Fine-mapping analysis revealed complex pleiotropic effect and tissue-specific regulatory mechanism of TNFSF15 in primary biliary cholangitis, Crohn's disease and leprosy.

Yonghu Sun,Kyuyoung Song,Fangfang Bao,Astrid Irwanto,Katsushi Tokunaga,Yuki Hitomi,Gongqi Yu,Hyunchul Choi,Huimin Zhang,Chuan Wang,Masao Nagasaki,Myunghee Hong,Honglei Wang,Licht Toyo-Oka,Yosua Denise,Byong Duk Ye,Furen Zhang,Minoru Nakamura,Anand Kumar Andiappan,Hong Liu,M Kawashima ,Xi’an Fu ,Olaf Rötzschke ,Zhenming Yue ,Kazuyuki Kojima ,Suk Kyun Yang ,Yang Yu ,Jing Liu

doi:10.1038/srep31429

Abstract

Genetic polymorphism within the 9q32 locus is linked with increased risk of several diseases, including Crohn’s disease (CD), primary biliary cholangitis (PBC) and leprosy. The most likely disease-causing gene within 9q32 is TNFSF15, which encodes the pro-inflammatory cytokine TNF super-family member 15, but it was unknown whether these disparate diseases were associated with the same genetic variance in 9q32, and how variance within this locus might contribute to pathology. Using genetic data from published studies on CD, PBC and leprosy we revealed that bearing a T allele at rs6478108/rs6478109 (r2 = 1) or rs4979462 was significantly associated with increased risk of CD and decreased risk of leprosy, while the T allele at rs4979462 was associated with significantly increased risk of PBC. In vitro analyses showed that the rs6478109 genotype significantly affected TNFSF15 expression in cells from whole blood of controls, while functional annotation using publicly-available data revealed the broad cell type/tissue-specific regulatory potential of variance at rs6478109 or rs4979462. In summary, we provide evidence that variance within TNFSF15 has the potential to affect cytokine expression across a range of tissues and thereby contribute to protection from infectious diseases such as leprosy, while increasing the risk of immune-mediated diseases including CD and PBC.

Highlights

Previous genome-wide association studies (GWAS) have revealed multiple associations of diverse disease phenotypes on chromosome 9q32, including primary biliary cholangitis(PBC)[9], Crohn’s disease (CD)[10,11,12,13], and leprosy[14]
We identified and ranked single-nucleotide polymorphisms (SNP) associated with each disease and used in silico functional annotation combined with in vitro experimentation to show that the diverse disease associations within TNFSF15 are probably driven by two independent causal variants that contribute to regulating TNFSF15 expression across a range of tissues and cell types
The most significantly associated variants were different SNPs in each case (Figure S1): for Japanese PBC, the most significant SNP was rs55768522, which is in perfect linkage disequilibrium (LD) (r2 = 1) with the previously reported SNP, rs49794629; for Korean CD, the most significant SNP was rs6478108, which is in perfect LD with the previously reported rs6478109; while for leprosy, the most significant SNP is rs4366152, which was in high LD with the previously reported SNP, rs6478108 (r2 ≥ 0.97, Figure S2)

Summary

Introduction

Previous genome-wide association studies (GWAS) have revealed multiple associations of diverse disease phenotypes on chromosome 9q32, including primary biliary cholangitis (previously known as primary biliary cirrhosis1–8)(PBC)[9], Crohn’s disease (CD)[10,11,12,13], and leprosy[14]. While PBC, CD and leprosy share some pathological features, such as granuloma formation[24], the different effects of variance at 9q32 upon risk of these diseases, and the discovery of these risk variants in specific populations, mean that important questions on the overall role of 9q32 polymorphisms, those affecting TNFSF15, remain unanswered. To resolve this situation we performed an integrated fine-mapping study of the TNFSF15 locus using published genetic data from PBC patients and controls in a Japanese population, CD patients and controls in a Korean population, and leprosy patients and controls in a Chinese population. We identified and ranked single-nucleotide polymorphisms (SNP) associated with each disease and used in silico functional annotation combined with in vitro experimentation to show that the diverse disease associations within TNFSF15 are probably driven by two independent causal variants that contribute to regulating TNFSF15 expression across a range of tissues and cell types

Methods

Results

Conclusion