Genome-wide association study of widespread pain links LRRC3B as possible novel pain associated gene

Abstract

Purpose: The hallmark clinical symptom of clinical osteoarthritis (OA) is pain, which is one of the leading causes of disability in OA. Pain not only has a physical effect on the individual, but can also impact the individual mentally and socially. Despite this high burden, there is not a lot known on the role of genetics in musculoskeletal pain. In this study, we aimed to investigate the genetic determinants of muscoskeletal pain and the way we perceive it. In order to examine this, we have performed a genome-wide association study (GWAS) using the number of self-reported pain sites, as a proxy for musculoskeletal pain. Methods: We have included 8,930 individuals (57.3% women) from the Rotterdam Study (RS) cohorts, RSI, RSII and RSII, with self-reported pain sites and genetic information available. Participants could indicate on 15 body sites if they were experiencing pain on these sites. Resulting in a score from 0-no pain at any site to 15-pain at every site. This score was used as outcome in the GWAS analysis. Genotype reference was imputed to HRC v1.1, GWAS analysis were adjusted for age and sex, in addition a sex stratified analysis was also performed. As the number of pain sites score did not have a normal distribution, top signals from the GWAS were additionally re-analyzed using a Poisson regression model. Top signals were followed up using gene expression information from GTEx database. Results: The meta-analysis GWAS results of the Rotterdam study cohorts, did not show any genome-wide significant or suggestive associations (p<1E-07) with number of pain sites. However the mean number of pain sites in women is 1.58 (SD=2.34), which is significantly higher than in men (0.84 SD=1.56, p<1E-04). Thus we performed sex stratified analysis. In the female stratified GWAS one genome-wide significant signal was associated with number of pain sites, also after Poisson regression analysis (Beta= 0.50, ppoisson=4.3*10-08) and was not associated with number of pain sites in men (beta=0.093, p=0.60). The lead variant rs1488240, with risk allele A has a MAF of 9% and is a novel pain associated variant. This variant is located near the Leucine-Rich Repeat Containing 3B (LRRC3B) gene and has a variant in high linkage disequilibrium which is a coding variant in LRRC3B ((rs35497952, r2 = 0.83, synonymous). GTEx expression data shows high expression of LRRC3B in multiple nervous system tissues. The strongest gene expression was reported in basal ganglia(n=194) and the cerebellum (n=209). No significant or suggestive signals were found in the male stratified analysis. In the female stratified analysis next to our one novel pain associated signal, we were also able to replicate 42 known pain associated variants(p<0.05). Conclusions: We have identified one novel pain associated variant, a synonymous variant in the LRRC3B gene, and nominally replicate 42 known pain associated variants. LRRC3B is highly expressed in the cerebellum and basal ganglia of the brain, which are both known to be involved in pain processing and modulation. In addition, LRRC3B has protein interactions and co-expression with IGDCC3 which is linked to functions in axon guidance, and is highly expressed in mouse nervous systems and limb development. Suggesting LRRC3B might play a role in chronic musculoskeletal pain.