Broader anti-EBV TCR repertoire in multiple sclerosis: disease specificity and treatment modulation

Abstract

Epstein-Barr virus (EBV) infection has long been associated with the development of multiple sclerosis (MS). Patients with MS have elevated titres of EBV-specific antibodies in serum and show signs of CNS damage only after EBV infection. Regarding CD8+ T cells, an elevated but ineffective response to EBV was suggested in MS patients, who present with a broader MHC-I-restricted EBV-specific T-cell receptor beta chain (TRB) repertoire compared to controls. It is not known whether this altered EBV response could be subject to dynamic changes, e.g. by approved MS therapies, and whether it is specific for MS. Peripheral blood TRB repertoire samples (n = 1317) of healthy donors (n = 409), patients with MS (n = 710) before and after treatment, patients with neuromyelitis optica spectrum disorder (n = 87), MOG antibody-associated disease (MOGAD) (n = 64) and Susac's syndrome (n = 47) were analysed. Apart from MS, none of the evaluated diseases presented with a broader anti-EBV TRB repertoire. In MS patients undergoing autologous haematopoietic stem-cell transplantation, EBV reactivation coincided with elevated MHC-I-restricted EBV-specific TRB sequence matches. Therapy with ocrelizumab, teriflunomide or dimethyl fumarate reduced EBV-specific, but not CMV-specific MHC-I-restricted TRB sequence matches. Together, these data suggest that the aberrant MHC-I-restricted T-cell response directed against EBV is specific to MS with regard to neuromyelitis optica, MOGAD and Susac's syndrome and that it is specifically modified by MS treatments interfering with EBV host cells or activated lymphocytes.

Keywords: Epstein-Barr virus; Susac’s syndrome; T-cell receptor; multiple sclerosis; myelin-oligodendrocyte-glycoprotein antibody-associated disease; neuromyelitis optica spectrum disorder.

Figure 1

Broader EBV-specific TRB repertoire is MS-specific with regard to differential diagnoses. HLA-matched comparisons of EBV-specific TRB sequences from patients with NMOSD (A), MOGAD (B) or SuS (C). Each patient group matched 3:1 with healthy donors (HD) and 1:1 with multiple sclerosis (MS) patients. This resulted in a group comparison of 87 neuromyelitis optica spectrum disorder (NMOSD) patients versus 87 MS patients and 261 healthy donors (A), 64 MOG antibody-associated disease (MOGAD) patients versus 64 MS patients and 192 healthy donors (B), and 47 Susac syndrome (SuS) patients versus 47 MS patients and 141 healthy donors (C). Shown is the number of EBV-specific TRB sequence matches in a sample, as well as the level of significance for the cohort differences between healthy donors versus MS and healthy donors versus NMOSD/MOGAD/SuS from linear models adjusted for age, sex, sequencing depth, current therapy and HLA. The boxes indicate the 25% and 75% percentile with a line at the median, the whiskers extend from the hinge to the largest/smallest value no further than 1.5× interquartile range from the hinge. *P < 0.05, **P < 0.01 or ***P < 0.001. An overview of the patient groups including HLA backgrounds is given in Supplementary Table 1. EBV = Epstein-Barr virus; n.s. = not significant; TRB = T-cell receptor beta chain.

Figure 2

EBV-specific TRB sequences spike in patients undergoing AHSCT. Two independent, previously published cohorts (A, n = 24 in Muraro et al. and B, n = 13 in Massey et al.) of multiple sclerosis (MS) patients before autologous haematopoietic stem-cell transplantation (AHSCT) and 2, 6, 12, 24 and 36 months after AHSCT were analysed. The number of Epstein-Barr virus (EBV)-specific T-cell receptor beta chain (TRB) sequence matches per 10⁴ unique TRB sequences are shown. Red dots and boxes in A indicate MS patients with EBV reactivation or reinfection according to the primary publication. Orange dots and boxes in B indicate MS patients with high EBV viral load in the period of observation. The significance level (***P < 0.0001) in A or (*P < 0.05; **P < 0.01) indicates the difference in counts of EBV-specific TRB sequence matches in MS patients at the indicated time point with a linear mixed model and the covariates age, sex, therapy (AHSCT after baseline), time point, EBV infection status and a random intercept for each study participant. Black and red/orange dots represent individual patient samples. The boxes indicate the 25% and 75% percentile with a line at the median, the whiskers extend from the hinge to the largest/smallest value no further than 1.5× interquartile range from the hinge. Grey lines connect the same individual.

Figure 3

Treatment with ocrelizumab, teriflunomide and dimethyl fumarate reduces EBV-specific but not CMV-specific TRB sequence matches. (A) Epstein-Barr virus (EBV)-specific T-cell receptor beta chain (TRB) sequences of healthy donors (n = 16) and multiple sclerosis (MS) patients (other panels, n = 198) are shown at multiple time points. MS patients either received no treatment (untreated MS) (n = 24) or one of glatiramer acetate (n = 30), interferon beta (n = 65), ocrelizumab [investigator-initiated trial (IIT) cohort: n = 27; ENSEMBLE cohort: n = 17], teriflunomide (n = 27) or dimethyl fumarate (n = 8). The time points correspond to pretreatment (Time = 0) and ON treatment (Time = 1, 2, 3, 6, 12 and 24 months of continuous treatment). A linear mixed model on the presence of EBV-specific TRB sequence matches was developed for the data in A. Covariates of the linear mixed model included age, sex, therapy (after baseline), EBV-specific TRB sequences at baseline, MS status (MS patient or healthy donors), cell type [CD8⁺ T cell or peripheral blood mononuclear cell (PBMC)], sequencing depth, study, all relevant HLAs from the sequence database, and a random intercept for each study participant. (B) A selection of the effect size of each of the covariates in the linear mixed model of A. Asterisks indicate adjusted q-values: *q < 0.05 or **q < 0.01. Grey lines connect the same individual. Red lines indicate a linear regression. An overview of the patient groups, the analysed samples and the full model parameters are given in Supplementary Table 1. CMV = cytomegalovirus.