Distinct and shared B cell responses of tuberculosis patients and their household contacts

Abstract

This study was aimed at identifying the B cell responses which could distinguish between 'latent tuberculosis infection (LTBI)' and active TB disease. Study subjects were smear-positive TB patients (n = 54) and their disease-free household contacts (HHCs, n = 120). The sera were used for determination of antibody levels (ΔOD values) against Mycobacterium tuberculosis membrane (MtM) antigens by ELISA and for visualisation of seroreactive MtM antigens by immunoblotting. B cell subsets in whole blood samples were determined by flow cytometry. In TB sera, levels of IgG antibodies were significantly higher than IgM and IgA whereas IgM and IgA antibody levels were comparable. Conversely, HHC sera had significantly higher IgM antibody levels than IgG and IgA. The ratio of IgM to IgG antibodies in HHCs were also significantly higher than in patients. Immunoblotting revealed that some of the MtM antigens (<10, ~12 and ~25 kDa) reacted with TB as well as HHC sera whereas some other antigens (~16, ~36, ~45 and ~60 kDa) reacted with most of TB and a subset of HHC sera. Frequencies of classical memory B cells (cMBCs, CD19+CD27+) were significantly higher, and of IgG+ cMBCs were significantly lower in HHCs than in patients. Frequencies of IgA+ cMBCs in HHCs and patients were comparable but both were significantly higher than the corresponding frequencies of IgG+ cMBCs. Frequencies of IgA+ atypical MBCs (aMBCs, CD19+CD27-) in HHCs and patients were also comparable and significantly higher than the IgG+ aMBCs. The plasmablast (CD19+CD27++CD38++) frequencies in HHCs and patients were comparable. These results suggest that the IgM/IgG antibody ratio, antibody binding to selected MtM antigens and relative frequencies of MBC subsets could indicate protective or pathogenic immune responses following the primary infection with Mtb. Responses that orchestrate protection leading to a 'quiescent' LTBI may provide clues to an effective vaccination strategy against TB.

Fig 1

[A] Serum IgG, IgM and IgA antibody responses (ΔOD values) of HHCs and TB patients against MtM antigens. [B] IgM/IgG antibody ratios in HHCs and TB patients. Inset figure depicts proportion (%) of subjects showing a ratio of >1 (numbers on top of the bars denote % values). Statistical analyses were done by either the Wilcoxon test (for paired datasets) or Mann-Whitney test (for unpaired datasets). Two-tailed P values are shown on top of the relevant columns. NS, not significant (P > 0.05).

Fig 2

Correlation between IgG, IgM and IgA antibodies (ΔOD values) in HHCs (panels A, C and E) and TB patients (panels B, D and F). Spearman’s rho and P values are shown for each XY pair. NS, not significant (P > 0.05).

Fig 3. Comparative immunoblot analysis of TB (n = 20, T1-T20) and HHC (n = 20, H1-H20) sera against MtM antigens.

Boxes A-G highlight antigen-antibody reactions which were shared, to a variable extent, by both the subject categories. Bands in the molecular weight range of ~25 kDa (box D), ~12 kDa (box F) and <10 kDa (box G) were shared by nearly all sera, and the other major bands (boxes A-C and E) were shared by a subset of the sera.

Fig 4. Identification of B cell subsets in whole blood.

[A] RBCs were lysed, and forward (FSC) and side (SSC) scatter of WBCs was used for gating the lymphocytes. [B, C] From lymphocyte gate, B cells (stained for CD19) were gated and further analysed for staining of CD27 and CD38 in order to identify the plasmablasts (CD19+CD27++CD38++). [D] Lymphocytes were segregated on the basis of staining for CD19 and CD27 to select the classical memory B cells (CD19+CD27+) which were further segregated as IgA+ [E] and IgG+ [F] ‘class-switched’ classical memory B cells. The CD19+CD27- B cells [G] were further segregated as IgA+ [H] and IgG+ [I] atypical memory B cells.

Fig 5. Frequencies of B cell subsets in peripheral blood of TB patients (TB, n = 47) and their household contacts (HHC, n = 113).

Panel A shows frequencies of classical memory B cells (cMBCs, CD19+CD27+) as % of lymphocytes, class-switched (IgA+ or IgG+) cMBC as % of cMBCs and plasmablasts (CD19+CD27++CD38++) as % of B cells. Panel B shows frequencies of IgA+ or IgG+ atypical memory B cells as % of CD19+CD27- B cells. The gating strategy is described in Fig 4. Statistical analyses were done by either the Wilcoxon test (for paired datasets) or Mann-Whitney test (for unpaired datasets). P values for statistically significant differences are shown on top of the respective columns. NS, not significant (P > 0.05).

Fig 6. Correlations between B cells subsets.

Panels A and B show significant positive correlations between IgA+ classical and atypical memory B cells in HHCs and TB patients. Panels C and D show lack of correlation between IgG+ classical and atypical memory B cells in HHCs and TB patients. Spearman’s rho and P values are shown in each panel. NS, not significant (P > 0.05).