Cntnap2-KO Autism Model Mice Exhibit Perturbed Parvalbumin-Positive Interneuron Input and Output in Primary Visual Cortex V1

Background: Sensory processing dysfunction is a key feature and common challenge in autism. The refinement and maintenance of sensory processing depends on GABAergic inhibition. Studies in humans and animals suggest that altered excitatory/inhibitory balance or impaired inhibition could underlie the development of autism, and altering inhibitory circuitry can change the neuronal response to sensory stimulation. Low-level sensory processing alterations could exert larger downstream effects on higher-level behaviors relying sensory input. Changes in genes related to synaptic function and connectivity are risk factors for autism. Cntnap2 encodes CASPR2, a neurexin-family cell-adhesion molecule expressed throughout the brain. Cntnap2 knockout (KO) mice are an established model of autism that exhibit phenotypes including repetitive/stereotyped behavior, hyperactivity, reduced social interaction, seizures, synaptic spine instability, and changes in excitatory and inhibitory transmission, but the role of Cntnap2 in visual processing remains unknown.

Objectives: We sought to investigate the effects of Cntnap2 deletion on inhibitory circuitry in primary visual cortex, V1. Because we have previously demonstrated that Cntnap2 deletion leads to a developmentally-dependent reduction in tonic inhibitory currents and decreased spontaneous IPSC frequency in layer 2/3 pyramidal cells, we set out to investigate the impact of Cntnap2 deletion on the function of fast-spiking parvalbumin-positive (PV+) interneurons within V1.

Methods: We used whole-cell patch-clamp electrophysiology to record from L2/3 pyramidal cells and PV+ interneurons in slices made from V1 of Cntnap2-WT and -KO mice at 8 weeks of age. AAV injection was used to express fluorophores and/or ChR2 in PV+ interneurons in Cntnap2^-/-×PV^Cre mice, for identification and optogenetic stimulation.

Results: We have previously shown that L2/3 pyramidal cells from 8-week-old Cntnap2-KO mice exhibit lower spontaneous IPSC frequency and reduced tonic inhibitory conductance mediated by δ-subunit containing GABA_ARs. Stimulation of the feed-forward L4 to L2/3 pathway to evoke monosynaptic EPSCs and disynaptic IPSCs in pyramidal neurons resulted in a higher E/I ratio in KO mice compared to WT controls, indicating reduced inhibition. We did not observe an effect of genotype on intrinsic somatic properties of PV+ interneurons of V1, but optogenetic stimulation of PV+ terminals to evoke IPSC trains in pyramidal cells in L2/3 revealed increased paired-pulse depression at these inhibitory synapses. We again used stimulation of feed-forward L4 to L2/3 synapses to measure the NMDA/AMPA ratio in PV+ interneurons, and found that KO cells exhibited a significantly reduced NMDA/AMPA ratio.

Conclusions: Attenuated GABA_AR tonic currents and spontaneous IPSC frequency indicate fewer GABAergic synapses in Cntnap2-KO animals. Changes in evoked E/I balance and PV+ interneuron-generated IPSCs in V1 pyramidal cells suggests deficits in inhibition by fast-spiking interneurons. While PV+ interneuron properties were largely unaffected by the deletion of Cntnap2, the reduction of the NMDA/AMPA ratio suggests major deficits in neuronal information processing and integration in the cortical inhibitory circuit. Increased paired-pulse depression at PV-IN inhibitory synapses suggests higher initial GABA release and could indicate axonal dysfunction. Future studies should investigate the cellular/molecular mechanisms that bridge the gap between Cntnap2 mutation and inhibitory dysfunction, as well as the effects of diminished inhibition on the function of the V1 cortical circuit.