Chronic subanesthetic ketamine induces schizophrenic-like social deficits and alters translatome in the lateral septum

Schizophrenia is often marked by poor social functioning that can have a severe impact on quality of life and independence, but the underlying neural circuity is not well understood. Here we used a translational model of subanesthetic ketamine mice to delineate neural pathways in the brain linked to social deficits in schizophrenia. Mice treated with chronic ketamine exhibit profound social and sensorimotor deficits as previously reported. Using three-dimensional c-Fos immunolabeling and volume imaging (iDISCO), we show that ketamine treatment resulted in hypoactivation of the lateral septum (LS) in response to social stimuli. Chemogenetic activation of the LS rescued social deficits after ketamine treatment, while chemogenetic inhibition of previously active populations in the LS (i.e. social engram neurons) recapitulated social deficits in ketamine-naïve mice. We then examined the translatome of LS engram neurons and found upregulation of genes encoding potassium inwardly-rectifying channels and solute transporters as well as dysregulation of genes implicated in apoptotic processes, all of which could contribute to the hypoactivation of LS social engram neurons after chronic ketamine exposure. A number of ketamine-induced differentially expressed genes (DEGs), including those involved in mitochondrial function and neuroinflammatory pathways, were shared with human schizophrenia and mouse models of neurodevelopmental disorders. To isolate differentially translating mRNAs from social interaction-activated neurons in the lateral septum (LS), we employed the ribo-TRAP approach, which allowed us to selectively isolate translating mRNAs in activated neurons using Cre-dependent expression of EGFP-tagged L10a ribosomal subunits in Fos-iCreER mice. After TRAPing activated neurons, we administered saline or ketamine (30 mg/kg/day, i.p.) for 10 days. Because the expected number of activated neurons was relatively low, we pooled five LS tissue punches into one tube. Consequently, we processed 10 ketamine and 10 saline LS tissue punches to create 2 ketamine and 2 saline pooled samples, respectively, for ribo-TRAP.