Patient-derived hiPSC neurons with heterozygous CNTNAP2 deletions display altered neuronal gene expression and network activity

Variants in CNTNAP2, a member of the neurexin family of genes that function as cell adhesion molecules, have been associated with multiple neuropsychiatric conditions such as schizophrenia, autism spectrum disorder and intellectual disability; animal studies indicate a role for CNTNAP2 in axon guidance, dendritic arborization and synaptogenesis. We previously reprogrammed fibroblasts from a family trio consisting of two carriers of heterozygous intragenic CNTNAP2 deletions into human induced pluripotent stem cells (hiPSCs) and described decreased migration in the neural progenitor cells (NPCs) differentiated from the affected CNTNAP2 carrier in this trio. Here, we report the effect of this heterozygous intragenic deletion in CNTNAP2 on global gene expression and neuronal activity in the same cohort. Our findings suggest that heterozygous CNTNAP2 deletions affect genes involved in neuronal development and neuronal activity; however, these data reflect only one family trio and therefore more deletion carriers, with a variety of genetic backgrounds, will be needed to understand the molecular mechanisms underlying CNTNAP2 deletions.

Variants in CNTNAP2, a member of the neurexin family of genes that function as cell adhesion molecules, have been associated with multiple neuropsychiatric conditions such as schizophrenia, autism spectrum disorder and intellectual disability; animal studies indicate a role for CNTNAP2 in axon guidance, dendritic arborization and synaptogenesis. We previously reprogrammed fibroblasts from a family trio consisting of two carriers of heterozygous intragenic CNTNAP2 deletions into human induced pluripotent stem cells (hiPSCs) and described decreased migration in the neural progenitor cells (NPCs) differentiated from the affected CNTNAP2 carrier in this trio. Here, we report the effect of this heterozygous intragenic deletion in CNTNAP2 on global gene expression and neuronal activity in the same cohort. Our findings suggest that heterozygous CNTNAP2 deletions affect genes involved in neuronal development and neuronal activity; however, these data reflect only one family trio and therefore more deletion carriers, with a variety of genetic backgrounds, will be needed to understand the molecular mechanisms underlying CNTNAP2 deletions.

SHORT REPORT
The shared genetic architecture underlying neuropsychiatric disorders implicates common molecular mechanisms. 1 For example, while homozygous null mutations in CNTNAP2 lead to cortical dysplasia-focal epilepsy syndrome, 2, 3 heterozygous intragenic deletions are associated with schizophrenia, intellectual disability, language deficits, seizures, and autism traits. 4 Critically, CNTNAP2 variants are not completely penetrant. 2,5 Animal studies indicate a role for CNTNAP2 in axon guidance, dendritic arborization, and synaptogenesis. [6][7][8] We obtained fibroblast samples from a family trio with two carriers of heterozygous intragenic CNTNAP2 deletions, one affected and one unaffected, and an unaffected non-carrier control ( Table 1). The CNTNAP2 carriers display discordant clinical phenotypes; the daughter (DL7078) presented with schizoaffective disorder (depressed subtype) while the father (DL8735) was neurotypical. 9 We previously used sendai viral vectors to reprogram fibroblasts from this trio into hiPSCs that were then differentiated via dual-SMAD induction into NPCs and neurons. We characterized decreased migration in NPCs and allele-biased expression of the mutant CNTNAP2 transcript by qPCR in neurons from the affected CNTNAP2 carrier in this trio. 9 Here, we report the effect of this heterozygous intragenic deletion in CNTNAP2 on global gene expression and neuronal activity in this same cohort.
CNTNAP2 is highly expressed in Ngn2-induced neurons, a population of nearly pure excitatory neurons, 10 relative to hiPSCderived NPCs 11 and hiPSC-astrocytes 12 (qPCR FC = 82.5, p < 0.0001, ANOVA with Tukey's Post-Hoc) (Fig. 1a). RNA was harvested after 21 days of Ngn2-induction. The New York Genome Center prepared RNAseq libraries using the Kapa Total 350 bp kit, followed by 2 × 125 bp Illumina RNA sequencing to a read depth of 40 M reads per sample on the HiSeq 2500.
We queried the expression of CNTNAP2 in Ngn2-induced neurons from each member of this family trio, hypothesizing that heterozygous intragenic deletions may affect the expression of CNTNAP2. Surprisingly, overall CNTNAP2 expression was increased in the CNTNAP2 deletion carriers compared with the non-carrier mother (log 2 FC = 1.24, padj = 0.003) (Fig. 1b).
Differential expression analysis was performed using DESeq2 13 and the top 500 differentially expressed genes were used to perform gene ontology using DAVID 14,15 (SI Table 1). The most significant subset of genes mapped to terms relating to DNA binding and central nervous system (CNS) development (FC = 1.8, p < 0.00001 and FC = 1.9, p = 0.0003) ( Fig. 1c; SI Table 2). Within the gene subset involved in CNS development, there are some interesting candidate genes such as CNTN6 and CNTN4, which are involved in regulating cell surface interactions during nervous system development and are also thought to be important in synaptogenesis (SI Table 3).
Given the differences in gene expression of critical neuronal and synaptic genes, we applied an Axion multi-electrode array (MEA) (see similar applications to Amyotrophic Lateral Sclerosis 16 and Parkinson's disease 17 ) to record population-wide neuronal activity under conditions similar to those used in our RNAseq analyzes. 21-day-old Ngn2-induced neurons from both the affected (daughter, DL7078) and unaffected (father, DL8735) CNTNAP2 deletion carriers showed significantly increased spontaneous network level activity (an increase of 210 and 253%, respectively) relative to the non-carrier (mother, DL5535) and an unaffected unrelated control (female, NSB3113) (N = 12 wells/ condition; p < 0.001, ANOVA with Tukey's Post-Hoc) (Fig. 2a). These results were confirmed in hiPSC-derived forebrain neuron populations, which are comprised of a mixture of mature and immature excitatory neurons, inhibitory neurons and astrocytes, with the majority of cells presumed to be excitatory neurons. 18,19 Here, NPCs were seeded (65,000 cells/mL) onto 12-well MEA plates and differentiated for 28 days; hiPSC-derived neurons from the affected and unaffected CNTNAP2 deletion carriers showed significantly increased spontaneous population wide neuronal activity relative to the non-carrier mother (increases of 344 and 182% relative to the non-carrier mother; N = 12 wells/condition; p < 0.001, ANOVA with Tukey's Post-Hoc) (Fig. 2a, b). All measurements were performed as biological triplicates for each hiPSC line and averaged across experiments. In both populations, neurons robustly stained for neuronal markers βIII-TUBULIN, NeuN, and MAP2 (representative images in Fig. 2c).
Our genetic analyzes and functional assays together show that heterozygous deletion of CNTNAP2 may impact neuronal activity. The significant increase in spontaneous spiking activity in the unaffected carrier father and carrier daughter may underlie aspects of the aberrant behavior displayed by the proband. Additionally, this alteration in spike activity may in part explain observations of disrupted neuronal synchrony in CNTNAP2-null mice. 6 Here we demonstrated that hiPSC-derived neurons from individuals with heterozygous intragenic deletions in CNTNAP2 display differential expression of genes involved in synaptic transmission and altered neuronal activity, consistent with reports of disrupted cortical neuronal activity in CNTNAP2-null mice, 6 and potentially independent of clinical outcome. Our report reflects results from just one family trio; a greater variety of disease-associated CNTNAP2 mutations, on an array of genetic backgrounds, will be needed to understand the full breath of genotype-phenotype relationships with respect to CNTNAP2. 18,[20][21][22]