LAUSR

Abstract It is well known that the risk of opioid dependence (OD) is genetically influenced. Several GWAS, including ours, have succeeded at identifying common risk variants for OD. We undertook the present whole exome sequence (WES) study to identify associations for OD risk with variants not readily genotypable by microarray. So far, >1600 WES have been completed: >1100 European Americans (EAs) and >500 African Americans (AAs). Using the WES data, we conducted single variant (score) and gene-based (SKAT-O) tests in these subjects, with DSM-IV OD diagnosis and DSM-IV criterion count as outcomes. After Bonferroni correction for the number of single SNP tests, one variant was significantly associated with both outcomes in EAs in the gene NRXN3. NRXN3 is of great biological interest due to its role in neuronal synapse function and has been associated with risk for both alcohol and opioid dependence. The MAF of NRXN3 rs368371631 was 2.7% overall (~10% in exposed controls and 0.4% in cases). It was not identified as a risk marker in our GWAS studies because it is not on any of the imputation reference panels available; it is in a low coverage area and only appears in one of the two primary sequencing databases. One variant in DMPK was significantly associated with OD in AAs. In addition to being the putative gene for myotonic dystrophy, DMPK is thought to play a role in synaptic plasticity. Gene-based tests identified an additional seven protein-coding genes that were significantly associated, including NLGN2 and PRDM12, which are promising biological candidates. NLGN2 has a role in reinforced activity-dependent inhibitory, but not excitatory, synaptic responses. PRDM12 has been shown to control sensory neuron development and pain perception in a Xenopus model and to be essential to pain perception in humans; variation in this locus can cause clinical insensitivity to pain. However, to our knowledge this gene has not previously been implicated in the risk of OD or any other substance use disorder. Our preliminary results support the utility of WES in an OD case/control study, most notably implicating PRDM12 in a gene-based test.