CNTNAP2 in neurodevelopmental disorders: Genetics, excitation/inhibition mechanisms and preclinical evidence

This review synthesizes recent genetic, molecular, cellular and preclinical evidence on CNTNAP2 function and dysfunction across the neurodevelopmental spectrum from highly penetrant recessive disorder to common autism spectrum disorder risk variation with the focus on human-specific expression patterns, mechanistic convergence on excitatory/inhibitory imbalance and potential translational targets. Biallelic CNTNAP2 loss-of-function defines CASPR2-deficiency disorder (CCD), a stereotyped syndrome of refractory epilepsy, profound language regression, intellectual disability, autistic features and cortical dysplasia that is distinct from the effects of heterozygous variants and common polymorphisms. Single-cell transcriptomics have revealed human-enriched CNTNAP2 expression in anterior frontal and temporal cortices, parvalbumin-positive interneurons and cortico-striato-thalamic circuits, a pattern that contrasts sharply with its rodent cortical expression. At the molecular level, CASPR2 scaffolds juxtaparanodal Kv1 channels through TAG-1/MAGUK interactions and stabilizes interneuron dendrites via CASK, its loss leads to interneuron depletion, impaired perisomatic inhibition and network hypersynchrony in both mouse and human iPSC models. Forebrain organoids derived from homozygous patients demonstrate potentially reversible cortical overgrowth driven by excessive PAX6-positive progenitor proliferation. In preclinical models, early intervention with oxytocin and AMPA receptor modulators has been shown to rescue social deficits and restore excitatory/inhibitory balance in knockout mice. In summary, CNTNAP2 exemplifies a nodal regulator linking axonal excitability, interneuron development and prefrontal network synchrony to a broad spectrum of neurodevelopmental phenotypes. Biallelic loss produces highly penetrant CCD, while common variation modifies autism and language risk. The human-specific expression patterns demand caution in rodent modeling, but circuit-level excitatory/inhibitory biomarkers and targeted interventions represent promising preclinical-stage clinical opportunities that require further validation in genotype-defined cohorts.