the potential to identify significant linkage equivalent greater than 35. These minimal develop- regions containing autism susceptibility loci. mental levels assure that ADI-R results are valid and reduce the likelihood of including individuals with only severe mental retardation. A best estimate Materials and methods clinical research diagnosis for autism was determined Family ascertainment by the clinicians at each of the research sites utilizing Families were ascertained at three sites: the Center for all available case materials. Subjects were excluded if Human Genetics (CHG) at Duke University Medical there was evidence of developmental disorders with Center, the WS Hall Psychiatric Institute and the known phenotypic overlap with autism (e.g., Prader– Center for Human Genetics Research (CHGR) at Willi syndrome, Angelman syndrome, tuberous Vanderbilt University. The Institutional Review sclerosis complex, Rett Syndrome and fragile X

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 378 syndrome), neurological, severe sensory disorders or r2 = 0.16 was set as the cutoff point for the LD level motor disorders. between markers for the multipoint analysis.45 The genetic marker distance was calculated by the local Genotype generation Ensembl program,46 based on 1 cM sex-average inte- A high-density SNP genotyping assay using the grated maps from deCode Genetics.47 Marker allele Affymetrix GeneChip Human Mapping 10K array frequencies were estimated from the data set using all was performed at the Translational Genomics Re- individuals.48 As the sample was comprised of search Institute. A total of 11 665 SNPs were geno- pedigrees of varying sizes, we assessed identity-by- typed on 192 samples; 11 563 have their sequence descent sharing (LOD*) between all pairs of affected

localization, of which 15 markers had only one allele. individuals within a family using the Spairs sharing The genotype efficiency for each SNP was calculated statistic49 and the exponential model,50 as implemen- before error checking. A total of 9581 markers had ted in MERLIN.43 A whole genome-wide simulation efficiencies > 0.9, of which 76 markers had a minor was conducted in 17 male-only affected families for allele frequency < 0.01. Thus, a total of 9505 markers both multipoint parametric and non-parametric ana- across the genome were used in the analysis. lysis to generate an empirical P-value for the marker Mendelian pedigree inconsistencies were identi- with the maximal LOD score by employing the fied using PEDCHECK.36 All of the genotype incon- simulation option embedded in MERLIN,43 using a sistencies identified were zeroed out from the total of 1000 replicates. The haplotyping function database. Further, inter- and intra-familial genetic embedded in MERLIN,43 was used to construct the relationships were verified, using RELPAIR37,38 at the whole chromosome 12 haplotypes to identify the beginning of the study, and by using 10% of all minimal shared region across the male-only families genotyped markers across the genome. Incorporating sharing positive LOD scores across the peak linkage the inconsistency results from PEDCHECK with the region. findings from RELPAIR, one family and one triad from another family were removed from analysis owing to Results multiple pedigree inconsistencies. In the end, 26 extended families with 65 affected (male-to-female Supplementary Figure 1 shows all pedigrees that ratio: 55:10) and a total of 184 samples were included were included in the final analysis. Out of a total of in the final analysis. This resulted in a total of 52 184 sampled individuals, 65 there had autism. affected-relative pairs; including 26 cousin, 10 full- Table 1 shows the results from two-point para- sibling, two half-sibling and 11 other pairs (second- metric analysis for all markers with HLOD X2.0 in cousin pairs, etc.). Finally, 17 of the 26 autism the overall data set. A peak heterogeneity HLOD of families contained only males affected with autism, 2.82 was found at rs2877739 on chromosome 14q12 and they were classified as male-only families. under the dominant model. Suggestive linkage evi- dence (HLODX2) was found on chromosomes Linkage analyses 1q23.2–23.3 (156–160 cM); 2q35–36.3 (211–230 cM); The selected families were analyzed for linkage using a multianalytical approach. Both parametric and non- Table 1 The two-point HLOD scores X2.0 in the overall parametric linkage analyses were performed. Para- data set metric two-point linkage analysis was performed using the FASTLINK program of the LINKAGE soft- Chromosome Marker Location HLOD Model ware package.39,40 Analyses were performed using genotype data only from affected individuals with 1 RS2369611 156.380 2.402 DOM 1 RS1858232 159.258 2.209 REC both dominant and recessive genetic models and 2 RS956133 211.470 2.153 DOM allowed for a disease allele frequency of 0.001 for the 2 RS720501 230.320 2.083 DOM 14 dominant model and 0.01 for the recessive model. 2 RS1114562 230.360 2.043 DOM These two-point limit of detection (LOD) scores were 2 RS1917126 230.490 2.185 DOM used to calculate genetic heterogeneity LOD scores 5 RS2078477 127.283 2.117 REC (HLOD) using HOMOG.41 Non-parametric two-point 6 RS1902946 16.080 2.618 REC analysis was conducted using MERLIN.42 Parents 2.589 DOM who qualified as ASP were coded as unknown with 6 RS654559 93.377 2.488 REC respect to autism in the linkage analysis. These 6 RS654118 93.378 2.488 REC individuals were classified as unknown, i.e. given 11 RS1986446 100.930 2.520 REC 2.652 DOM that only limited clinical data were available (e.g. No 12 RS722918 76.720 2.540 REC ADI-R data). However, analyses coding parents as 2.265 DOM affected did not substantially change the linkage 12 RS974349 80.290 2.177 DOM results. 12 RS1515561 102.346 2.102 REC Multipoint parametric and non-parametric linkage 14 RS2877739 20.870 2.817 DOM analyses were performed using the version of MER- LIN43,44 that allows for the adjustment of linkage Abbreviations: DOM, dominant model; HLOD, heterogene- disequilibrium (LD) between markers. In this study, ity, LOD score; REC, recessive model.

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 379 5q23.2 (127 cM); 6p25.3 (16 cM); 6q15 (93 cM); peak at rs1445442 (HLOD = 3.02 under the recessive 11q22.1 (101 cM); 12q14.2–14.3 (76-80 cM); 12q22 model). This region at 12q13.13–q15 (67–84 cM) was (102 cM). For two-point non-parametric analysis, the only region showing significant genome-wide none of the markers gave us LOD*X2.0. linkage evidence. Figure 1 presents the results of the multipoint non- For chromosome 12, the overall data set was parametric and parametric analysis results for the stratified by the gender of affected individuals in overall data set. For non-parametric multipoint the families. Under both multipoint parametric and analysis, a significant linkage peak (LOD* > 3.0) was non-parametric analysis, the maximum LOD score found at chromosome 12q at rs1445442 with increases to HLOD = 4.51 under the recessive LOD* = 3.2. Suggestive linkage evidence (LOD* > 2.0) model; genome-wide empirical P-value = 0.001 and was seen on chromosome 7p14.1 (63–64 cM). For LOD* = 4.22; empirical P-value = 0.0001 in male-only multipoint parametric analysis, significant linkage affected families at rs146122 (Figure 2). Using the one evidence was found on chromosome 12q with the LOD rule,51 the confidence interval of the peak region

chromosome 01 chromosome 02 chromosome 03

0.6 0.6 1.5 0.4 0.4 0.2 1.0 0.2 0.0 0.0 0.5 –0.2 –0.2 –0.4 0.0 –0.4 –0.6 –0.6 –0.5 050 100 150 200 250 050 100 150 200 250 0 50 100 150 200 cM cM cM

chromosome 04 chromosome 05 chromosome 06

1.0 1.0

1.0 0.5 0.5

0.0 0.5

0.0 –0.5 0.0

–0.5 –1.0 0 50 100 150 200 050 100 150 200 050 100 150 200 cM cM cM

chromosome 07 chromosome 08 chromosome 09

2.5 0.5 2.0 0.5

1.5

1.0 0.0 0.0

0.5

0.0 –0.5 –0.5 –0.5 0 50 100 150 050 100 150 050 100 150 cM cM cM

chromosome 10 chromosome 11

1.0 1.0

0.5 0.5

0.0 0.0

–0.5 –0.5

0 50 100 150 0 50 100 150 cM cM Figure 1 Results for non-parametric and parametric multipoint analysis in the entire data set. Solid gray line: multipoint non-parametric LOD*; solid black line: multipoint heterogeneity LOD score for dominant model; dot black line: multipoint heterogeneity LOD score for recessive model.

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 380 chromosome 12 chromosome 13 chromosome 14

1.0 0.6 3 0.4 0.5 0.2 2 0.0 0.0 1 –0.2

–0.5 –0.4 0 –0.6 –1.0 02040 60 80 100 120 0 50 100 150 0 20 40 6080 100 120 cM cM cM

chromosome 15 chromosome 16 chromosome 17 1.5 0.4

1.0 0.5 0.2 0.5 0.0 0.0 0.0

–0.2 –0.5 –0.5

–1.0 –0.4 0 20 40 6080 100 120 140 0 20 40 6080 100 120 140 0 20 40 6080 100 120 140 cM cM cM

chromosome 18 chromosome 19 chromosome 20

1.0 0.5 1.5 0.8

0.6 1.0 0.0 0.4 0.5 0.2

0.0 –0.5 0.0

–0.2 –0.5

0 20 40 60 80 100 120 0 20 40 60 80 100 0 20406080 100 cM cM cM chromosome 22 chromosome 21 1.0 0.6

0.4 0.5

0.2

0.0 0.0

–0.2 –0.5

001020 3040 5060 70 20 40 60 cM cM Figure 1 Continued.

is from 74.68 (rs717274) to 81.58 cM (rs1405467). multipoint LOD scores, the most likely minimal Supplementary Figure 2 provides the genome-wide candidate region is from 75 to 79 cM. Figure 4 multipoint non-parametric and parametric plots in displays an example of a segregating haplotype and by-gender data sets. Coding the parents diagnosed shared regions among affected individuals. as ASP as affected did not alter the LOD score (HLOD[rec] = 4.52). Discussion Eleven of the 17 male-only families consistently gave positive LOD scores throughout the linked More than 12 genome screens have been performed in region (67–84 cM). The remaining six male-only autism using microsatellites and data sets comprised families and eight of the nine with female affected almost exclusively of affected sib-pair families (par- gave negative multipoint LOD scores throughout this ents and their affected offspring). This is the first region. Figure 3 displays the shared region on genome-wide screen report using both a high-resolu- chromosome 12 (shared region defined as the haplo- tion 10K SNP panel and extended families comprised type transmitted from the same founder) between of avuncular and cousin pairs. It is also the first or among the affected in the male-only families. autism report employing chromosome-wide multi- Based on the male-only families exhibiting positive point parametric linkage analysis and using chromo-

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 381 some-wide haplotype construction to look at the that the gender dichotomy is an important factor in minimal shared region across affected individuals to the genetics of autism.54 narrow the linkage peak region. A significant genome- Extended pedigrees can contain more genetic wide linkage peak was found at 12q14.2, which has information than sib-pair families and can provide been undetected by previous genome screens, substantially more power to detect linkage, particu- although the results did not meet the corrected larly when there is genetic heterogeneity. These genome-wide significance in the entire data set.52,53 extended families may well represent a gene or a This linkage evidence was significantly enhanced in few genes of relatively large effect. This can be an families with only male affected individuals, result- advantage in detecting linkage even with a small ing in a HLOD = 4.51 under the recessive model with a sample size such as the current study of only 26 genome-wide empirical P = 0.001. That the linkage families.55 Previous studies have demonstrated that evidence is significantly enhanced in male-only small sample sizes can identify regions of interest in affected families, further supports previous evidence genomic screens for complex traits when the genetic effect is strong. For example, the initial genomic screen in late-onset Alzheimer’s disease that estab- chromosome 12 lished linkage to chromosome 19 and subsequently 5 56 Peak HLOD 4.51 led to the identification of apolipoprotein E, a major Peak HLOD* 4.22 susceptibility gene for late-onset Alzheimer’s disease, 4 57 1 LOD Unit included only 31 multiplex families. However, we cannot exclude the possibility that the identified 3 linkage peak may be specific only for those autism families with an extended family history and, thus, 2 may represent a small proportion of patients.

1 For genome-wide linkage screens, false-positive linkage evidence is a major concern. In high-density

0 SNP panels, LD between SNPs becomes a new significant source of LOD score inflation. To address 42 0 50 100 150 this issue, we employed a new version of MERLIN cM that performs chromosome-wide multipoint analysis while taking into account LD between markers, Figure 2 Gender-specific multipoint parametric and non- parametric analysis results on chromosome 12 by gender. although it is unlikely to make a difference in these Solid line: parametric heterogeneity LOD score (under data because the majority of parents (8/105) are recessive model); dotted line: non-parametric LOD*; black: genotyped. male affected only families; gray: families with female Inspection of the pedigrees (Supplementary affected. Figure 1) found several examples of affected sib-pairs

Figure 3 The two lines mark the shared region between male affected in the 11 male-only families with positive by-family multipoint LOD scores across the linkage region.

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 382

Figure 4 Example of a segregating haplotype and shared region between the affected (FAM16).

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 383 who shared the same two haplotypes from their input on this project. We also thank Drs Robert Delong parents. These data explain why the HLOD is more and Gordon Worley for referring patients and their significant under the recessive model rather than the families to the study. We thank the Translational dominant model despite the hypothesized vertical Genomics Research Institute (TGen) for their gener- transmission. osity and effort in performing the Affymetrix chip Defining the minimal linkage region for detailed assays. This research was supported, in part, by evaluation of potential candidate genes in a complex National Institutes of Health (NIH) program project trait such as autism is problematic. A number of Grant NS26630, and Grants R01 NS36768, R01 factors complicate the effort including genetic hetero- AG20135 and R01 NS42165; by the National Alliance geneity, small family sizes and variable phenotypes. of Autism Research (NAAR); and by a gift from the These studies, however, are critical because they Hussman Foundation. The research conducted in this guide the search for contributing susceptibility genes study complies with current US laws. We also by identifying high-priority regions upon which to gratefully acknowledge the resources provided by focus further efforts. Our approach to identify a region the AGRE consortium and the participating Autism within which to begin our detailed molecular in- Genetic Resource Exchange (AGRE) families. The vestigation included an examination of the 11 male- AGRE is a program of Cure Autism Now (CAN). This only families that contributed to the positive LOD work used the core resources of the GCRC (MO1 score establishing linkage in the region. The remain- RR-00095) and the CHGR at VUMC and the CHG at ing six male-only families, as well as eight of the nine DUMC. families with female affected had negative multipoint LOD scores across the linkage region. Based on these criteria, we identified a 4 cM (75–79 cM) long segment References as the most likely location of the chromosome 12 1 Yeargin-Allsopp M, Rice C, Karapurkar T, Doernberg N, Boyle C, autism susceptibility gene. The approach, albeit Murphy C. Prevalence of autism in a US metropolitan area. JAMA important, is not infallible. Small positive scores, 2003; 289: 49–55. owing to limited family size, could represent false- 2 Steffenburg S, Gillberg C, Hellgren L, Andersson L, Gillberg IC, positive results that might mask the true minimal Jakobsson G et al. A twin study of autism in Denmark, Finland, Iceland, Norway, and Sweden. J Child Psychol Psychiatry 1989; 30: candidate region. Thus, it is critical that this region be 405–416. viewed as a starting point that can be sequentially 3 Bailey A, Le Couteur A, Gottesman I, Bolton P, Simonoff E, Yuzda extended to include additional genes if the contribut- E et al. Autism as a strongly genetic disorder: evidence from a ing susceptibility variant is not found. British twin study. Psychol Med 1995; 25: 63–77. 4 Bolton P, Macdonald H, Pickles A, Rios P, Goode S, Crowson M et Within our identified most significant linkage peak al. A case-control family history study of autism. J Child Psychol region (most likely minimal shared region: 75–79 cM) Psychiatry Allied Disciplines 1994; 35: 877–900. from 60 710 030 bp (rs348691) to 64 239 801 bp 5 Pickles A, Bolton P, Macdonald H, Bailey A, Le Couteur A, Sim CH (rs4129000), more than 19 known and predicted et al. Latent-class analysis of recurrence risks for complex genes are found (Ensembl, June 2006 http://www. phenotypes with selection and measurement error: a twin and family history study of autism. Am J Hum Genet 1995; 57: ensembl.org/Homo_sapiens/index.html). These in- 717–726. clude the AVPR1A (Arginine vasopressin receptor 6 Folstein SE, Rosen-Sheidley B. Genetics of autism: complex 1A (61 826 000–68 832 000)) gene, which has been aetiology for a heterogeneous disorder. Nat Rev Genet 2001; 2: recently reported as a strong candidate for involve- 943–955. ment in autism susceptibility in association studies, 7 Risch N, Spiker D, Lotspeich L, Nouri N, Hinds D, Hallmayer J et 58,59 al. A genomic screen of autism: evidence for a multilocus etiology. and deserves continued scrutiny. SRGAP1 SLIT- Am J Hum Genet 1999; 65: 493–507. ROBO GTPase-activating protein, RHO 1 (62 524 808– 8 Pritchard JK. Are rare variants responsible for susceptibility to 62 823 751 bp), a member of the SLIT family of complex diseases? Am J Hum Genet 2001; 69: 124–137. proteins known to be involved in neuronal guidance 9 International Molecular Genetic Study of Autism Consortium. A full genome screen for autism with evidence for linkage to a region among other functions, and WIF1 (Wnt inhibitory on chromosome 7q. Hum Mol Genet 1998; 7: 571–578. factor precursor (63 730 674–63 928 374)), involved in 10 International Molecular Genetic Study of Autism Consortium. A developmental Wnt regulation, all lie within the genomewide screen for autism: strong evidence for linkage to critical region. chromosomes 2q, 7q, and 16p. Am J Hum Genet 2001; 69: 570–581. 11 Collaborative Linkage Study of Autism. An autosomal genomic In conclusion, the novel linkage peak on chromo- screen for autism. Am J Med Genet 2001; 105: 609–615. some 12q14 in this genome-wide screen further 12 Liu J, Nyholt DR, Magnussen P, Parano E, Pavone P, Geschwind D supports the hypothesis that there is substantial locus et al. A genomewide screen for autism susceptibility loci. Am J heterogeneity in autism, and that extended families, Hum Genet 2001; 69: 327–340. along with gender effects, may help delineate these 13 Meyers RM, Risch N, Spiker D, Lotspeich L. A full genome screen for susceptibility genes to autism by linkage analysis in ninety different loci. multiplex families. Am J Hum Genet 1998; 63: A302. 14 Shao Y, Wolpert CM, Raiford KL, Menold MM, Donnelly SL, Ravan SA et al. Genomic screen and follow-up analysis for autistic Acknowledgments disorder. Am J Med Genet 2002; 114: 99–105. 15 Auranen M, Vanhala R, Varilo T, Ayers K, Kempas E, Ylisaukko- We thank the patients with autism, and the family Oja T et al. A genomewide screen for autism-spectrum disorders: members who agreed to participate in this study, as evidence for a major susceptibility locus on chromosome 3q25–27. well as the personnel of the CHG at DUMC, for their Am J Hum Genet 2002; 71: 777–790.

Molecular Psychiatry Genome screen in extended autism families DQ Ma et al 384 16 Philippe A, Martinez M, Guilloud-Bataille M, Gillberg C, Rastam 36 O’Connell JR, Weeks DE. PedCheck: a program for identification of M, Sponheim E et al. Genome-wide scan for autism susceptibility genotype incompatibilities in linkage analysis. Am J Hum Genet genes. Paris Autism Research International Sibpair Study. Hum 1998; 63: 259–266. Mol Genet 1999; 8: 805–812. 37 Epstein MP, Duren WL, Boehnke M. Improved inference of 17 Yonan AL, Alarcon M, Cheng R, Magnusson PKE, Spence SJ, relationship for pairs of individuals. Am J Hum Genet 2000; 67: Palmer AA et al. A genomewide screen of 345 families for autism- 1219–1231. susceptibility loci. Am J Hum Genet 2003; 73: 886–897. 38 Boehnke M, Cox NJ. Accurate inference of relationships in sib-pair 18 McCauley JL, Li C, Jiang L, Olson LM, Crockett G, Gainer K et al. linkage studies. Am J Hum Genet 1997; 61: 423–429. Genome-wide and ordered-subset linkage analyses provide sup- 39 Lathrop GM, Lalouel JM, Julier C, Ott J. Strategies for multilocus port for autism loci on 17q and 19p with evidence of phenotypic linkage analysis in humans. Proc Natl Acad Sci USA 1984; 81: and interlocus genetic correlates. BMC Med Genet 2005; 6:1. 3443–3446. 19 Barrett S, Beck JC, Bernier R, Bisson E, Braun TA, Casavant TL et 40 Schaffer AA, Gupta SK, Shriram K, Cottingham RW. Avoiding al. An autosomal genomic screen for autism.Collaborative linkage recomputation in linkage analysis. Hum Hered 1994; 44: 225–237. study of autism. Am J Med Genet 1999; 88: 609–615. 41 Ott J. Linkage analysis and family classification under hetero- 20 Santangelo SL, Tsatsanis K. What is known about autism: genes, geneity. Ann Hum Genet 1983; 47(Part 4): 311–320. brain, and behavior. Am J Pharmacogenomics 2005; 5: 71–92. 42 Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin – rapid 21 Ott J. Analysis of Human Genetic Linkage, 1st edn. The Johns analysis of dense genetic maps using sparse gene flow trees. Hopkins University Press: Baltimore, 1985. Nat Genet 2002; 30: 97–101. 22 Shao Y, Cuccaro ML, Hauser ER, Raiford KL, Menold MM, Wolpert 43 Abecasis GR, Wigginton JE. Handling marker-marker linkage CM et al. Fine mapping of autistic disorder to chromosome 15q11– disequilibrium: pedigree analysis with clustered markers. Am J q13 by use of phenotypic subtypes. Am J Hum Genet 2003; 72: Hum Genet 2005; 77: 754–767. 539–548. 44 Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin – rapid 23 Buxbaum JD, Silverman JM, Smith CJ, Kilifarski M, Reichert J, analysis of dense genetic maps using sparse gene flow trees. Nat Hollander E et al. Evidence for a susceptibility gene for autism on Genet 2002; 30: 97–101. chromosome 2 and for genetic heterogeneity. Am J Hum Genet 45 Boyles AL, Scott WK, Martin ER, Schmidt S, Li YJ, Ashley-Koch A 2001; 68: 1514–1520. et al. Linkage disequilibrium inflates type I error rates in 24 Buxbaum JD, Silverman J, Keddache M, Smith CJ, Hollander E, multipoint linkage analysis when parental genotypes are missing. Ramoz N et al. Linkage analysis for autism in a subset families Hum Hered 2005; 59: 220–227. with obsessive-compulsive behaviors: evidence for an autism 46 Xu H, Gregory SG, Hauser ER, Stenger JE, Pericak-Vance MA, susceptibility gene on chromosome 1 and further support for Vance JM, Zuchner S, Hauser MA. SNP selector: a web tool for susceptibility genes on chromosome 6 and 19. Mol Psychiatry selecting SNPs for genetic association studies. Bioinformatics 2004; 9: 144–150. 2005; 21: 4181–4186. 25 Buxbaum J. The genetics of autism spectrum disorders. Medscape: 47 Kong A, Gudbjartsson DF, Sainz J, Jonsdottir GM, Gudjonsson SA, Psychiatry and Mental Health 10(2). 12-29-2005. Richardsson B et al. A high-resolution recombination map of the 26 Terwilliger JD, Goring HH. Gene mapping in the 20th and 21st . Nat Genet 2002; 31: 241–247. centuries: statistical methods, data analysis and experimental 48 Broman KW. Estimation of allele frequencies with data on design. Hum Biol 2000; 72: 63–132. sibships. Genet Epidemiol 2001; 20: 307–315. 27 Hodge SE. Do bilineal pedigrees represent a problem for linkage 49 Whittemore AS, Halpern J. A class of tests for linkage using analysis? Basic principles and simulation results for single-gene affected pedigree members. Biometrics 1994; 50: 118–127. diseases with no heterogeneity. Genet Epidemiol 1992; 9: 191–206. 50 McPeek MS, Strahs A. Assignment of linkage disequilibrium by 28 John S, Shephard N, Liu G, Zeggini E, Cao M, Chen W et al. Whole- the decay of haplotype sharing with application to fine-scale genome scan, in a complex disease, using 11 245 single-nucleotide genetic mapping. Am J Hum Genet 1999; 65: 858–875. polymorphisms: comparison with microsatellites. Am J Hum 51 Ott J. Analysis of Human Genetic Linkage, 3rd edn. The Johns Genet 2004; 75: 54–64. Hopkins University Press: Baltimore, Maryland, 1999. 29 Schaid DJ, Guenther JC, Christensen GB, Hebbring S, Rosenow C, 52 Lander E, Kruglyak L. Genetic dissection of complex traits: Hilker CA et al. Comparison of microsatellites versus single- guidelines for interpreting and reporting linkage results. Nat nucleotide polymorphisms in a genome linkage screen for prostate Genet 1995; 11: 241–247. cancer-susceptibility Loci. Am J Hum Genet 2004; 75: 948–965. 53 Kruglyak L, Daly MJ. Linkage thresholds for two-stage genome 30 Middleton FA, Pato MT, Gentile KL, Morley CP, Zhao X, Eisener scans. Am J Hum Genet 1998; 62: 994–997. AF et al. Genomewide linkage analysis of bipolar disorder by use 54 Stone JL, Merriman B, Cantor RM, Yonan AL, Gilliam TC, of a high-density single-nucleotide-polymorphism (SNP) genotyp- Geschwind DH et al. Evidence for sex-specific risk alleles in ing assay: a comparison with microsatellite marker assays and autism spectrum disorder. Am J Hum Genet 2004; 75: 1117–1123. finding of significant linkage to chromosome 6q22. Am J Hum 55 Baron M. Optimal ascertainment strategies to detect linkage to Genet 2004; 74: 886–897. common disease alleles. Am J Hum Genet 1999; 64: 1243–1248. 31 Sawcer SJ, Maranian M, Singlehurst S, Yeo T, Compston A, Daly 56 Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell MJ et al. Enhancing linkage analysis of complex disorders: an PC, Small GW et al. Gene dose of apolipoprotein E type 4 allele evaluation of high-density genotyping. Hum Mol Genet 2004; 13: and the risk of Alzheimer’s disease in late onset families. Science 1943–1949. 1993; 261: 921–923. 32 Sawcer S, Ban M, Maranian M, Yeo TW, Compston A, Kirby A 57 Namjou B, Nath SK, Kilpatrick J, Kelly JA, Reid J, James JA et al. et al. A high-density screen for linkage in multiple sclerosis – Stratification of pedigrees multiplex for systemic lupus erythema- International multiple sclerosis genetics consortium. Am J Hum tosus and for self-reported rheumatoid arthritis detects a systemic Genet 2005; 77: 454–467. lupus erythematosus susceptibility gene (SLER1) at 5p15.3. 33 Lord C, Pickles A, McLennan J, Rutter M, Bregman J, Folstein S et Arthritis Rheum 2002; 46: 2937–2945. al. Diagnosing autism: analyses of data from the Autism Diagnostic 58 Kim SJ, Young LJ, Gonen D, Veenstra-VanderWeele J, Courchesne Interview. J Autism Dev Disord 1997; 27: 501–517. R, Courchesne E et al. Transmission disequilibrium testing of 34 Rutter M, LeCouteur A, Lord C. Autism Diagnostic Interview, arginine vasopressin receptor 1A (AVPR1A) polymorphisms in Revised (ADI-R). Western Psychological Services: Los Angeles, autism. Mol Psychiatry 2002; 7: 503–507. CA, 2003. 59 Wassink TH, Brzustowicz LM, Bartlett CW, Szatmari P. The search 35 Sparrow SS, Balla D, Cicchetti D. Vineland Adaptive Behavior for autism disease genes. Ment Retard Dev Disabil Res Rev 2004; Scales, Interview edn. AGS Publishing: Circle Pines, MN, 1984. 10: 272–283.

Supplementary Information accompanies the paper on the Molecular Psychiatry website (http://www.nature.com/mp)

Molecular Psychiatry