The NRG1 exon 11 missense variant is not associated with autism in the Central Valley of Costa Rica

We are collecting autism trios with the goal of performing a population genetic study of autism in the isolated founder population of the CVCR. As this population was founded by a small number of individuals in the 16^th century, and grew exponentially in isolation until the 1970s, disease alleles present at the founding should be widely distributed in the current population. This project was approved under the guidelines of the Ministry of Health of Costa Rica, the Ethical committee of the National Children's Hospital in San Jose (Hospital Nacional de Ninos or HNN), and the Institutional Review Board at Mount Sinai School of Medicine in accordance with the Declaration of Helsinki. These approvals remain active.

Details of the case ascertainment and evaluation process have been previously described [1]. Briefly, families of individuals with possible autism contact us, or are contacted by us after expressing interest in the study, and are formally asked to participate using established informed consent criteria. All interviews and exams take place in the Neurodevelopmental Unit of the HNN. Possibly affected individuals and their parents are interviewed by experienced neuropediatricians using the Autism Diagnostic Interview Revised (ADI-R) [7]. The Autism Diagnostic Observation Schedule (ADOS) [8] is also administered to subjects and both of these assessments are videotaped for independent scoring by the best estimators. IQ tests are administered that are appropriate for the age and level of verbal communication of the subjects. The Vineland Adaptive Behavioral Scales [9] are administered and compared with the results of IQ tests to confirm diagnoses of mental retardation.

Subjects were included in this study if they met DSM IV-TR [10] criteria for autism in all three symptom domains (social interaction, communication or language, and behavioural abnormalities) as measured by the ADI-R and the ADOS. Additionally, age of onset for at least one symptom domain must have been less than 36 months. A diagnosis of autism spectrum was given to individuals who met ADI-R criteria but had an age of onset greater than 36 months, or were no more than one point below ADI-R criteria for autism in the social domain, and in either the communication or repetitive behaviour domains, but not in both, or if they met full criteria using the ADI-R but were one point below the cut-off score for autism in one domain of the ADOS. We also included subjects with a DSM IV-TR diagnosis of Asperger's Syndrome. Although we will not include probands with a DSM IV-TR diagnosis of Pervasive Developmental Disorder Not Otherwise Specified (PDD-NOS) in a future whole genome association study of autism, we did include them in this exploratory study. Subjects with a nonverbal IQ <35 were excluded unless they had an adaptive behaviour score much higher than the IQ measure. This is because there are no local Costa Rican norms for IQ testing and thus IQ may be underestimated in this population [1].

All individuals also had to have CVCR ancestry defined as having at least 6/8 great-grandparents (GGPs) born in the CVCR, or all 4 GGPs from either the maternal or paternal side (in which case only transmission from the parent with CVCR ancestry is considered). Furthermore, individuals cannot be related by fewer than 6 generations, although we did genotype siblings, if affected, in the current study. The sample genotyped for this study includes 143 independent families. There were 129 cases with autism, 67% of which also had a diagnosis of mental retardation. The rest of the cases did not have mental retardation and included N = 7 with autism spectrum disorder, N = 4 with PDD-NOS, and N = 3 with Asperger's Syndrome. Three autism cases also had siblings with either PDD-NOS (2), or autism (1). The mean age of the subjects was 5.75 years and the range was 3–13. Sixteen of the 143 cases were female. The three multiplex families consisted of two male sib pairs and one male:female sib pair.

A TaqMan SNP genotyping assay was designed using ABI Assay-by-Design File Builder Version 2.0 from the following genomic sequence:

The sequences used for the amplification primers are as follows: Forward primer, TGCATCGCCCTCCTTGTG; Reverse primer, AGAAAGGCATGGAGAAAGAAGGTTT.

Identification of the G allele was accomplished using the probe CACCACACACATGAT-8 labelled with VIC. Identification of the T allele was performed using CCACCAAACACATGAT-8 labelled with FAM. The SNP Genotyping Assay Mix was obtained from Applied Biosystems (Foster City, CA) at a concentration of 40×, and diluted to 20× with 1× TE. For each reaction, 2.50 μl of 2× TaqMan Universal PCR Master Mix, No AmpErase UNG (Applied Biosystems, Foster City, CA), 0.25 ul of 20× SNP Genotyping Assay Mix, and 2.25 μl of genomic DNA at 10 ng/μl were used for a total reaction volume of 5 μl using 384-well plates. The plates were run in the ABI 7900 thermal cycler with the following conditions: 95°C for 10 minutes hold, then 40 cycles of 15 second denaturation at 92°C and 1 minute of annealing/extension at 60°C. At the completion of the amplification and the absolute quantification read, the allelic discrimination assay read was performed to generate the allelic discrimination read file to be analyzed on SDS 2.1.

Microsatellite markers for haplotype analysis included 317J8-2123, D8S1810, 420M9-1395 and 487-2. Details regarding genotyping of these microsatellites are available upon request. Note that the order of these markers has changed in the Human Genome Browser March 2006 assembly since the publication of NRG1 haplotypes by Li et al. [10]. D8S1810 and 420M9-1395 have switched positions. Details regarding the microsatellite markers we utilized are available at deCODE Genetics [14].

The exon 11 missense variant was present in two cases diagnosed with autism, one case diagnosed with an autism spectrum disorder (one point below one domain of the ADI-R), and one case diagnosed with PDD NOS, in addition to co-morbid attention deficit hyperactivity disorder and a tongue tic (Case 129). The allele was transmitted from the mother in all cases but appears to have occurred de novo in Case 129. The missense allele was also present in 12 unrelated, other than by marriage, parental individuals, 7 fathers and 5 mothers, from 11 families who did not transmit the allele to cases. To be clear, both parents of Case 25 carry the T allele and they are not related to each other by fewer than 6 generations. The frequency of the allele in the non-transmitted chromosomes was therefore 12/267 = 0.045, remarkably close to the frequency estimate of 0.042 by Walss-Bass et al [2]. Haplotype analysis revealed the presence of 6 haplotypes carrying the missense T allele when considering only the closest microsatellite 317J8-2123, 0.229 Mb proximal to the exon 11 missense variant (Table 1). Please see Figure 1 for the position of microsatellite markers typed in relation to the exon 11 variant. After microsatellite 317J8-2123, the next closest markers are clustered together with D8S1810 at ~0.964 Mb upstream followed by 420M9-1395 and then 487-2 ~1.04 Mb upstream from the first marker. These three 5' markers were typed as most associations to NRG1 with SCZ have been to markers in this region. Three parental chromosomes from 3 independent families (one transmitted to a case and the other two non-transmitted) shared alleles at all 4 microsatellite markers typed over a distance of ~1.04 Mb (Table 1). Two other independent parental non-transmitted chromosomes also may share a different ~1.04 Mb haplotype however phase is uncertain.

We take a careful family psychiatric history on first, second and third degree relatives of all probands. Therefore, we reviewed the psychiatric history of all cases as well as persons carrying the T allele to look for the presence of psychotic or affective disorders. None of the 12 'T' carrier non-transmitting parents had a psychiatric disorder. However, AU025C, a 40 year old mother did have a sister with manic-depression and, AU113C, a 38 year old mother, did have a brother who had been hospitalized with SCZ (unfortunately DNA samples could not be obtained on these subjects). This inheritance pattern could represent incomplete penetrance in carriers of the missense allele. However, 113C and 25C did not share a haplotype with each other. Of note, 7 trios not segregating the T allele had a blood-related aunt or uncle with schizophrenia, which were not available for testing. An 8^th case had a parent with SCZ who did not carry the T allele and a parental sib with SCZ whose carrier status is unknown. A 9^th autism case had more than one first degree relative with SCZ, neither of whom carried the T allele. None of the three T-allele transmitting parents had a personal or family history of psychotic or affective disorders. Finally, Case 129 appears to carry a de novo occurrence of the T allele as we have observed no other non-mendelian transmissions from either parent to Case 129 in a panel of markers spanning chromosomes 15, 7, 17, 21 and 4 which should effectively rule out non-paternity.