LAUSR

Abstract Identifying clinical and genetic risk factors associated with antidepressant-induced mania (AIM) may improve individualized treatment strategies for bipolar depression. Several small studies have investigated the association of variation in the serotonin transporter gene (SLC6A4) with AIM. Here we extend these studies using a new sample from the Mayo Clinic Bipolar Disorder Biobank, and perform the first genome wide association study (GWAS) of AIM. Bipolar patients, confirmed by structured interview were retrospectively screened for AIM. AIM cases reported a manic/hypomanic episode within 60 days of starting or changing dose of antidepressant, while control had an adequate (≥60 days) exposure to an antidepressant with no associated manic/hypomanic episode. The sample with genome-wide genotyping included 525 Caucasian bipolar patients (174 AIM cases and 351 controls). No genome-wide significant associations were observed in our GWAS of AIM. The SNP with strongest evidence of association (rs10093489; p=5.79E-07) maps to an intron of ZMAT4. Another top signal represents a cluster of SNPs that map to the DVL1 (Dishevelled -1) gene region (top SNP rs307364; P = 1.25E-06). A set of candidate variants selected based on association with treatment response in prior antidepressant pharmacogenomics studies in major depressive disorder did not show evidence of association in the AIM GWAS, suggesting that different variants may contribute to antidepressant response in unipolar and bipolar depression. Ongoing analyses using polygenic methods are being used to further investigate this question. The potential role of DVL1 warrants further investigation, as the gene encodes a cytoplasmic scaffolding protein involved in Wnt signaling, which has been implicated in antidepressant effects in rodents. Further work is encouraged to better delineate the potential role of Wnt signaling in antidepressant response in mood disorders.