Lithium partially rescues gene expression and enhancer activity from heterozygous knockout of AKAP11 while inducing novel differential changes [ChIP-Seq]

Bipolar disorder (BD) is a complex psychiatric condition usually requiring long-term treatment. Lithium (Li) remains the most effective mood stabilizer for BD, yet it benefits only a subset of patients, and its precise mechanism of action remains elusive. Exome sequencing has identified AKAP11 (A-kinase anchoring protein 11) as a shared risk gene for BD and schizophrenia (SCZ). Given that both the AKAP11-Protein Kinase A (PKA) complex and Li target and inhibit Glycogen Synthase Kinase-3 beta (GSK3ß), we hypothesize that Li may partially normalize the transcriptomic and/or epigenomic alterations observed in heterozygous AKAP11-knockout (Het-AKAP11-KO) iPSC-derived neurons. In this study, we employed genome-wide approaches to assess the effects of Li on the transcriptome and epigenome of human iPSC-derived Het-AKAP11-KO neuronal culture. We show that chronic Li treatment in this cellular model upregulates key pathways that were initially downregulated by Het-AKAP11-KO, several of which have also been reported as downregulated in synapses of BD and SCZ post-mortem brain tissues. Moreover, we demonstrated that Li treatment partially rescues certain transcriptomic alterations resulting from Het-AKAP11-KO, bringing them closer to the WT state. We suggest two possible mechanisms underlying these transcriptomic effects: (1) Li modulates histone H3K27ac levels at intergenic and intronic enhancers, influencing enhancer activity and transcription factor binding, and (2) Li enhances GSK3ß serine 9 phosphorylation, impacting WNT/ß-catenin signaling and downstream transcription. These findings underscore Li's potential as a therapeutic agent for BD and SCZ patients carrying AKAP11 loss-of-function variants or exhibiting similar pathway alterations to those observed in Het-AKAP11-KO models. Overall design: We performed H3K27ac profiling (ChIP-seq) on three AKAP11-KO clones and their three WT counterpart replicates, both untreated and treated with Li (chronic treatment, 11 days, 1mM), to analyze differential acetylation patterns. Specifically, we aimed to (1) identify differentially acetylated peaks in AKAP11-KO-Li compared to AKAP11-KO-UNT, (2) detect peaks in AKAP11-KO-UNT vs. WT-UNT that lost their differential status in AKAP11-KO-Li vs. WT-UNT, and (3) identify new differential peaks in AKAP11-KO-Li vs. WT-UNT that were not differential in AKAP11-KO-UNT vs. WT-UNT.