Zebrafish Mutants of the Autism Risk Gene Cntnap2 Identify Gabaergic Defects and Estrogens As Phenotypic Suppressors

Background: Gene discovery in autism has accelerated dramatically, yet the underlying mechanisms remain unknown, limiting the development of targeted treatments. The resulting rapidly accumulating pool of reliable autism genes is providing a launching point for the illumination of biological pathways and rational drug discovery. Homozygous loss-of-function mutations in Contactin Associated Protein-2 (CNTNAP2), which encodes a member of the neurexin family of cell adhesion molecules, are strongly linked to autism and epilepsy. Loss of Cntnap2 in mice revealed abnormal migration of projection neurons to upper cortical layers, reduced GABAergic neurons, spontaneous seizures, and behavioral abnormalities. CNTNAP2 localizes voltage-gated potassium channels at the juxtaparanodal region of myelinated axons, yet its role in the human central nervous system and the consequences of its loss for autism pathophysiology are less well understood. Currently, mammalian model systems offer limited opportunities for large-scale in vivo drug screening. In contrast, zebrafish is a model vertebrate system well-suited for small molecule screens. 

Objectives: Our goals are: 1) to investigate the function of Cntnap2 in nervous system development; 2) establish a model for conducting pharmacological screens; and 3) identify phenotypic suppressors and novel pathways with relevance to autism.

Methods: Because an imbalance in excitatory and inhibitory signaling in the CNS has been proposed as a mechanism underlying autism, we investigated excitatory and inhibitory neuronal populations in the brains of zebrafish cntnap2 mutants using transgenic lines labeling GABAergic and glutamatergic cells. Next, to determine the effect of loss of Cntnap2 on seizure susceptibility, we treated wild-type and mutant larvae with pentylenetetrazol, a GABA-A antagonist that is used to induce seizures. Further, we adapted an automated, high-throughput assay to quantify a series of behavioral parameters, including rest-wake cycle behaviors, in cntnap2 mutants over multiple days. By comparing the behavioral profiles of cntnap2 mutants and wild-type fish exposed to >500 psychoactive agents, we conducted correlation analyses to identify drugs that induce responses that strongly correlate or anti-correlate with the mutant behavioral phenotype. Further, we tested a select group of psychoactive agents to identify drugs that specifically reverse the mutant behavioral phenotype. 

Results: Zebrafish cntnap2 mutants display GABAergic deficits, particularly in the telencephalon and hypothalamus at 4 days post fertilization. In addition, cntnap2 mutants display hypersensitivity to drug-induced seizures, consistent with decreased inhibitory tone. Further, high-throughput behavioral profiling reveals a prominent phenotype of nighttime hyperactivity in cntnap2 mutants, while pharmacological screening reveals dysregulation of both GABAergic and glutamatergic systems. Specifically, we identify significant enrichment of NMDA antagonists among compounds that correlate with the mutant behavioral phenotype and 4 estrogenic agents among the top 10 anti-correlating drugs. Moreover, we find that cntnap2 mutants display differential behavioral responses to GABA agonists and increased sensitivity to behavioral activation by NMDA antagonists. Interestingly, biochanin A, a phytoestrogen identified in our unbiased screen, specifically reverses the mutant behavioral phenotype.

Conclusions: Zebrafish cntnap2 mutants identify GABAergic deficits and estrogenic compounds as phenotypic suppressors. These results highlight the utility of the zebrafish model as a genetic tool for illuminating novel phenotypic and pharmacological pathways relevant to autism.