Introduction: The non-histone nuclear protein high mobility group box protein-1 (HMGB1) is typically associated with nucleosomes, but may shuttle between the nucleus and the cytoplasm, and under some conditions also be released extracellularly and participate in systemic inflammation. Monoclonal HMGB1-targeting antibodies can ameliorate murine polyarthritis and lupus-like disease. Interestingly, autoantibodies against HMGB1 have also been described in patients with systemic lupus erythematosus (SLE), but their clinical implications remain elusive. The main aims of this study were to detect serum anti-HMGB1 antibodies in patients with SLE and relate them to other types of antinuclear antibodies (ANA), and to disease activity.

Methods: 188 Swedish SLE patients meeting the 1982 American College of Rheumatology classification criteria and/or the 2012 Systemic Lupus International Collaborating Clinics classification criteria participated in the study. Anti-HMGB1 antibody levels were analysed in patient and control (n = 112) sera by an in-house ELISA using recombinant histidine-tagged HMGB1. SLE sera were also analysed for ANA by immunofluorescence (IF) microscopy (IF-ANA) using fixed HEp-2 cells, and by a line-blot assay for antigen fine-specificities. To quantify antibodies to double-stranded DNA, a fluoroenzyme-immunoassay was employed.

Results: At inclusion, 23 % of the SLE patients were anti-HMGB1 antibody positive compared to 5 % of the controls. Anti-HMGB1 antibodies occurred in 49 % of the IF-ANA positive SLE patients, and in 34 % of IF-ANA negative cases (p = 0.004). Levels of anti-HMGB1 antibodies correlated with anti-dsDNA antibody levels (r = 0.49; p < 0.001). Significant, but less pronounced correlations were found regarding anti-HMGB1 and SLE disease activity index (SLEDAI-2K: r = 0.15; p = 0.04), classical complement function (r = -0.24; p = 0.002) and complement protein C4 (r = -0.23; p = 0.002). Average anti-HMGB1 antibody levels were significantly higher among patients with homogenous ± other IF-ANA staining patterns (median 180 AU) compared to IF-ANA negative cases (median 83 AU) (p = 0.004). Rabbit anti-HMGB1 antibodies gave rise to cytoplasmic, but not nuclear, staining of HEp-2 cells.

Conclusions: We confirm that anti-HMGB1 antibodies are common in SLE and correlate with disease activity variables. Although anti-HMGB1 antibodies measured by ELISA often coincide with nuclear IF-ANA staining, our results indicate that anti-HMGB1 antibodies do not give rise to nuclear staining of the predominantly used commercial HEp-2 cell slides.