A Mycobacterium tuberculosis Specific IgG3 Signature of Recurrent Tuberculosis

Abstract

South Africa has the highest prevalence of HIV and tuberculosis (TB) co-infection globally. Recurrent TB, caused by relapse or reinfection, makes up the majority of TB cases in South Africa, and HIV infected individuals have a greater likelihood of developing recurrent TB. Given that TB remains a leading cause of death for HIV infected individuals, and correlates of TB recurrence protection/risk have yet to be defined, here we sought to understand the antibody associated mechanisms of recurrent TB by investigating the humoral response in a longitudinal cohort of HIV co-infected individuals previously treated for TB with and without recurrent disease during follow-up, in order to identify antibody correlates of protection between individuals who do not have recurrent TB and individuals who do. We used a high-throughput, "systems serology" approach to profile biophysical and functional characteristics of antibodies targeting antigens from Mycobacterium tuberculosis (Mtb). Differences in antibody profiles were noted between individuals with and without recurrent TB, albeit these differences were largely observed close to the time of re-diagnosis. Individuals with recurrent TB had decreased Mtb-antigen specific IgG3 titers, but not other IgG subclasses or IgA, compared to control individuals. These data point to a potential role for Mtb-specific IgG3 responses as biomarkers or direct mediators of protective immunity against Mtb recurrence.

Trial registration: ClinicalTrials.gov NCT01539005.

Keywords: IgG3; Mycobacterium tuberculosis; antibodies; recurrence; recurrent tuberculosis.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Fischinger, Cizmeci, Shin, Davies, Grace, Sivro, Yende-Zuma, Streeck, Fortune, Lauffenburger, Naidoo and Alter.

Figures

Figure 1

Mtb-specific antibody profile across recurrent/no recurrent TB patient cohort. (A) The annotated heatmap shows antibody measurements across time, groups and displays demographic data. The first row of the heatmap delivers information of grouping into individuals who go on to get recurrent TB (orange) versus who do not get recurrent TB (grey). Days to TB recurrence for the recurrence group is indicated in shades of orange, days from cure are colored in grey for all individuals. The row below in blue shades reveals if participants had (1) or did not (2) have previous TB before enrollment into the parent study. Age, gender, CD4 levels and viral load are indicated. Antibody effector functions, titer and Fc-receptor binding are displayed, antigens for the functions are PPD, EDAT6/CFP10 and LAM, additional antigens for titer and Fc-receptor binding include Hspx, Ag85, RV0826 and RV1363, in this order. (B) IgG1 and IgG3 titers are shown over time for the recurrent TB group (top) and the no recurrence group (bottom) for PPD, ESAT6/CFP10, LAM and Hspx. X axis depicts time since TB cure1 at enrollment into the study and color in the recurrence group represents time to TB recurrence in shades of orange.

Figure 2

Mtb-specific antibody IgG1 titers correlate across different antigens. Graphs show the most distant time point from time since cure (and closest time point to recurrence for the recurrent TB group) for each individual in both groups. (A) Nightingale rose charts show comparisons between IgG1 titers against different TB-antigens between individuals with recurrent and no recurrent TB. Petals are scaled based on the median of the percent rank of each feature per group. (B) The Spearman correlation heatmap depicts correlation r values for the no recurrence group (grey) and the recurrence group (orange) for IgG1 titers across antigens. Color depth indicates r values, asterisk represent significance, adjusted for multiple comparison using Bonferroni correction method, adjusted p value *< 0.05, **< 0.01, ***<0.001, ****<0.0001.

Figure 3

Individuals who do not get recurrent TB have higher titers of IgG3. All graphs show data for the most distant time point from time since cure. (A) Nightingale rose charts allow for comparison between groups for antibody titer (IgG1-IgG4, IgA, IgM), Fc-receptor binding levels (FCR2A, 2B, 3A, 3B) and the ability to induce effector functions (ADCP, ADNP) at the most distant time point from TB cure1. The left column shows flowers for the recurrent TB group, the right side for the no recurrence group. Relative responses for PPD (top), ESAT6/CFP10 (middle) and LAM (bottom) are shown. (B) Dot plots show univariate IgG3 levels (MFI) across different Mtb antigens, median is indicated by line. For Hspx, IgG3:IgG2 ratios are depicted additionally. Grey depicts no recurrence, orange recurrent TB. Each dot represents the average within each matching bin. P-values were calculated using Wilcoxon signed-rank test and adjusted for multiple comparison using Benjamini-Hochberg method.

Figure 4

Mixed effects model supports the indication of higher IgG3 levels in no recurrent TB individuals. Dot plots and the model depict the most distant time point from time since cure for every individual. The mixed effects models depict the overall differences in measured antibody features between individuals with (left side) and without (right side) recurrent TB. The models include recurrent and non-recurrent individuals taking into account age, gender, as well as days since TB cure1. The X-axis depicts the effect size between the groups and y-axis shows a measure of statistical significance.