Linkage and Association Studies Show Evidence of Neurexin and Neuroligin Involvement In Autism

Background: Autism is a complex neurodevelopmental disorder characterized by impaired social interaction, language/communication deficits and restricted, repetitive behavioral patterns. Previous studies established a strong influence of genomic variation in the etiology of autism and it is recognized as one of the most heritable complex neuropsychiatric disorders. A number of studies have linked members of the neurexin (NRXN) and neuroligin (NLGN) gene families to autism. Members of the NLGN family are proposed to act as splice site-specific ligands for beta-NRXNs. NLGN genes are located on the X chromosome and are involved in the formation and remodeling of central nervous system synapses.

Objectives: We sought to evaluate the involvement of NRXN and NLGN gene families in our autism dataset. Specifically, we were interested in determining if single base-pair variants located in these genes are over-represented in cases. Also, we are interested in assessing a potential sex-bias for this variation and the involvement of putative gene-gene interactions in the disease process.

Methods: We genotyped 26 SNPs located in NLGN3, NLGN4, and NRXN1 in our dataset consisting of 403 Caucasian American families. The Pedigree Disequilibrium Test (PDT) and the Family Based Association Test (FBAT) were used to test for association. Linkage analysis was conducted using two-point heterogeneity LOD scores (HLOD). Both recessive and dominant models with disease allele frequencies of 0.01 and 0.001, respectively, were analyzed. Taking into account the 4:1 ratio of males to females affected with autism, the HLOD, PDT, FBAT were also run in a subset of families containing only affected males. “Pseudo” controls were constructed, using the non-transmitted alleles of the parents, to confirm previously reported association results.

Results: FBAT identified significant associations in the entire dataset for two SNPs in NLGN4, rs5915658 (p=0.0006) and rs5915659 (p=0.005). This association remained significant after correcting for multiple comparisons (p=0.006 and p=0.05, respectively). Following FBAT in the male only subset, rs5915658 and rs5915659 remained significant (p=0.005 and p=0.03, respectively). FBAT in the entire dataset showed nominal significance (p=0.02) for rs7606758, in NRXN1. This association is also nominally significant in the male only subset (p=0.04). When evaluating the entire dataset, the highest HLOD score of 2.21 (recessive model) was observed for rs4971649 in NRXN1 on chromosome 2p16.3. In the male only subset, the HLOD score (recessive model) calculated for this same SNP was 1.82. No linkage was detected for NLGN3 and NLGN4 at Xq13 and Xp22 respectively. The case/pseudo-control analysis performed for NLGN3 and NLGN4 confirmed previously reported association results and allelic and genotypic associations for rs5915658 and rs5915659 were both significant.

Conclusions: The most significantly associated SNP, rs5915658, is located in intron 1 of the NLGN4 gene and is a good candidate for future functional studies. We are currently using logistic regression to evaluate the probability of epistasis, among NRXN1, NLGN3 and NLGN4, contributing to autism.