Association of NCAM1 Polymorphisms with Autism and Parental Age at Conception in a Chinese Han Population

Abstract

Aims: The neural cell adhesion molecule (NCAM) has been reported to be involved in the development of the central nervous system and its mRNA level might decrease in the serum of autistic patients. However, there was no evidence of the association of the NCAM1 gene polymorphisms with autism. In the present study, we enrolled 237 children with autism and 451 healthy control subjects. Then, we used the direct DNA sequencing for genotyping five tag single-nucleotide polymorphisms (SNPs) in the NCAM1 gene. Results: By using case–control association analyses, we found that three SNPs at the NCAM1 gene were associated with autism (rs4937786, p=0.015; rs12418058, p=0.0076; rs1436109, p=0.0023). Two of them remained significant after the Bonferroni multiple testing correction (rs12418058, p_corrcted=0.038; rs1436109, p_corrcted=0.012). Moreover, two of the SNPs were associated with the parental age at conception in autism (rs12418058, p=0.037; rs1436109, p=0.01). Conclusion: These results showed that NCAM1 might play an important role in the pathogenesis of autism.

Introduction

Autism is a kind of common pervasive neurodevelopmental disorder. The typical clinical characteristics of autism include deficits in social interaction, communication, and the presence of repetitive or stereotypic behaviors (Cooper, 1995). These clinical symptoms become obvious in the first 3 years of life in childhood (Lord et al., 2000). Previous research suggested that strong genetic components might be involved in the susceptibility to autism (Rutter, 2000; Folstein and Rosen-Sheidley, 2001; Veenstra-Vanderweele et al., 2003). Although the etiology of autism is unknown, family and twin studies have indicated autism as a highly heritable neuropsychiatry disorder with the heritability of about ∼90% (Steffenburg et al., 1989; Folstein and Rosen-Sheidley, 2001).

Up to now, hundreds of susceptibility genes have been implicated in the etiology of autism and this disease has also been identified as genetically heterogeneous (Miyauchi and Voineagu, 2013). Therefore, few of the susceptibility genes could be validated among various populations. Therefore, many studies focused on some special neurodevelopmental processes, such as synaptic transmission and neuronal cell adhesion. Accumulating evidence has suggested that synaptic genes might be involved in susceptibility to autism (Kim et al., 2008; Peca et al., 2011). Association studies and mutation analysis of candidate genes have implicated the substantial roles of the synaptic genes Neurexin 1 (NRXN1), Neuroligin 3 (NLGN3), Neuroligin 4 (NLGN4), SH3 and multiple ankyrin repeat domains 3 (SHANK3), and contactin-associated protein-like 2 (CNTNAP2) in autism (Kumar et al., 2009; Kenny et al., 2013).

The neural cell adhesion molecule (NCAM) has been reported to be involved in the development of the central nervous system and its mRNA level might decrease in the serum of autistic individuals (Plioplys et al., 1990). However, another study showed that NCAM mRNA levels were not altered either in serum or postmortem brain samples (Purcell et al., 2001) of autistic children. It has been reported that autism might be associated with the neuronal cell adhesion molecule (NRCAM, 7q31.1) gene polymorphisms (Marui et al., 2009). The CAMs are immunoglobin (Ig) superfamily members, which exist in the nervous systems of both vertebrates and invertebrates. CAMs usually act as surface membrane proteins and they include multiple Ig domains at the N termini and a transmembrane intracellular domain or a glycophosphatidylinositol-linked membrane anchor at the C terminus (Lane et al., 1996).

However, to our knowledge, no genetic study previously focused on the gene coding for the neuronal cell adhesion molecule 1 (NCAM1), as a susceptibility gene of autism. The NCAM1 gene has been implicated in many psychiatric disorders such as bipolar disorder and schizophrenia (Vawter, 2000). The NCAM1 (11q23.1) gene contains 18 exons and it spans approximately 314 kb. It exerts a number of important functions in the development of the central nervous system (Fujita et al., 2000) and it is involved in the plasticity of the adult brain (Doherty et al., 1995; Gascon et al., 2007). Interestingly, studies on animal models found that mice deficient for the NCAM showed behavioral abnormalities in the adulthood, including increased intermale aggression and neuroendocrine response (Stork et al., 1997). NCAM deletion in rats was found to be related to a cognitive and behavioral phenotype reflective of impulsivity, which may be one of the typical clinical characteristics of autism (Matzel et al., 2008). Thus, in the present study, we investigated a panel of markers in NCAM1 (rs4937786, rs12418058, rs1436109, rs584427, and rs605843) in association with autism as well as the parental age at conception in a Chinese Han sample of 237 patients with autism and 451 healthy control subjects.

Moreover, a body of evidence has suggested that advanced paternal age (APA) might increase the risk of autism, schizophrenia, and other neuropsychiatric disorders (Crow, 2003; Lopez-Castroman et al., 2010; Buizer-Voskamp et al., 2011; Hehir-Kwa et al., 2011; Kong et al., 2012; Buizer-Voskamp et al., 2013). Therefore, we also intended to explore the potential relationship between NCAM1 polymorphisms and the paternal age at conception in autistic children.

Materials and Methods

Subjects

The sample for this study consisted of 237 children affected with autism and 451 adult healthy control subjects. These probands and controls were recruited at the Beijing Children's Hospital, China. Among the 237 patients with autism, 207 were male and 30 were female. The age of the children at the time of testing ranged from 2 to 17 years. The assessments of autism were established by two senior psychiatrists using the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) criteria, Autism Behavior Checklist (ABC) (Krug et al., 1980), and Childhood Autism Rating Scale (CARS) (Schopler et al., 1980). Excluded criteria included children with fragile X syndrome, tuberous sclerosis, a previously identified chromosomal abnormality by karyotyping analysis, and non-Han Chinese ancestry. The paternal ages at conception (20–42 years) were reported by parents and recorded in the medical documents of autistic children.

Healthy control subjects were eligible for inclusion if they and their parents denied any history of psychiatric disorders. The healthy control subjects consisted of 451 (408 males and 43 females, aged 18–48 years old) Chinese Han subjects. All the healthy control subjects and cases came from the Northern regions of China.

Ethics statements

This study was approved by the Ethics Committee of the Beijing Children's Hospital. All subjects provided written informed consent for participation in this study. Written informed consents for children were obtained from their legal guardians.

Genotyping

Venous blood samples were obtained from children with autism and healthy control subjects. Genomic DNA was extracted using the QIAamp DNA Blood Mini Kit (Qiagen). Information on single-nucleotide polymorphisms (SNPs) was obtained from the dbSNP (www.ncbi.nlm.nih.gov/SNP/) and the international HapMap project (www.hapmap.org/).

PCR amplification was performed in a 25 μL volume containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl₂, 200 mM of each dNTP, 0.3 mM of each primer, 0.6 U of Hotstart Taq DNA polymerase, and 30 ng of the genomic DNA. The conditions used for PCR amplification were an initial denaturation phase at 94°C for 5 min, followed by 36 cycles at 94°C for 30 s, annealing at 55°C–61°C for 30 s, and extension at 72°C for 40 s, followed by a final extension phase at 72°C for 10 min.

PCR products were sequenced, respectively, by DNA sequencing after cleaning the PCR products using a BigDye Terminator Cycle Sequencing Ready Reaction Kit with AmpliTaq DNA polymerase (PE Biosystem). The fragments were separated by electrophoresis on an ABI PRISM genetic analyzer (Applied Biosystem).

Statistical analyses

Deviation from the Hardy–Weinberg equilibrium (HWE) for genotype frequency distributions was tested using the chi-square goodness-of-fit test. All those with frequencies of minor alleles greater than 5% were used as genetic markers in this study. The Haploview (version 4.0) program was used to calculate pairwise linkage disequilibrium (LD) and haplotype-based association analysis using the option of determining blocks based on the criteria defined by Gabriel et al. (Gabriel et al., 2002; Barrett et al., 2005). Allele and genotype frequencies for each polymorphism were compared between patients and controls using chi-square tests. The one-way analysis of variance (ANOVA) tests were used to compare the paternal age at conception among various genotype carriers by using the Statistical Product and Service Solutions (SPSS) software version 13.0. The power of sample size for association tests was evaluated using the Genetic Power Calculator program (http://pngu.mgh.harvard.edu/∼purcell/gpc/) (Purcell et al., 2003). For the disease locus, the minor allele frequency of 0.15, the prevalence of 0.006, the genotype relative risk of Aa=1.5 and AA=1.5, and a D-prime=1 were used to perform the analyses. The significant level was set at p<0.05 (two sided).

Results

The genotype distributions of all five SNPs at the NCAM1 gene detected met the criteria of the Hardy–Weinberg equilibrium (p>0.05). The statistical power of the common SNPs in our study was about ∼0.73.

By using the case–control association analyses, we found that three SNPs at the NCAM1 gene were associated with autism (rs4937786, p=0.015; rs12418058, p=0.0076; rs1436109, p=0.0023; Table 1). Two of them remained significant after the Bonferroni multiple testing correction (rs12418058, p_corrcted=0.038; rs1436109, p_corrcted=0.012).

Table 1.

Association of NCAM1 Polymorphisms with Autism

				Minor allele frequency
SNP	Location in gene	Chromosome position	Minor/major alleles	Case	Control	Chi-square	p-Value	OR (95% CI)
rs4937786	5′UTR	112258317	C/A	0.414	0.377	5.863	0.015	1.17 (1.01–1.35)
rs12418058	Intron	112413264	G/A	0.131	0.161	7.103	0.0076	0.79 (0.65–0.94)
rs1436109	Intron	112496828	A/C	0.150	0.118	9.274	0.0023	1.32 (1.08–1.57)
rs584427	Coding	112609206	A/C	0.284	0.292	0.361	0.547	0.96 (0.80–1.09)
rs605843	Intron	112630444	G/A	0.325	0.340	0.963	0.326	0.94 (0.79–1.08)

SNP, single-nucleotide polymorphism; 5′UTR, 5′ untranslated region; OR, odds ratio; 95% CI, 95% confidence interval.

The pairwise LD showed that D′-value between rs12418058 and rs1436109 was 0.78 (Fig. 1). Then, we made a haplotype-based association test and found that two haplotypes constructed by rs12418058-rs1436109 indicated an association with autism (CC, p=0.0291, P_permutation=0.0719; TA, p=0.0035, P_permutation=0.0116; Table 2). The results suggested that the rs12418058-rs1436109 haplotype TA remained associated with autism, even after 10,000 times of permutation tests.

FIG. 1.

The linkage disequilibrium (LD) structure of the regions of five tag SNPs of NCAM1 gene, according to the Haploview (solid spine of LD, D′>0.7; www.broad.mit.edu/mpg/haploview/). Markers with LD (D′<1 and LOD>2) are shown in black through grey (color intensity decreases with decreasing D′ value). Regions of low LD and low LOD scores (D′<1 and LOD>2) are shown in white.

Table 2.

Haplotype-Based Association of RS12418058-RS1436109 in NCAM1 with Autism

	Minor allele frequency
Haplotype	Case	Control	Chi-square	p-Value
TC	0.722	0.729	0.213	0.6444
CC	0.128	0.153	4.761	0.0291
TA	0.146	0.115	8.513	0.0035

Moreover, by using the ANOVA tests, we found that two of the SNPs were associated with the parental age at conception (rs12418058, p=0.037; rs1436109, p=0.01; Table 3). Both the risk allele carriers (A of rs12418058 and A of rs1436109) showed association with the APA at conception. These findings suggested obvious dose-dependent effects of genotypes of rs12418058 and rs1436109 in the NCAM1 gene on the parental age at conception in autistic children. That means the advanced parental age at conception may also be one of the risk factors for autism.

Table 3.

Association of NCAM1 Polymorphisms with Parental Age at Conception in Autistic Children

SNP	Genotype	Parental age at conception (year)	F-value	p-Value
rs12418058	GG	26.44±4.91
	GA	32.68±6.12	3.32	0.037
	AA	33.25±6.59
rs1436109	CC	25.44±4.91
	CA	31.68±6.12	6.66	0.01
	AA	34.09±6.23

Discussion

The present study aimed to investigate a possible relationship between autism and the NCAM1 gene polymorphisms (rs4937786, rs12418058, rs1436109, rs584427, and rs605843). To our knowledge, this is the first study investigating these associations. According to our results, NCAM1 polymorphisms might be associated with autism, especially in the Chinese Han population. In the present study, we used the direct DNA sequencing for genotyping five tag SNPs (rs4937786, rs12418058, rs1436109, rs584427, and rs605843) at the NCAM1 gene. As a result, two of them remained significant after the Bonferroni multiple testing correction (rs12418058, p_corrcted=0.038; rs1436109, p_corrcted=0.012). The LD and haplotype analyses also suggested the association of autism and NCAM1 (rs12418058-rs1436109, haplotype CC, P_permutation=0.0116). Considering our relatively small sample size, these findings need to be replicated in much larger samples or other populations.

Autism has been described as a kind of neural development disorder that is associated with synaptic dysfunction. Genetic studies have reported several susceptibility genes of autism such as neurexins (NRXN1 and NRXN3, Yan et al., 2008), neuroligins (NLGN3 and NLGN4, Feng et al., 2006), cadherins (CDH8, CDH9, CDH10, CDH13, and CDH15, Wang et al., 2009; Sanders et al., 2011), contactins (CNTN4, CNTN5, and CNTN6, Cottrell et al., 2011; van Daalen et al., 2011), also known as the CAMs. Accumulating evidence has suggested some association between mutations in different CAM genes and autism. However, up to now, it remains unknown as to what is the potential functional implication of the CAMs on behavioral abnormalities in autistic children, such as impairments in social interaction and communication, and stereotypic or repetitive behaviors.

The NCAM, as one of the immunoglobulin superfamily members, have been reported to be engaged in multiple neurodevelopmental processes, including the neuron migration, axonal and/or dendritic projection, and synaptic targeting (Kenny et al., 2013; Minhas et al., 2013). The NCAM has been reported to accumulate in both the pre- and postsynaptic membranes and regulates synapse formation, maturation, and function through intra- and extracellular scaffold proteins (Bukalo and Dityatev, 2012). The presynaptic NCAM has been reported to be involved in the vesicle recycling in both neuronal and endocrine cells (Chan et al., 2005). The postsynaptic NCAM recruits the N-methyl-D-aspartate receptors (NMDARs) and Ca²⁺/calmodulin-dependent protein kinase II alpha (CaM-Kllalpha), and other postsynaptic density components for both synaptic formation and plasticity (Muller et al., 1996; Bukalo et al., 2004; Sytnyk et al., 2006). On the other hand, NCAM can also control axonal branching in GABAergic synapses of the interneurons (Chattopadhyaya et al., 2013).

Marui et al. (2009) have reported that the NRCAM, another gene coding the NCAM, might be associated with autism Overall, these findings suggested that the early neurodevelopmental process, including NCAM, might play an important role in the pathogenesis of autism.

Moreover, our study also found that two of the NCAM1 SNPs were associated with the parental age at conception in autism (rs12418058, p=0.037; rs1436109, p=0.01). On the other hand, the findings of Buizer-Voskamp et al. and others suggested that the level of global variation burden there is no influence of increased paternal age (Crow, 2003; Malaspina et al., 2005; Lopez-Castroman et al., 2010; Buizer-Voskamp et al., 2011; Hehir-Kwa et al., 2011; Krishnaswamy et al., 2011; Buizer-Voskamp et al., 2013). Recent studies have also reported the association between APA at conception and de novo in schizophrenia and autism (Buizer-Voskamp et al., 2011; Buizer-Voskamp et al., 2013; Gulsuner et al., 2013). Along with our findings, these studies suggested that some early neurodevelopmental processes might involve abnormalities, which may be related to the APA at conception. However, the potential mechanism should be explored in the future.

Nevertheless, two limitations might influence the results we obtained such as a small sample size and the heterogenous genetic backgrounds among different populations. Therefore, to control false-positive findings, many more samples should be genotyped and the replication in other populations should also be implemented in the future.

In summary, our findings have provided preliminary evidence for a significant association of NCAM1 polymorphisms with autism as well as paternal age at conception. Considering the important role of the NCAM1 gene in brain development, our results therefore indicated that the NCAM1 gene is a strong candidate gene for autism.

Acknowledgments

The authors thank all subjects who participated in this study and their colleagues for their assistance in recruiting patients in the study. This research was supported by research grants from the National Natural Science Foundation (grant numbers 81222017, 30870897).

Author Disclosure Statement

No competing financial interests exist.