**1. Introduction**

Tuberculosis (TB) results in an estimated 1.5 million deaths each year, making it the number one cause of death by a single infectious organism [1]. The burden of TB is disproportionately higher in Africa, where the co-infection rate with HIV is also the highest; more than 50% of the TB cases are HIV positive (+) [1]. The risk of active TB increases up to 30 times with HIV infection, with an annual risk of 15% compared to a 10% lifetime risk in HIV-uninfected participants [2]. Although antiretroviral therapy (ART) significantly reduces this risk [3], the incidence of TB while on ART remains higher than that in HIV-uninfected individuals [4,5]. HIV also specifically targets *Mycobacterium tuberculosis* (*Mtb*)-specific CD4 T-cells, resulting in impaired immune responses to TB in co-infected individuals [6,7].

Untreated HIV infection induces changes in the activation, memory, and functional profile of both CD4+ and CD8+ T-cells, with a decrease in naïve cells, and accumulation of highly differentiated cells [8]. It is hypothesized that this depletion of memory cells sustains the incidence of opportunistic infections in participants with advanced HIV. Uncontrolled replication of *Mtb* is associated with loss of CD27 expression [9], probably related to increased cellular homing to sites of disease. Recently, a ratio based on CD27 median fluorescent intensity (MFI) on CD4+ T-cells compared to that on IFN-γ+CD4<sup>+</sup> T-cells has been shown to differentiate active disease from infection [10]. However, whether this approach has good diagnostic potential in HIV+TB remains to be determined.

Studies have also shown that activated effector *Mtb*-specific CD4+ T-cells characterized by the phenotype CD38+HLA-DR+ effectively distinguish latent TB infection (LTBI) from active TB cases [11–13] Furthermore, the frequency of *Mtb*-specific CD4+ T-cells expressing CD38 and HLA-DR declined rapidly within the first month of anti-tuberculosis treatment (ATT) [11,14]. CD38 expression on bulk CD8+ T-cells has long been recognized as a marker for disease progression in HIV-infected participants [15]. In addition, activated HLA-DR+ bulk and antigen-specific CD4+ T-cells have been shown to be a risk factor for TB progression in BCG-vaccinated infants [16] and to effectively discriminate latent from active TB in adults, respectively [13,14]. However, these studies were conducted predominantly in HIV− or HIV+ ART naïve individuals. It remains unclear what the impact of combined ART and ATT on the activation profile of *Mtb*-specific T-cells is.

ART seems to partially correct the HIV-induced T-cell defects and the activation profile of bulk T-cells from patients on ART is higher than that in HIV-uninfected individuals [17]. In a study of HIV-infected women who were followed up for a year post ART initiation, viral suppression and CD4+ T-cell increase did not lead to a concomitant decline in early (CD27+CD45RO+) and terminally differentiated (CD27−CD45RO+) memory CD4+ T-cells [17]. Since the memory phenotype determines the extent to which pathogen-specific responses are restored while on ART, persistent defects in the memory subsets of *Mtb*-specific CD4+ T-cells could account for the risk of active TB while on ART [18]. Therefore, it remains unclear how long-term ART affects the normalization of *Mtb*-specific T-cell memory phenotypes. Thus, in order to understand the dynamics of *Mtb*-specific T-cells during combined ART and ATT, we assessed the functional, activation, and differentiation profile of *Mtb*-specific T-cells in HIV/TB co-infected participants on ART compared to HIV− active TB. We also assessed whether the effects of ATT on immune responses were restricted to *Mtb*-specific T-cells by adding a Gag stimulation condition as a control for HIV-specific responses.
