BackgroundGalactosylation of immunoglobulin G (IgG) is reduced in rheumatoid arthritis (RA) and assumed to correlate with inflammation and altered humoral immunity. IgG hypogalactosylation also increases with age. To investigate dependencies in more detail, we compared IgG hypogalactosylation between patients with RA, patients with axial spondyloarthritis (axSpA), and healthy control subjects (HC), and we studied it in RA on the background of HLA-DRB1 shared epitope (SE), anticitrullinated protein antibodies (ACPA), and/or rheumatoid factor (RF) status.MethodsPatients with RA (n = 178), patients with axSpA (n = 126), and HC (n = 119) were characterized clinically, and serum IgG galactosylation was determined by capillary electrophoresis. Markers of disease activity, genetic susceptibility, and serologic response included C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), DAS28, SE, HLA-B27, ACPA, and RF. Expression of glycosylation enzymes, including beta 1–4 galactosyltransferase (B4GALT3) activity, were estimated from transcriptome data for B-cell development (GSE19599) and differentiation to plasma cells (GSE12366).ResultsIgG hypogalactosylation was restricted to RA and associated with increasing CRP levels (p < 0.0001). In axSpA, IgG hypogalactosylation was comparable to HC and only marginally increased upon elevated CRP. Restriction to RA was maintained after correction for CRP and age. Treatment with sulfasalazine resulted in significantly reduced IgG hypogalactosylation (p = 0.003) even after adjusting for age, sex, and CRP (p = 0.009). SE-negative/ACPA-negative RA exhibited significantly less IgG hypogalactosylation than all other strata (vs SE-negative/ACPA-positive, p = 0.009; vs SE-positive/ACPA-negative, p = 0.04; vs SE-positive/ACPA-positive, p < 0.02); however, this indicated a trend only after Bonferroni correction for multiple testing. In SE-positive/ACPA-negative RA IgG hypogalactosylation was comparable to ACPA-positive subsets. The relationship between IgG hypogalactosylation and disease activity was significantly different between strata defined by SE (CRP, p = 0.0003, pBonferroni = 0.0036) and RF (CRP, p < 0.0001, pBonferroni < 0.0012), whereas ACPA strata revealed only a nonsignificant trend (p = 0.15). Gene expression data indicated that the key enzyme for galactosylation of immunoglobulins, B4GALT3, is expressed at lower levels in B cells than in plasma cells.ConclusionsIncreased IgG hypogalactosylation in RA but not in axSpA points to humoral immune response as a precondition. Reduced B4GALT3 expression in B cells compared with plasma cells supports relatedness to early B-cell triggering. The differential influence of RA treatment on IgG hypogalactosylation renders it a potential diagnostic target for further studies.Electronic supplementary materialThe online version of this article (10.1186/s13075-018-1540-0) contains supplementary material, which is available to authorized users.