Distinct Longitudinal Changes in Immunoglobulin G N-Glycosylation Associate with Therapy Response in Chronic Inflammatory Diseases

Abstract

Immunosuppressants and biologicals are widely used therapeutics for various chronic inflammatory diseases (CID). To gain more detailed insight into their downstream effects, we examined their impact on serum immunoglobulin G (IgG) glycosylation. We analyzed IgG subclass-specific fragment crystallizable (Fc) N-glycosylation in patients suffering from various CID using the LC-MS approach. Firstly, we compared IgG Fc N-glycosylation between 128 CID patients and 204 healthy controls. Our results replicated previously observed CID-related decrease in IgG Fc galactosylation (adjusted p-value range 1.70 × 10^-2-5.95 × 10^-22) and sialylation (adjusted p-value range 1.85 × 10^-2-1.71 × 10^-18). Secondly, to assess changes in IgG Fc N-glycosylation associated with therapy and remission status, we compared 139 CID patients receiving either azathioprine, infliximab, or vedolizumab therapy. We observed an increase in IgG Fc galactosylation (adjusted p-value range 1.98 × 10^-2-1.30 × 10^-15) and sialylation (adjusted p-value range 3.28 × 10^-6-4.34 × 10^-18) during the treatment. Furthermore, patients who reached remission displayed increased Fc galactosylation levels (p-value range 2.25 × 10^-2-5.44 × 10^-3) in comparison to patients with active disease. In conclusion, the alterations in IgG Fc glycosylation and the fact these changes are even more pronounced in patients who achieved remission, suggest modulation of IgG inflammatory potential associated with CID therapy.

Keywords: IgG glycosylation; autoimmune diseases; chronic inflammatory diseases; inflammatory bowel disease; personalized medicine; response.

Figure 2

Changes in IgG1 Fc glycan profiles during treatment in three cohorts of UC and CD patients. Median glycan values for each time point are bolded. Y-axis: relative change in glycan value normalized to baseline (1st timepoint); X-axis: duration of follow-up (weeks). Results of meta-analysis are shown in Table 1, while the results of statistical analysis for individual cohorts are shown in Table S5.

Figure 3

Changes in IgG2/3 Fc glycan profiles during treatment in CID (UC and CD) patients that entered remission (R) vs. patients with active disease (NR) in three cohorts. Median glycan values for each time point are bolded. Y-axis: relative change in glycan value normalized to baseline (1st timepoint); X-axis: duration of follow-up (weeks). Associations of IgG Fc glycan profiles with disease activity during treatment, for each patient cohort separately, are shown in Table S9, while the results of meta-analysis of all three cohorts are presented in Table 2.

Figure 1

IgG Fc-glycan patterns in controls, patients suffering from ulcerous colitis, and patients suffering from Crohn’s disease (Cohort 1 at baseline) for all IgG subclasses. Each box represents the 25th to 75th percentiles (interquartile range—IQR). Lines inside boxes stand for the median. The whiskers are the lowest and highest values within boxes ±1.5 × the IQR. Dots are outliers (>1.5 × IQR). Significant differences (p < 0.05) are marked with an asterisk *. Normalized size effect and p-values are given in Table S1. C—controls; UC—ulcerous colitis; CD—Crohn’s disease.