Genomics of severe and treatment-resistant obsessive-compulsive disorder treated with deep brain stimulation: a preliminary investigation

Abstract

Individuals with severe and treatment-resistant obsessive-compulsive disorder (trOCD) represent a small but severely disabled group of patients. Since trOCD cases eligible for deep brain stimulation (DBS) probably comprise the most severe end of the OCD spectrum, we hypothesize that they may be more likely to have a strong genetic contribution to their disorder. Therefore, while the worldwide population of DBS-treated cases may be small (~300), screening these individuals with modern genomic methods may accelerate gene discovery in OCD. As such, we have begun to collect DNA from trOCD cases who qualify for DBS, and here we report results from whole exome sequencing and microarray genotyping of our first five cases. All participants had previously received DBS in the bed nucleus of stria terminalis (BNST), with two patients responding to the surgery and one showing a partial response. Our analyses focused on gene-disruptive rare variants (GDRVs; rare, predicted-deleterious single-nucleotide variants or copy number variants overlapping protein-coding genes). Three of the five cases carried a GDRV, including a missense variant in the ion transporter domain of KCNB1, a deletion at 15q11.2, and a duplication at 15q26.1. The KCNB1 variant (hg19 chr20-47991077-C-T, NM_004975.3:c.1020G>A, p.Met340Ile) causes substitution of methionine for isoleucine in the trans-membrane region of neuronal potassium voltage-gated ion channel KV2.1. This KCNB1 substitution (Met340Ile) is located in a highly constrained region of the protein where other rare missense variants have previously been associated with neurodevelopmental disorders. The patient carrying the Met340Ile variant responded to DBS, which suggests that genetic factors could potentially be predictors of treatment response in DBS for OCD. In sum, we have established a protocol for recruiting and genomically characterizing trOCD cases. Preliminary results suggest that this will be an informative strategy for finding risk genes in OCD.

Keywords: OCD; Obsessive-compulsive disorder; deep brain stimulation; genetic; treatment-resistant.

Figure 1.

A KCNB1 missense variant identified in trOCD case #2 colocalizes with known pathogenic missense variants. KCNB1 encodes the potassium voltage-gated channel subfamily B member 1 and has six transmembrane domains. Shown in green are amino acids changed by missense variants classified as pathogenic or likely pathogenic in ClinVar. Shown in red is the variant identified in this study (hg19 chr20-47991077-C-T, NM_004975.3:c.1020G>A, p.Met340Ile). Like most of the pathogenic or likely pathogenic variants, M340I localizes to a transmembrane domain that helps form the pore through which potassium passes.

Figure 2.

Gene-disruptive rare CNVs called in exome sequence data from trOCD cases #3 (A) and #4 (B). The red vertical dashes indicate the boundaries of the CNV call. Patient #3 has a previously described 768 kb neurodevelopmental deletion in 15q11.2 that disrupts 4 genes. Patient #4 has a novel 287 kb duplication in 15q26.1 that disrupts 5 genes.