Despite significant efforts made to control tuberculosis (TB), the increasing burden of diabetes mellitus (DM) threatens the progress registered in reducing the global burden of TB, especially in low and middle-income countries (LMICs) (1). According to the 2021 International Diabetes Federation (IDF) estimates, approximately 537 million adults (aged between 20 and 79) live with DM. This figure is projected to rise to 783 million by 2045, with the most significant increase in Africa (2). Tuberculosis remains one of the leading causes of death from a single infectious agent, Mtb, worldwide (3). Globally, approximately 7.5 million people were newly infected with Mycobacterium tuberculosis (Mtb) or diagnosed with TB in 2022, with nearly 1.3 million deaths occurring (3). Epidemiologically, DM confers a 3-fold increase in the risk of developing TB disease and is associated with TB treatment failure and drug resistance (4). Indeed, it was recently reported that participants aged ≥ 40 years had increased odds of TB-DM comorbidity (5) and that Africans with DM have an increased latent TB infection (LTBI) risk (6). The risk for the development of active TB (ATB) is thought to be due to the immune-compromised status, but the underlying susceptibility mechanisms remain largely unknown.

The quality of the T-cell response is essential for Mtb immunity. CD4⁺ and CD8⁺ T cells are pivotal for immune control in Mtb-infected humans and murine TB models (7, 8). T-cell memory phenotypes are induced during LTBI and Bacillus Calmette-Guerin (BCG) vaccination that play a protective role in humans and in mice models (9–12). It is reported that LTBI is characterised by differential expression of functional markers, including decreased HLA-DR expression, a marker that distinguishes LTBI and ATB (13, 14), upregulated PD-1 expression, a marker that inhibits T-cell effector functions (15, 16), as well as downregulated Th1 (7) and Th17 (17, 18) cytokine production. Examining cytokine T-cell polyfunctionality is essential as these cells have been associated with resistance to infection (19, 20). Elevated frequencies of mono-functional and dual-functional CD4⁺ Th1 cells are reportedly a hallmark of active TB and DM (TB-DM) comorbidity (21). This shows that type 2 DM modulates T-cell immune responses to Mtb, which could profoundly affect TB pathogenesis. However, the underlying immunological mechanisms for TB susceptibility during DM remain to be elucidated, specifically with phenotypes and functional markers during LTBI.

In this study, we hypothesised that type 2 DM modulates the Mtb-specific memory phenotype and functional profiles of T cells among participants with LTBI, leading to impaired responses and potentially promoting TB susceptibility, progression or reactivation. We aimed to assess the Mtb-specific CD4⁺ and CD8⁺ T-cell memory phenotypes and functional profiles. We compared the T-cell memory, activation, degranulation, exhaustion and cytokine polyfunctionality profiles among participants with LTBI-DM comorbidity.

Immunological dysregulation is one mechanism that accounts for TB susceptibility and severity in DM, but it is not well elucidated and remains poorly characterised. We performed an extended analysis of the memory phenotypes and functional responses of Mtb-specific CD4⁺ and CD8⁺ T cells to identify immunological differences between LTBI-DM, LTBI-only and DM-only participants. Our study identified three key points: 1) Type 2 DM alters the memory phenotype of CD4⁺ and CD8⁺ T cells; 2) Type 2 DM impairs T-cell activation and degranulation but promotes T-cell exhaustion; 3) Type 2 DM impairs the CD4⁺ and CD8⁺ T-cell Th1, Th2 and Th17 cytokine responses, as well as the polyfunctional (triple, dual, mono) capacity of the CD4⁺ T-cell, and central and effector memory CD4⁺ T-cell subsets. We showed that type 2 DM is associated with profound impairment of Mtb-specific T-cell responses, which could increase TB susceptibility.

This study reveals that naïve CD4⁺ T cells were decreased, whereas the CD4⁺ and CD8⁺ T-cell central and effector memory phenotypes were increased in the LTBI-DM compared to the LTBI-only participants. The reduction in naïve CD4⁺ T cells is similar to a study by Kumar and colleagues, who reported decreased naïve CD4 T cells in active TB with DM participants (25). The decrease indicates a potential compromise towards delayed or insufficient immune responses against Mtb reactivation, allowing Mtb to potentially proliferate and increase susceptibility to active TB disease (26). The significant increase of central and effector memory T-cell frequencies in LTBI-DM participants implies a shift towards an activated memory T-cell phenotype. Memory T cells are crucial for long-term immune surveillance (27, 28) and rapid response upon re-exposure to Mtb (29). This increase may reflect an immune response to chronic Mtb stimulation or a compensatory mechanism in response to impaired naïve T-cell function. This could have implications for both TB protection and disease progression, as an increased T-cell memory phenotype could potentially contribute to Mtb-related chronic inflammation, resulting in T-cell memory cells with impaired immune function, including exhaustion, activation, homing and cytokine production (30). Type 2 DM orchestrated T-cell memory alteration may potentially decrease the overall robustness of the T-cell memory response, potentially increasing susceptibility to active TB disease.

The functional profiles and fitness of the T cells are significant factors to consider when assessing Mtb-specific responses in the face of DM. Our study reports significant upregulation of PD-1 on T cells in the LTBI-DM participants, a consensus to several studies that reported upregulation of PD-1 expression on T cells during Mtb infection and active TB disease (15, 16). PD-1 impairs T-cell proliferation during active TB disease (16) and Th1 immune function during Mycobacterium bovis BCG vaccination (31). Type 2 DM promoting increased PD-1 expression could have severe implications for other T-cell functional responses, including activation, degranulation and cytokine production. Interestingly, we report that type 2 DM impairs T-cell activation and degranulation. CD8⁺ T-cell HLA-DR expression was decreased in the LTBI-DM participants compared to the LTBI-only group, an association with a lower activation state, and consistent with another human study that reported impaired HLA-DR expression on H37Rv-infected monocyte-derived macrophages of DM patients (32). HLA-DR is an activating receptor that binds and presents antigens to T cells, thereby activating immune responses, including cytokine and cytotoxicity functions to clear Mtb-infected cells (33). Its expression has also been characterized with effector T cells (34). The decrease in the CD8⁺ T-cell activation state in the face of DM could impair their cytotoxic functions (33), leading to increased risk for LTBI acquisition and ATB progression. However, our study reports that fewer CD8⁺ (but not CD4⁺) T cells were activated. This needs a cautious interpretation, as TB immune impairment is often related to CD4⁺ T-cell HLA-DR dysfunction (35). Interestingly, HLA-DR expression has previously been described as a biomarker that distinguishes LTBI from ATB (36). Whether HLA-DR expression could be used as a biomarker for identifying and distinguishing TB phenotypes in coincident DM remains to be assessed. In addition, our study reports that type 2 DM is associated with compromised CD4⁺ and CD8⁺ T-cell CD107a, a marker for degranulation and cytotoxicity function (37). Similar results have been reported for which type 2 DM compromises the cytotoxic effects of CD8⁺ T and NK cells during active TB (38). CD4⁺ and CD8⁺ T cells have been reported to kill Mtb-infected monocytes directly by perforin and Fas/Fas Ligand independent pathways (39). It is important to note differences in the expression profiles of PD-1 and HLA-DR in LTBI-DM and DM groups. These differences may reflect distinct mechanisms of immune activation in the DM group that are not directly related to Mtb-specific immune responses in the LTBI-DM group. PD-1 and HLA-DR can be influenced by various factors, including metabolic dysregulation caused by DM (40, 41). Taken together, impairment of HLA-DR expression and CD107a production by DM could promote heightened Mtb replication and increased TB risk.

Next, we assessed the effect of DM on CD4⁺ and CD8⁺ T-cell cytokine production, and we observed marked differences in cytokine expression profiles for IFN-γ, IL-2, TNF, IL-13 and IL-17A. CD4⁺ T-cell IFN-γ and TNF, as well as CD8⁺ T-cell IFN-γ and IL-17A production, were decreased, whereas CD8⁺ T-cell IL-13 production was increased in the LTBI-DM participants compared to LTBI-only participants. CD4⁺ and CD8⁺ T-cell IFN-γ production mediates TB protection by controlling the Mtb burden and promoting host survival in mice (7) and humans (8). In addition, T-cell-derived TNF plays a crucial role in the early control of TB infection and promotes the formation of mature granulomas and the activation of infected macrophages in mice (42). Similarly, T-cell IL-17A, a Th17 family cytokine, recruits immune cells to Mtb-infected sites by upregulating chemokine expression, thereby contributing to granuloma formation and stability (43, 44). On the contrary, increased production of IL-13 is associated with lung damage and the formation of necrotic lesions in mice, which promotes and is consistent with human TB pathology (45, 46). Impairment of the CD4⁺ and CD8⁺ T-cell cytokine responses by DM in the face of TB infection could promote Mtb replication, thus promoting TB pathology.

Lastly, we assessed the effect of DM on combinations of polyfunctional Th1 cytokine co-expression profiles of CD4⁺ and CD8⁺ T cells, as well as CD4⁺ T-cell memory phenotypes. Several studies that have profiled the role of polyfunctional CD4⁺ T cells in producing multiple Th1 cytokines (IFN-γ, IL-2, TNF) during TB infection have associated polyfunctional CD4⁺ T cells with protection against TB (47–51). It is conceivable that polyfunctional T cells are more effective at controlling infection than those producing single cytokines. Whether these can be used as targets for TB vaccination in the face of DM remains to be assessed in more extensive studies. Our study is among the first to evaluate the impact of type 2 DM on CD4⁺ and CD8⁺ T-cell polyfunctionality, as well as the CD4⁺ T-cell central and effector memory polyfunctionality. Interestingly, BCG vaccination in mice and humans has been reported to induce polyfunctional CD4 central and effector memory T cells that confer protective memory immunity against TB in a mice model (11, 12). Our data reveals that DM significantly impairs the dual (IFN-γ⁺IL-2⁺ and IL-2⁺TNF⁺) and mono (IFN-γ⁺ and TNF⁺)-functional capacity of Mtb-specific CD4⁺ T cells in the LTBI-DM compared to the LTBI-only participants. Additionally, DM significantly impaired the triple (EM: IFN-γ⁺IL-2⁺TNF⁺), dual (CM: IFN-γ⁺IL-2⁺ and IL-2⁺TNF⁺; EM: IFN-γ⁺ TNF⁺ and IL-2⁺TNF⁺), and mono (CM: IFN-γ⁺, IL-2⁺ and TNF⁺; EM: IFN-γ⁺ and TNF⁺)-functional capacity of the Mtb-specific CD4⁺ T-cell central and effector memory responses, contributing to first evidence of DM immune impairment on polyfunctional CD4⁺ T-cell memory responses. The results are consistent with a study by Kumar et al. and colleagues (52) that reported diminished frequencies of dual- and mono-functional CD4⁺ T cells in LTBI-DM participants. Moreover, Kamboj et al. (53) reported improved Mtb clearance after restoring dual functional IFN-γ⁺TNF⁺ CD4⁺ T cells, further highlighting the importance of polyfunctional T cells as correlates of TB protection. This study demonstrates DM immune-modulatory effects and impairment of both Mtb-specific CD4⁺ T cells and their central and effector memory polyfunctional responses during TB progression. This may promote increased TB disease risk and increase active TB progression.

This study faces limitations, including a limited sample size. It is also important to note that the data generated after in vitro culture may not represent what occurs in vivo. In addition, HbA1c and other DM-related parameters were not collected for participants in the LTBI-only group as these were from another control group comprised of household contacts of TB index patients (KTB study). As a result, our analysis could not adjust for HbA1c levels across all groups. Hence, there remains a possibility of residual confounding related to diabetes severity, which could influence some of the observed immune differences between groups. Lastly, this focused exclusively on T-cell responses to peptides derived from ESAT6 and CFP10 peptides, representing only a subset of the numerous antigens expressed by Mtb. Consequently, the findings related to T-cell responses in this study may not be fully generalizable to the overall T-cell response to Mtb.

In summary, this study advances the understanding of immune impairment in the LTBI-DM comorbidity. Type 2 DM impairs the memory phenotype and polyfunctional profiles of Mtb-specific CD4⁺ and CD8⁺ T cells, which could influence the LTBI-DM immunopathology towards increased TB disease risk.