1120: Transcriptome-Wide Association Study of Sjögren’s Disease Risk Alleles Identifies Novel Genes with Altered Expression in Minor Salivary Gland and Other Tissues

Marcin Radziszewski¹, Mandi Wiley¹, Bhuwan Khatri¹, Kandice Tessneer¹, Astrid Rasmussen¹, Professor Simon Bowman², Lida Radfar³, Roald Omdal⁴, Marie Wahren-Herlenius⁵, Blake Warner⁶, Torsten Witte⁷, Roland Jonsson⁸, Maureen Rischmueller⁹, Patrick Gaffney¹, Judith James¹, Lars Ronnblom¹⁰, R. Hal Scofield³, Xavier Mariette¹¹, Marta Alarcon-Riquelme¹², Fai Ng¹³, Gunnel Nordmark¹⁰, Umesh Deshmukh¹, A. Darise Farris¹ and Christopher Lessard¹, ¹Oklahoma Medical Research Foundation, Oklahoma City, OK, ²University Hospitals Birmingham, Birmingham, United Kingdom, ³University of Oklahoma Health Sciences Center, Oklahoma City, OK, ⁴Stavanger University, Stavanger, Norway, ⁵Karolinska Institutet, Stockholm, Sweden, ⁶National Institutes of Health, Bethesda, MD, ⁷MH-Hannover, Hannover, Germany, ⁸University of Bergen, Bergen, Norway, ⁹RheumatologySA, Adelaide, Australia, ¹⁰Uppsala University, Uppsala, Sweden, ¹¹Paris-Saclay University, Rueil Malmaison, Ile-de-France, France, ¹²Center for Genomics and Oncological Research (GENYO), Granada, Spain, ¹³Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, United Kingdom

Background/Purpose: Sjögren's disease (SjD) is an autoimmune disease characterized by reduced function of exocrine glands, but also has systemic manifestations affecting multiple organs, including abnormal pulmonary or liver functions in 75% or 50% of patients, respectively [1,2]. Genetic and transcriptional changes influence epithelial-immune cell interactions in SjD pathogenesis, but mechanisms remain understudied.

Methods: SNPs from 16 regions with SNP-SjD associations (P< 5x10^-8) in our genome-wide association studies (GWAS) [3,4] were interrogated for expression quantitative trait loci (eQTLs) in Genotype-Tissue Expression (GTEx) minor salivary gland (MSG) data. Genes that were identified as eQTLs in the MSG and showed differential expression or accessibility in RNA-seq and ATAC-seq, respectively, from the submaxillary salivary gland epithelial cell line, A253, were analyzed by pathway enrichment analysis using gProfiler. Transcriptome-wide association study (TWAS) was performed using GWAS summary statistics and MSG, liver, and lung eQTL GTEx data.

Results: In total, 5884 genome-wide significant (GWS) SNPs spanning 10 risk loci were identified as MSG eQTLs using two discovery thresholds: p(FDR)< 0.05 provided by eQTL study (3566 SNPs) and p(FDR) >0.05 and p< 0.05 in eQTL study (2318 SNPs). MSG eQTLs for 155 unique genes exhibited coalescence of RNA- and ATAC-seq data in A253 cells. Many SNPs altered expression of the nearest gene, such as IRF5 and TNPO3 on chromosome 7 at 128Mb. This locus had 12 additional genes that were eQTLs in MSG. In contrast, other loci had no reported eQTLs for the nearest gene, but several reported eQTLs for distant loci, such as TYK2 on chromosome 19 at 10Mb that showed no change in TYK2 expression but eQTLs for 8 distant genes, including ICAM1. Pathway enrichment analysis revealed Butyrophilin (BTN) family interactions (P_Adj=1.564x10^-5), including the BTN2A1, BTN2A2, BTN3A1, BTN3A2 and BTN3A3 gene cluster in the MHC region. TWAS of MSG and SjD GWAS summary statistics (after Bonferroni correction) showed association between SjD and BTN3A2 (P_Adj=1.19x10^-50), among many loci in the MHC region. Several long non-coding RNAs (lncRNAs) were also significant. LINC02210 (P_Adj=1.21x10^-6) on chromosome 17 exhibited opposing z-scores in the TWAS of liver (6.02) relative to lung (-5.99) and MSG (-6.27).

Conclusion: SjD risk alleles influence disease by altering tissue-specific gene expression. Interestingly, we observed MSG eQTLs for several BTN family genes, which act as cell-surface binding partners to regulate cell-cell interactions, such as those between epithelial cells and activated T cells [5]. Further, tissue-specific changes in lncRNA expression suggest that these uncharacterized transcripts may play important pathogenic roles. Additional studies in MSG and other affected tissues are needed for further insights into SjD pathology.

References:
1. Verstappen. Nat Rev Rheumatol 2021;17(6):333-348.
2. Negrini et al. Clin Exp Med. 2022; 22(1): 9–25.
3. Lessard, et al. Nat Genet 2013 Nov;45(11):1284-92.
4. Khatri, et al. Annals of Rheumatic Diseases 2020;79:30-31.
5. Arnett HA, Viney JL. Nature Reviews Immunology 2014;14:559-569.

Disclosures: M. Radziszewski, None; M. Wiley, None; B. Khatri, None; K. Tessneer, None; A. Rasmussen, None; P. Bowman, Novartis, AstraZeneca, AbbVie/Abbott, Galapagos; L. Radfar, None; R. Omdal, None; M. Wahren-Herlenius, None; B. Warner, Pfizer, Astellas Bio; T. Witte, AbbVie, Amgen, Bristol-Myers Squibb, Celgene, Chugai, Janssen, Lilly, MSD, Mylan, Novartis, Pfizer, Roche, UCB, Galapagos; R. Jonsson, None; M. Rischmueller, None; P. Gaffney, None; J. James, Bristol-Myers Squibb(BMS), AstraZeneca, Novartis, Progentec Biosciences; L. Ronnblom, None; R. Scofield, None; X. Mariette, AstraZeneca, Bristol Myers Squibb, Galapagos, GSK, Novartis, Pfizer; M. Alarcon-Riquelme, None; F. Ng, None; G. Nordmark, None; U. Deshmukh, None; A. Farris, Janssen; C. Lessard, Janssen.