OA24 Age-associated T-bet+ B cells in healthy and autoimmune pregnancy

Abstract

Abstract Background/Aims Inflammatory rheumatic diseases (IRD), including systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA), affect females of reproductive age and are characterised by loss of immune tolerance and an increased burden of adverse pregnancy outcomes (APO). Mechanisms by which IRD may affect immune adaptations of healthy pregnancy to maintain host and fetal defence whilst tolerating male fetal antigens are poorly characterised. This study aimed to investigate B cell subpopulations in healthy and IRD in pregnancy, with a focus on IgD-CD27- (double negative, DN) T-bet+ B cells, considered pathogenic in IRD. Given disparate relationships between disease activity and risk of APO in SLE and RA pregnancy, the objective of this pilot study was to determine if pregnancy alters the frequency and/or the phenotype of DN T-bet+ B cells in IRD compared with healthy control pregnancies. Methods Peripheral blood mononuclear cells were obtained from IRD pregnancy (SLE n = 5; RA n = 8), IRD non-pregnancy (SLE n = 12; RA n = 10), and healthy pregnant (n = 8) people at University College London Hospital by informed consent. Cells were analysed by flow cytometry to identify B cell subpopulations and their expression of multiple markers including T-bet, IFN-γR, TLR7, TLR9, Ki67, CD38 and CD19 was assessed. Results There were no significant differences in B cell subpopulations, categorised based on the expression of IgD and CD27, between the cohorts. In healthy pregnancy, the median frequency (%) of CD19+T-bet+, IgD-CD27+T-bet+, DNT-bet+ cells was 5, 1.3 and 2, respectively. In contrast, the median frequency of CD19+T-bet+ cells was significantly elevated in SLE pregnancy at 13% compared with 5% (p < 0.05) in healthy pregnancy[GI3] [RV4] . Similarly, the median frequency of DNT-bet+ B cells was elevated in SLE pregnancy at 6% compared with 2% (p = 0.009) in healthy pregnancy. RA pregnancy, but not SLE pregnancy, had significantly lower frequency of Ki67+ DN T-bet+ cells compared to non-pregnant RA (p=0.04) and SLE (p=0.07). ROC analysis demonstrated that for both RA (p = 0.016, AUC = 0.84, 95% CI [0.65, 1.00]) and SLE (p = 0.015, AUC=0.88) the frequency of DN T-bet+ Ki67+ B cells can discriminate between pregnancy and non-pregnancy samples. In pooled analysis (n = 43), but not in single cohorts, DN T-bet+ cells had significantly high TLR7 and TLR9 expression. High frequencies of CD19hi DN T-bet+ cells were found but were unaffected by pregnancy. Clustering revealed canonical DN T-bet+ clusters and CD38hi T-bet+ clusters enriched in SLE and pregnancy-altered. Conclusion We have immunophenotyped B cell populations in healthy pregnancy and identified different patterns of alteration in frequency and proliferation of DNTbet+ cells in IRD compared with healthy pregnancy that are disease specific. This pilot study reveals previously unrecognized immunological changes in disease pregnancy warranting further larger studies of their mechanistic basis and clinical relevance. Disclosure S. Benton: None. C. Raine: None. K. Shah: None. S. Luong: None. A. Venkatkrishnan: None. M. Leandro: None. M. Castelino: None. D. Sen: None. D.A. Isenberg: None. A. Akbar: None. I. Giles: Grants/research support; IG has received an unrestricted research grant, travel and speaker fees from UCB Pharma. V.R. Reddy: None.