A prospective comparative study based on the evaluation of clinical and laboratory data with the analysis of sera samples which were collected in 2017 - 2019 at the St. Petersburg Research Institute of Phthisiopulmonology and the City Hospital No. 2.

Exclusion criteria: a period of more than 2 years from the detection of radiographic changes in the lungs, receiving anti-tuberculosis therapy, the presence of HIV infection, syphilis, neoplastic diseases, and decompensated diabetes mellitus.

According to the study design, 46 patients with pulmonary tuberculosis with a bacterial excretion were included ( Table 1 ). The control group was comprised of 40 healthy volunteers with no chronic diseases, contacts with tuberculosis and changes in laboratory parameters.

According to Table 1 , clinical manifestations of the disease and comorbidities were detected in 89.1% and in 21.7% of patients respectively.

The study was approved by the Independent Ethical Committee of the St. Petersburg Research Institute of Phthisiopulmonology (protocol No. 34.2 dated 01/19/2017) and the Local Ethical Committee of St. Petersburg State University (protocol No. 01-126 30.06.17). All the participants in the study signed an informed consent.

The diagnosis of pulmonary tuberculosis was verified by the M. tuberculosis detection in the sputum and/or MBT DNA according to molecular genetics and bacteriological methods, with the presence of typical X-ray changes (27).

All study participants were surveyed according to the standardized “ASIA Research Questionnaire”, to assess the impact of ASIA - trigger factors, as well as to determine the compliance of existing clinical manifestations with the ASIA syndrome diagnostic criteria (16). In addition, the history of a professional contact with trigger factors (long-term contact with printer toner, dust, metals, chemicals, cars etc.) before the appearance of clinical and radiological manifestations of the disease were studied ( Table 2 ).

The level of the antibodies to a mutated citrullinated vimentin (anti-MCV) was measured in the sera of 28 patients with Tbc and 40 healthy controls. The patients with positive for anti-MCV antibodies were evaluated for the presence of antibodies to cyclic citrullinated peptide (anti-CCP). Antibodies to anti-MCV were measured using ELISA (ORGENTEC, Germany), anti-CCP with ELISA (Euroimmune, Germany).

Additionally the level of 17 most prevalent autoantibodies were determined. The following antibodies were analyzed: autoantibodies to thyroglobulin (-a-TG), autoantibodies to thyroperoxidase (-a-TPO), IgG antibodies against double-stranded DNA (-dsDNA) were determined using ELISA (Euroimmune, Germany); antinuclear antibodies (-ANA), antibodies to neutrophil cytoplasm (-ANCA), antibodies to smooth muscle (-ASMA), antimitochondrial antibodies (-AMA), antibodies to gastric parietal cells (-APCA) using indirect immunofluorescence (Euroimmune AG, Germany); profile of antinuclear antibodies (SS-a,SS-B,Scl-70, Sm, CENP-B) using immunoblot (Euroimmune, Germany); antibodies to b2-glycoprotein (-b2GP), antibodies to liver and kidney microsomes (-LKM), cardiolipin antibodies IgM, IgG (-ACLA-G, ACLA-M), antibodies to C1q complement factor (-a_C1q) using ELISA (ORGENTEC, Germany).

The manufacturers were used for determination of positive autoantibody concentration.

200 μL of peripheral blood of 41 Tbc patients and 37 healthy donors was stained for B cell subsets studies with fluorochrome-conjugated monoclonal antibodies: – anti-IgD, anti-CD38, anti-CD183, anti-CD27, anti-CD24, anti-CD19, anti-CD5 and anti-CD45 (25). Red blood cell were lysed, and blood samples were washed by centrifugation and fixed with PBS containing 2% of neutral buffered formalin solution (Sigma-Aldrich),. Sample acquisition was performed using a Navios flow cytometer (Beckman Coulter, Inc.),. At least 5000 CD19+ B cells were analyzed in each sample. Gating strategy was described previously in details (28) and on Figure 1 .

Flow cytometry data were analyzed by Kaluza software v2.3 (Beckman Coulter, Inc.),. To determine the frequencies of circulating B cell subsets the relative expression of IgD and CD38 (“Bm1-Bm5” classification) was used, which allows to identify IgD+CD38- “naïve” Bm1 cell, IgD+CD38+ “activated naïve” Bm2 cells, IgD+CD38++ pre-germinal-center Bm2’ cells, IgD-CD38++ centroblasts and centrocytes (so-called “Bm3+Bm4” cells), and two subsets of memory B cells – IgD-CD38+ early memory (eBm5) and IgD-CD38- resting memory cells (Bm5) as suggested by Bohnhorst et al. (29).

For follicular Th cell subsets, 200 μL of whole peripheral blood was stained with the following fluorochrome-conjugated monoclonal antibodies cocktail: anti-CD183 (CXCR3), anti-CD25, anti-CD185 (CXCR5), anti-CD194 (CCR4), anti-CD196 (CCR6), anti-CD4, anti-CD8, anti-CD3, anti-CD197 (CCR7) and anti-CD45RA as it was described previously in details (25). The red blood cell were lysed, and blood samples were washed by centrifugation and fixed with PBS containing 2% of neutral buffered formalin solution (Sigma-Aldrich). The sample acquisition was performed using a Navios flow cytometer (Beckman Coulter, Inc., USA). At least 40000 CD3+CD4+ Th cells were analyzed in each sample. Gating strategy was described previously in details (28) and on Figure 2 .

Statistical analysis was performed using GraphPad Prism 6 (Graph Pad Software), Statistica 10 (Statsoft) and MedCalc – version 18.2.1 (Ostend, Belgium) values. Flow cytometry data were by Kaluza software v2.3 (Beckman Coulter, Inc.),.

The Mann-Whitney-U and Fisher’s exact tests were used for non-parametric data. Quantitative data were presented as m ± sd. The degree of association was calculated using confidence intervals, as well as the χ² test with Yeats correction. To determine the relationship between the values, Spearman correlation analysis was performed. Differences or relationship indicators were considered statistically significant at p-value less than 0.05.

Previously, the relationship between the development of autoimmune pathology after the introduction of an attenuated strain of M. bovis was shown. M. bovis is the main causative agent of tuberculosis in cattle and it is used for immunization in humans to date, both for the prevention of Tbc and for the treatment of oncological pathology and even severe COVID-19 (8, 26, 35, 36).

In the experiment, Mtb is quite often used as an adjuvant, for example, in a complete Freund adjuvant in animal models of autoimmune diseases (37), which is presumably related to the fact that these antigens overcome tolerance to host antigens when co-administered. In experimental models, Mtb immunization can cause autoimmune joint lesions by the cross-reactivity with proteoglycan in cartilage (35).

Currently, there is an evidence of the trigger role of M. tuberculosis in the development of diseases such as sarcoidosis, systemic lupus erythematosus, rheumatoid arthritis, primary biliary cirrhosis and many others (11, 19, 38–40).

No reliably known mechanism which are responsible for the formation of antibodies in tuberculosis. This antibodies have been identified to date. There are assumptions about possible mimicry between the antigenic structure of mycobacteria and the host tissue’s own antigens (12, 21, 41).

At the same time, a number of studies have shown that in 40% of cases in tuberculosis patients, antibodies typical for granulomatosis with polyangiitis, systemic lupus erythematosus and other autoimmune diseases were detected (4, 35). Statistically significant increase in plasma concentrations of antibodies in tuberculosis patients was diagnosed to ribonucleoproteins (15%), anti-SSA (64%) and anti-ACA-IgM antibodies (59%) (42–44). In rare cases, antibodies to neutrophil cytoplasm, antibodies to beta-2-glycoprotein (anti-b2GPI), antibodies to cyclic citrullinated peptide, anticardiolipin antibodies are registered. At the same time, the frequency of detection of these autoantibodies was in some cases similar to that in patients with autoimmune pathology, as well as anti-tuberculosis treatment can lead to normalization of some AAB (35, 45, 46).

Previously, Elkholy et al. reported that the frequency of CD3-CD19+ B cell in peripheral blood from patients with active pulmonary tuberculosis was significantly lower than in a control group (47). In contrast, Wu et al. found that CD19+ B cells were increased in Tbc patients vs a control group (48). We noticed no differences in relative and absolute numbers on total CD19+ B cells subset between M. tuberculosis infected patients and a healthy control, but we found dramatic alterations in B cell subsets composition.

Our data indicated that peripheral blood B cells from Tcb patients showed the decreased levels of “naïve’ B cell that expressed unique B cell receptors and were able to recognize and to initiate immune response to new specific antigens in secondary lymphoid organs. Furthermore, M. tuberculosis infected patients had low levels of memory eB5 cells, that was essential for long-term protection against reinfection. Thus, the recognition of new antigens and effective activation of immunological memory which have been encountered previously and they could functionally impaired during M. tuberculosis infection. Similarly, Joosten et al. reported that in Tbc patients the level of atypical circulating subsets (CD21−CD27− or IgD−CD27−) and activated (IgD−CD27+) B-cellsreduced numbers of IgD+CD27− naïve B-cells if compared to control (49). Furthermore, we found in Tbc patients IgD+CD38++ pre-germinal-center Bm2’ cells were activated, that could leave the blood stream, enter secondary lymphoid organs and initiate humoral immunity in response to antigen recognition. We should mention that Bm2’ cells are highly heterogeneous, and may contain regulatory B cell subset with CD24++CD38++ phenotype, as it was shown previously by Blair et al. (50). Similarly, several paper also reported that peripheral blood circulating Bregs with CD19+CD1d+CD5+ phenotype were increased in patients with tuberculosis (51, 52). Interestingly, Breg subsets were found and increased in patients with various autoimmune diseases, including ankylosing spondylitis (53), primary Sjögren’s syndrome (54), diabetes mellitus (55), multiple sclerosis and neuromyelitis optica spectrum disorders (56), etc., suggesting a close involvement of CD24++CD38++ B cells in the autoimmune pathogenesis.

Tfh cells have been shown to mediate protective immunity against tuberculosis via accumulation in infected lungs and production of proinflammatory cytokines, while mice deficient in Cxcr5 had increased susceptibility to Tbc due to defective T cell localization within the lung parenchyma (57). Recently, Sun et al. noted that the frequencies of Tfh cells with the following phenotypes CD4+CXCR5+, CD4+CXCR5+ICOS+ and CD4+CXCR5+PD-1+Tfh as well as peripheral blood IL-21 levels were increased in Tbc patients if compared to healthy controls (58). Controversially, Kumar et al. observed that active pulmonary tuberculosis was characterized by diminished frequencies of Tfh cells ex vivo and in response to Tbc antigens and by diminished frequencies Tfh cells producing IL-21 (23). Though, we found no differences in circulating memory Tfh cells between M. tuberculosis infected patients and healthy controls, however, we noticed alterations in Tfh cell subsets. However, the absolute number and the frequency of Th1 cells within the CD45RA–CCR7+CXCR5+ CD4+ T cells compartment were significantly lower in patients with tuberculosis when compared to a healthy control, while the relative number of Tfh2 cell was increased. Indeed, imbalance of cytokines that were realized in the site of M. tuberculosis infection and were produced by activated dendritic cells in secondary lymphoid tissue could dramatically influence Th cell polarization and affect all types of Th cell subsets, including different subsets of Tfh cells (59).

CXCR3+ Tfh1 diminished the functions to provide help to B cells in a mouse malarial model (60) and human in vitro experiments, when they were shown to induce apoptosis in activated naïve B cells (61), while Tfh2 and Tfh17 induced differentiation of naive B cells to plasma cells, activated class-switching and stimulated antibodies production (32). Thus, the ratio of Tfh2 and Tfh17 that provided help to B cells over Th1 (able to block B cell activation) was considered to be important for effective humoral immune responses. Furthermore, numerous studies also suggested that an altered balance of blood CXCR5+ CD4+ Tfh cell subsets was closely linked with autoimmunity. For instance, in SLE patients with an active Tbc (SLEDAI score>8) the frequency of Tfh2 was significantly increased that was accompanied with Tfh1 cell subset percentage decrement, and alterations in relative numbers of Tfh2 and Tfh1 cells were associated with the presence of high Ig levels and autoantibodies in patient’s sera (62). Furthermore, patients with relapsing-remitting and secondary progressive multiply sclerosis decreased frequencies of Th1-like Tfh cells (63). Next, patients with IgG4-related disease exhibited significantly the increased level of Tfh2 cells and significantly the decreased level of Tfh1 cells compared to a healthy control group (64). Thereby, altered Tfh subset balance was associated with a wide range of autoimmune diseases in humans, but still it is poorly study.

The anti-MCV antibodies were defined in Tcb patients and not in a control group, that could imply an autoimmune response in pathogenesis of the disease (65–68).

The presence of an autoimmune component associated with an increase in the level of autoantibodies may be significant for the correction of therapy and serves as a criterion for considering the appointment of immunosuppressive therapy in the future.

Probably, the reason for the elevation of autoantibodies may be molecular mimicry, which is described in chronic infections with a hyperactive immune response. Those findings are coherent with our data. The heat shock proteins Mtb-HsP60, Mtb-HsP65, catalase (mKatG) can be considered as the candidate mycobacterial antigens, possibly involved in the cross-reaction (69–71).

The presence of antibodies in patients with tuberculosis may reflect the relationship between the pathogenesis of those diseases with the possibility of cross-reactivity between vimentin and M. tuberculosis peptides (71–73).