1734: Identification of Sjögren’s Disease-Associated T Cell Receptors Through Deep Sequencing and Single-Cell Transcriptomics

Ananth Aditya Jupudi¹, Michelle Joachims¹, Christina Lawrence¹, Charmaine Lopez-Davis¹, Bhuwan Khatri¹, Astrid Rasmussen¹, Lida Radfar², Kiely Grundahl¹, R. Hal Scofield², Judith James¹, Joel Guthridge¹, Christopher Lessard¹, Linda F. Thompson¹ and A. Darise Farris¹, ¹Oklahoma Medical Research Foundation, Oklahoma City, OK, ²University of Oklahoma Health Sciences Center, Oklahoma City, OK

Background/Purpose: Sjögren's disease (SjD) is a chronic rheumatic autoimmune disorder that primarily targets the lacrimal and salivary glands (SG) resulting in dry eyes and dry mouth. The SG tissue of SjD patients is characterized by the presence of hallmark focal lymphocytic infiltrates predominated by CD4⁺ T cells. This implies a pathogenic role for these cells which is further corroborated by the association of SjD with HLA DR3/DQ2. In this study, we aim to identify T cell receptors (TCRs) with shared antigen-specificity in SjD cases.

Methods: Peripheral blood (PB) samples were collected from 19 cases (16 DR3⁺DQ2⁺, 2 DR3^–/DQ2^–, 1 DR3⁺DQ2^–) meeting the 2016 ACR/EULAR classification criteria for SjD and 19 DR3⁺DQ2⁺ matched healthy controls (HC). CD3⁺CD4⁺CD45RA^– antigen-experienced T cells were bulk-sorted from subjects followed by RNA extraction and cDNA synthesis using the SMARTSeq system. TCRβ libraries were generated for TCR deep sequencing. We implemented the 'Molecular Identifier Guided Error Correction pipeline' (MIGEC) to extract CDR3 sequences, V/J segment usage, and clonotype information from raw sequencing data. The incorporation of unique molecular identifiers (UMIs) into cDNA allowed for the correction of amplification bias and sequencing errors. In addition to the PB data, we included paired TCR α and β sequences derived from the SG of 20 SjD cases (n=10 by TCR RT-PCR, n=10 by single-cell RNAseq) for analysis. We prepared a curated meta-dataset from all available TCRs to identify those with shared antigen specificity using the 'GLIPH2' (Grouping of Lymphocyte Interactions by Paratope Hotspots) algorithm.

Results: PB-TCR deep sequencing data from 12 SjD cases and 12 HC passed quality control. We analyzed 1,877,614 CDR3β chain sequences using GLIPH2 and identified 55,419 TCR-clusters with shared antigen-specificity found exclusively in the patient TCR repertoire and not in HCs. Of these, 215 clusters contained TCRs from ≥3 SjD cases that shared amino acid sequence similarity and had significant GLIPH2 scores for clonal expansion, cluster size, Vb gene enrichment, and conserved CDR3 length. Furthermore, we discovered that 6 clusters with significant GLIPH2 scores for the same features contained TCRs that were commonly enriched in both the PB and SG tissue of patients. Notably, 3 TCRs were found to be enriched both in the blood and SG of the same patient. One of these was part of a clonal expansion in the SG, over-represented in PB, and expressed dual TCRα chains.

Conclusion: This study presents evidence of clonally-expanded T cells containing dual TCRα chains in the affected SG tissue and peripheral blood of the same patient. Such TCRs have been proposed to contribute to autoimmunity in several studies. We discovered an SjD-associated TCR repertoire shared between the SG and PB across multiple patients. Furthermore, we found disease-associated TCRs enriched in the PB of SjD cases. These data corroborate the systemic nature of SjD and enable the discovery of shared SjD T cell epitopes.

Disclosures: A. Jupudi, None; M. Joachims, None; C. Lawrence, None; C. Lopez-Davis, None; B. Khatri, None; A. Rasmussen, None; L. Radfar, None; K. Grundahl, None; R. Scofield, None; J. James, Bristol-Myers Squibb(BMS), AstraZeneca, Novartis, Progentec Biosciences; J. Guthridge, None; C. Lessard, Janssen; L. Thompson, None; A. Farris, Janssen.