CNTNAP2 in high functioning autism UCCAP Anna Werling 1, Regina Taurines 4, Elise Bobrowski 1,3 , Ronnie Gundelfinger 1, Marcel Romanos 4, Edna Grünblatt 1,2 , Susanne Walitza 1,2 1 University Clinics of Child and Adolescent Psychiatry, University of Zurich, Switzerland; 2 Neuroscience Center Zurich, University of Zurich and ETH Zurich, Switzerland Zurich 3 Department of Experimental Psychology, University of Regensburg, Regensburg, Germany; 4 Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy, University of Würzburg, Germany

Introduction: Autism spectrum disorder (ASD) is a neurodevelopmental disorder collectively characterized by the trias of impairments in social interaction and communication and by repetitive or unusual behaviors. ASDs include early infantile Autism, Asperger’s disorder and atypical Autism. ASD is often associated with intellectual disability and speech/language impairments. According to multiple family and twin studies with a substantial heritable component with 70-90% of autism etiology, ASD is considered to be one of he most strongly genetically influenced multifactorial childhood psychiatric disorder. Despite the high heritability, it is interfered to be complex and no major gene has been observed to be relevant for the majority of ASD diagnoses. In the past decades various studies have implicated a small amount of with either rare highly-penetrant mutations, low-penetrant common variants or copy-number variants (CNV) that together explain only about 15-30% of the population prevalence. The Contactin Associated -like 2 (CNTNAP2) gene has been discussed to be associated in ASD and other neurodevelopmental disorders. In this study, we aimed to elucidate the genetic association of CNTNAP2 gene within high functioning ASD (HFA) to eliminate intelligence related factors and combine all published results to-date in a meta-analysis for ASD and HFA . Fig. 1: The CNTNAP2 gene Fig. 2: CNTNAP2 gene encodes a member of the family

Fig. 2 Members of the neurexin family function as cell adhesion molecules and receptors and are localized at the juxtaparanodes of myelinated axons. They Fig. 1 The CNTNAP2 gene is located on the long arm 7 at position 35 (7q35) and is mediate interactions between neurons and glia during nervous system one of the largest genes in the . SNPs like rs7794745 and rs2710102 development and are also involved in localization of potassium channels within are considered to be associated with autism . differentiating axons.

Tab. 1: Clinical data of the case-control study and DFAM analysis Fig. 3: Meta-analysis results of the CNTNAP2 rs7794745 (A) and Würzburg samples Male (% total) Female (% total) rs2710102 (B) associations with HFA only studies presented as a forest plot Sex ASD, n=29 27 (93.1) 2 (6.9) Controls, n=35 32 (91.4) 3 (10.67) Mean SD Age (in y) ASD 11.52 3.64 Controls 10.86 2.83 IQ, Mean ASD 100.57 13.94 controls 110.31 13.23 Zürich samples Male (% total) Female (% total) Sex ASD, n=76 64 (84.2) 12 (15.8) Controls, n=98 76 (77.6) 22 (22.4) Mean SD Age (in y) ASD 11.24 3.10 Controls 11.74 3.03 Fig. 3: Blue boxes are the odds ratio (OR) and whiskers the confidential intervals IQ, SON ASD, n=75 108.68 16.03 (CI). The synthesis OR is given in orange. The meta-analysis of both SNPs did not Controls, n=98 112.18 12.59 result in significant association with either HFA (OR=0.95 p=0.37 n-studies=3 n=1308; OR=1.009 p=0.515 n-studies=3 n=1308, respectively) nor with ASD IQ, CFT ASD, n=68 105.16 13.66 (OR=0.99 p=0.53 n-studies=7 n=8576; OR=1.01 p=0.34 n-studies=5 n=7041, Controls, n=96 106.81 10.97 respectively) (data not shown). DFAM analysis Male (% total) Female (% total) Sex ASD,n=44 39 (88.6) 5 (11.4) Sibs, n=57 28 (49.1) 29 (50.9) Mean SD Age (in y) ASD 11.3 2.83 Sibs 10.5 3.12 IQ, Mean ASD 106.87 16.03 Sibs 107.89 8.55

Tab. 2: Case-control association study for HFA (Fisher’s Exact test)

SNP Minor allele P-value OR (95% CI) Würzburg samples rs7794745 T 0.856 1.124 (0.55-2.29) Conclusions rs2710102 C 0.858 0.929 (0.46-1.89) Zürich samples Since our study focused on a special population of ASD, rs7794745 T 0.020 1.757 (1.12-2.77) the HFA, we could eliminate any effects due to ID or rs2710102 C 0.736 1.104(0.71-1.72) language disabilities, which resulted in no significant A case-control association study for HFA was performed in two independent populations association. Therefore, we might conclude that if (Swiss, n=174, German, n=64). Individuals were genotyped for two SNPs on the CNTNAP2 CNTNAP2 plays a role in neurodevelopmental disorders, gene. The rs2710102 deviated in both populations for Hardy-Weinberg equilibrium (p- value<0.001), while rs7794745 did not deviate (data not shown). A significant association than rather in individuals displaying ID or developmental for the carriers of the T-allele of the rs7794745 with HFA was found only in the case- delays, as recently discussed. control sample from Zurich. Moreover, we performed a family based association (DFAM) analysis with siblings (n=57 References: Toma et al.(2013) Psychiatr Genet; Sampath et al. (2013) PlosOne; sibs, n=45 cases). No association could be proved by DFAM with any of the CNTNAP2 SNPs Anney et al. (2012) Hum Mol Genet; Li et al. (2010) Psychiatr Genet; Arking et al. with HFA. (2008) Am J Hum Genet.