Concurrent TB and HIV therapies control TB reactivation during co-infection but not chronic immune activation

Sharan, Riti; Zou, Yi; Singh, Bindu; Shivanna, Vinay; Dick, Edward J.; Hall-Ursone, Shannan; Luo, Xi; Guo, Guanlin; Khader, Shabaana A.; Alvarez, Xavier; Rengarajan, Jyothi; Lai, Zhao; Mehra, Smriti; Wu, Hulin; Kaushal, Deepak

doi:10.1038/s41467-025-67188-4

Download PDF

Article
Open access
Published: 12 December 2025

Concurrent TB and HIV therapies control TB reactivation during co-infection but not chronic immune activation

Nature Communications volume 17, Article number: 499 (2026) Cite this article

6977 Accesses
1 Citations
88 Altmetric
Metrics details

Subjects

Abstract

Most HIV-negative individuals exposed to Mycobacterium tuberculosis (Mtb) control infection as latent TB infection (LTBI), but HIV co-infection greatly increases progression to tuberculosis (TB), the leading cause of death in people living with HIV (PLHIV). Although combination antiretroviral therapy (cART) reduces LTBI reactivation, immune control of Mtb is not fully restored, as shown by persistent TB incidence in PLHIV on cART. In macaques, skewed pulmonary effector memory CD4⁺ T-cell (T_EM) responses and new TB lesions persist despite cART. We hypothesize that concurrent anti-TB therapy with cART would improve bacterial control and immune restoration compared to cART alone. Using rhesus macaques (RM) with LTBI and Simian Immunodeficiency Virus (SIV) co-infection, we tested three months of weekly isoniazid and rifapentine (3HP) plus daily cART. Concurrent cART+3HP improves clinical and microbiological outcomes but fails to fully restore lung CD4⁺ T-cell immunity. Treated RMs retain caseous granulomas with high FDG uptake and incomplete CD4⁺ T-cell reconstitution, marked by persistent activation, exhaustion, and inflammation. CD4⁺ T_EM cells remain depleted. Concurrent therapy induces Type I IFN signatures and enhances Mtb-specific T_H1/T_H17—but reduces TNFα—responses. These findings reveal persistent pulmonary immune defects underlying TB risk in HIV co-infection and identify potential targets for host-directed adjunctive therapies.

CD8⁺ T cells sustain vaccination-induced immunity against dissemination of contained tuberculosis in immunosuppressed hosts

Article Open access 24 March 2026

HIV co-infection increases the risk of post-tuberculosis mortality among persons who initiated treatment for drug-resistant tuberculosis

Article Open access 11 October 2024

Survival rate and predictors of mortality among TB/HIV co-infected adult patients: retrospective cohort study

Article Open access 01 November 2022

Introduction

Combinatorial antiretroviral therapy (cART) reduces the risk of reactivation of latent tuberculosis infection (LTBI) in humans^1,2,3,4. However, tuberculosis (TB) remains a major cause of morbidity and mortality in people living with HIV (PLHIV)^5,6,7. Anti-tubercular treatment has been shown to reduce disease incidence by 30%–50%^8,9 and is recommended for PLHIV in high-burden countries. Observational studies in humans suggest that concurrent administration of cART and isoniazid preventive therapy (IPT) for LTBI lowers the risk of developing TB compared to cART alone¹⁰. A randomized double-blind placebo-controlled trial⁸ showed that administering IPT in conjunction with cART resulted in significantly lower numbers of incident TB cases than cART plus placebo. Thus, concurrent cART+IPT leads to improved outcomes with clear protective effects and clinical benefit to HIV-infected individuals. Although the WHO recommends concurrent IPT and cART in TB-endemic settings, uptake remains poor, and the immune mechanisms underlying the benefits of concurrent cART and IPT have not been defined. Another caveat of this approach has been a lack of completion of the treatment regimen by the majority of those who initiate the 6-month course of daily isoniazid while on cART. To enable treatment adherence and completion, the World Health Organization (WHO) recommended a 12-dose weekly regimen of Isoniazid and Rifapentine (3HP) as a treatment-shortened option for treating TB. Concurrent administration of cART and 3HP was safe and well-tolerated, with over 95% completion rate^11,12. In our RM model of Mycobacterium tuberculosis (Mtb)/SIV co-infection, though 3HP effectively reduced persistent LTBI¹², it did not sterilize the lungs of a third of the-treated RMs¹³. Recent work in Mtb/SIV co-infected RMs¹⁴ shows that, in addition to the depletion of CD4⁺ T cells, HIV-driven chronic immune activation correlates with LTBI reactivation^{14,15,16,17,18}. Though cART controls viral replication, it leads to insufficient reconstitution of protective CD4⁺ T effector memory (T_EM) responses in the lungs¹⁵ and fails to rescue from virus-driven immune activation¹⁹. Administration of anti-tubercular therapy (ATT), concurrently with cART, significantly reduces reactivation better than cART among individuals with LTBI¹¹. However, long-term sterilization of bacteria and immune reconstitution in the lungs has not been shown in these individuals. As such, it is important to model these conditions of concurrent treatment of Mtb/HIV co-infection with cART/ATT. Such experiments have the potential to both demonstrate experimentally long-term Mtb sterilization in this setting, as well as uncouple bacterial and viral control from immune activation.

In this study, we leveraged our nonhuman primate (NHP) model of Mtb/SIV co-infection to study the effect of simultaneous cART and 3HP treatment on immune responses to Mtb. Performing these studies in humans is virtually impossible due to the lack of determination of the timing of Mtb and HIV co-infections, verification of bacterial and viral loads and performance of invasive longitudinal studies to investigate the lung compartment. Our very low-dose aerosol infection NHP model recapitulates the spectrum of human lung pathological lesions, including LTBI and its reactivation to active TB by HIV^{13,14,15,17,20}. We carried out detailed studies of the immune responses by longitudinal sampling of bronchoalveolar lavage (BAL) in the absence of cART as well as during and after cART and cART+3HP. Importantly, we investigated the functional capacity of antigen-specific T-cell immunity in the lung microenvironment. Our findings clearly show that while cART and 3HP control viral and bacterial replication, respectively, immune reconstitution is partial and the functional capabilities of the reconstituted CD4⁺ T cells remain highly impaired. In particular, skewed CD4⁺ T_EM responses persist despite concurrent cART and anti-TB treatment in Mtb/SIV co-infected macaques and the ongoing inflammation in the lung is not ameliorated. It is well known that PLHIV and TB remain at risk for progression due to subsequent infections or TB reactivation even after improved clinical and microbiological attributes. We conclude that persisting immune activation/inflammation is the mechanisms that cause this susceptibility. Clearly, host-directed therapies against immune activation and lung inflammation, adjunctive to TB therapy and cART, must be developed to better treat PLHIV with TB.

Results

Concurrent treatment with cART and 3HP improves clinical and microbiological attributes of Mtb/SIV co-infection

To assess the impact of concurrent cART and 3HP therapy on LTBI reactivation in Mtb/SIV co-infection, we co-infected a total of 6 new Indian-origin RMs and treated them with cART+3HP for 12 weeks (n = 6). The data of RMs in groups: LTBI without SIV co-infection (n = 4), cART-naive co-infected RMs (n = 8), co-infected RMs treated with cART alone for 9 weeks (n = 4) were used from historical animals from completed studies (refs. ^15,19, Supplementary Data 1). The study design is outlined in Fig. 1A. All the RMs were infected with a low dose of Mtb (~10 CFU deposited in the lungs). cART naive, cART and cART+3HP group RMs were subsequently infected with SIV (300 TCID₅₀ SIV_mac239), intravenously. Infection was confirmed by a positive tuberculin skin test at weeks 3 and 5 after Mtb infection. All study RMs developed LTBI infection characterized by serum C-reactive protein (CRP) of 5 µg/mL or lower (Fig. 1B), no significant change in percentage body temperature (Supplemental Fig. 1A) and body weight (Supplementary Fig. 1B) up to 9 weeks after Mtb infection and <1–2 Log₁₀CFU of Mtb in the BAL at weeks 3, 5, and 7 post Mtb infection (Supplementary Fig. 1C). Upon establishment of latency at week 9 post Mtb-infection, RMs in cART naive, cART and cART+3HP groups were co-infected with 300 TCID₅₀ SIV_mac239 via the intravenous route^14,15,20. RMs in the LTBI group were not co-infected with SIV. SIV infection was confirmed by measuring the plasma viral loads via reverse transcriptase quantitative PCR (RT-qPCR). Post-SIV co-infection, all the RMs except in the LTBI group were either treated with cART alone for 9 weeks or cART+3HP, once weekly orally for 12 weeks (Fig. 1A), as described²¹. All animals were euthanized by week 24 post Mtb infection, which was the predetermined study endpoint, or at earlier time points if they became clinically unwell and/or showed signs of TB reactivation, as assessed by veterinarians in accordance with humane endpoint criteria. Clinical, pathological and immunological response was compared in the four experimental groups: LTBI, cART naive, cART and cART+3HP.

Elevated serum CRP levels associate with active TB and an increase in bacterial burdens in NHPs^15,17,20. CRP levels in RMs in the cART+3HP group were significantly lower than in the cART-treated RMs (P < 0.001) at weeks 13 (P < 0.001), 15 (P < 0.0001) and 17 post Mtb infection (P < 0.0001) (Fig. 1B). Importantly, CRP levels in cART+3HP RMs were not significantly different from the RMs in LTBI group (P = 0.44) (Fig. 1B). To evaluate the impact of cART+3HP on bacterial burden, we cultured BAL fluid, lung tissue, bronchial lymph nodes, and lung granulomas on 7H11 agar plates, following previously established protocols^14,15,17,20. Among the six RMs treated with cART+3HP, five had no detectable bacterial burden in their lungs at necropsy (Fig. 1C). In contrast, only one of four cART-treated RMs and two of 14 cART-naive RMs remained sterile at the end of the study (Fig. 1C). Therefore, RMs in the cART+3HP group (87.5% sterile) showed bacterial clearance comparable to the RMs in the LTBI group (75% sterile). Additionally, the six RMs in cART+3HP RMs had no detectable bacilli in lung granulomas (Fig. 1D), BAL (Fig. 1E) and bronchial lymph nodes (Fig. 1F) at necropsy. Notably, the bacterial burden in cART+3HP RMs was significantly lower than cART-treated RMs in lung (P = 0.01) (Fig. 1C), lung granulomas (P = 0.001) (Fig. 1D) and bronchial lymph node (P < 0.0001) (Fig. 1F). These findings suggest that cART+3HP significantly reduces bacterial burden in key sites of infection.

To assess the effectiveness of cART regimen in controlling viral replication when administered alongside 3HP treatment, viral loads were measured in the plasma of all six RMs. The viral loads in the cART+3HP-treated RMs were compared to those in the cART-only treated group at weeks 11, 13, 15 and necropsy (Supplementary Fig. 1D). No significant differences in the virus decay rate were observed between the groups, indicating that 3HP did not impact the efficacy of cART in controlling viral replication. We also evaluated cytotoxicity markers in blood samples collected at the end of the treatment to determine the safety of cART+3HP administration (Supplementary Data 2). Serum levels of the albumin/globulin (A/G) (g/dL) ratio, aspartate aminotransferase, glutamic-oxaloacetic transaminase (ALT/SGOT) (units per liter of serum), blood urea nitrogen/creatinine (BUN/creat) (µmol/L) ratio, and alkaline phosphatase (Alk phos) (units per liter) remained within normal range at treatment completion (Supplementary Data 2). To assess the impact of cART+3HP treatment on the lung pathology, lung tissue sections collected at necropsy were stained with hematoxylin and eosin (H&E) (Fig. 1G) and analyzed by board-certified pathologists. Pathological findings aligned with the clinical and microbiological observations. The cART+3HP-treated RMs displayed a few small, non-necrotizing caseous granulomas (approximately 0.5—1 cm in size) scattered across the lung lobes, with rare aggregates of lymphocytes and macrophages (Fig.1G). One RM showed multifocal lymphocyte accumulations and non-necrotizing active granulomas in the liver. In contrast, the hilar and bronchial lymph nodes, spleen, showed normal pathology, similar to LTBI-only controls. Notably, the cART+3HP-treated RMs had significantly reduced percentage lung involvement compared to the cART-only and cART naive RMs (Fig. 1H). The lung tissue of cART-only and cART-naive RMs exhibited numerous large granulomas with necrotic cores. Overall, cART+3HP administration was safe, effectively controlled the bacterial burden, and improved lung pathology compared to cART-only treatment.

We conducted Positron Emission Tomography with Computed Tomography (PET/CT), as described earlier^13,22,23, to investigate lung lesions in three of the six RMs at four time points: weeks 6 (LTBI), week 12 (LTBI + SIV co-infection, 1 week of cART+3HP treatment), week 16 (LTBI + SIV co-infection, 5 weeks of cART+3HP treatment) and week 22 (LTBI + SIV co-infection, 11 weeks of cART+3HP treatment) (Fig. 1I). Lung lesions in all RMs remained stable, showing no or minimal progression in size by week 6 after Mtb infection (Fig. 1I). The stability of these lung lesions at week 6, suggestive of LTBI, was supported by additional diagnostic parameters (positive Tuberculin Skin Test, CRP levels, and bacterial load explained in Fig. 1 and Supplementary Fig. 1C), which together confirm the LTBI diagnosis in this macaque model. At week 12, there was a significant increase (P = 0.01) in both the volume of lung lesions (Supplementary Fig. 1E) and 18F-fluorodeoxyglucose (18F-FDG) uptake (Supplementary Fig. 1F). However, by week 16 post Mtb infection (five weeks of cART+3HP treatment), 18F-FDG uptake decreased, though this reduction was not statistically significant (P = 0.003). No further increase in lesion volume (Supplementary Fig. 1E) or 18F-FDG uptake (Supplementary Fig. 1E) was observed by week 22 post Mtb infection (ten weeks of cART+3HP treatment). These PET/CT results indicate a significant decrease in lesion volume post-cART+3HP treatment, though the metabolic potential of the lesions remained unchanged, suggesting that concurrent therapy led to a gradual resolution of caseous lesions formed after SIV co-infection (week 12), while the inflammation persisted in the few remaining lesions. The continued metabolic activity within these lesions could be interpreted in several ways. While we propose that it reflects ongoing inflammation rather than active bacterial replication, the presence of PET-avid lesions despite therapy raises the possibility of undertreated active tuberculosis. The distinction between “mild active TB” and “latent” TB is complex, and our macaque model represents LTBI as observed in humans. It is possible that a more intensive treatment regimen could yield different outcomes regarding lesion resolution and metabolic activity. Future studies should examine the impact of more aggressive anti-TB therapies on both lesion size and metabolic activity in this model, to better understand whether the observed lesions reflect a failure of immune control or indicate undertreated active TB, thus offering further insights into the transition between active and latent TB.

Immune reconstitution by cART + 3HP in pulmonary compartment of Mtb/SIV co-infected RMs

Immunophenotyping of T cells was conducted to evaluate both the extent and quality of immune reconstitution in the pulmonary compartment of Mtb/SIV co-infected RMs treated with cART+3HP compared to those treated with cART alone. After 12 weeks of cART+3HP treatment, CD4⁺ T-cell frequency in the BAL was restored to levels similar to those observed in LTBI (Fig. 2A). However, in the lung tissue, CD4⁺ T-cell frequency remained significantly lower than in LTBI controls (Fig. 2B) (P = 0.0021). A significant increase in the percentage of CD8⁺ T cells was observed in the BAL of cART+3HP-treated RMs compared to cART-only-treated RMs (Fig. 2C) (P = 0.04), but no such difference was seen in the lung (Fig. 2D). Additionally, the percentage of CD8⁺ T cells in the lung was not significantly different across the LTBI, cART, and cART+3HP treatment groups in the Mtb/SIV co-infected RMs. We have previously shown that chronic immune activation drives LTBI reactivation upon SIV co-infection in RMs^14,15,20. Hence, to assess the impact of cART+3HP on T-cell activation, we studied expression of HLA-DR and CD69 on CD4⁺ T cells in BAL at week 11 post Mtb infection (or 2 weeks post SIV co-infection, prior to initiation of cART+3HP) and at necropsy (end of 12 weeks of cART+3HP treatment) in all four study groups. All Mtb/SIV co-infected groups exhibited significantly higher frequencies of HLA-DR⁺- and CD69⁺- CD4⁺ T cells at week 11 (peak viremia) compared to the LTBI group (Fig. 2E, F). cART+3HP effectively reduced the percentage of CD4⁺ T cells expressing HLA-DR and CD69 compared to cART naive RMs, but not to the levels seen in LTBI or cART-treated RMs. These findings suggest that while cART+3HP treatment can partially restore T-cell frequencies in the pulmonary compartment, it does not fully normalize immune responses in the lung, highlighting the complexity of immune reconstitution in Mtb/SIV co-infection.

Fig. 2: Treatment of Mtb/SIV co-infected RMs with cART+3HP increases migration of TH17 and TH1* cells into pulmonary compartment compared to cART naive RMs. — **Fig. 2: Treatment of *Mtb*/SIV co-infected RMs with cART+3HP increases migration of T_H17 and T_H1* cells into pulmonary compartment compared to cART naive RMs.**

High expression of the PD-1 marker on T cells is commonly associated with increased exhaustion and dysfunction in chronic infections like HIV, even with cART treatment [21, 22]. To investigate the effects of cART+3HP on T-cell exhaustion in Mtb/SIV co-infection, we assessed the percentage of PD-1⁺ T cells in BAL cells at week 11 (peak viremia) and necropsy (Fig. 2G). At necropsy, RMs treated with cART and cART+3HP showed significantly higher percentages of PD-1⁺CD4⁺ T cells compared to LTBI RMs. However, the addition of 3HP to cART did not alleviate T-cell exhaustion in the pulmonary compartment, as there was no significant difference in PD-1⁺CD4⁺ T-cell expression in BAL between cART and cART+3HP-treated RMs (Fig. 2G). This lack of improvement occurred despite near undetectable levels of Mtb and SIV at the protocol’s conclusion in the-treated RMs. Overall, our findings suggest that cART+3HP fails to control immune activation following SIV co-infection in LTBI, potentially leading to CD4⁺ T-cell exhaustion in the lungs. We hypothesize that the duration and intensity of immune activation may impair T cells’ ability to mount the typical effector responses in Mtb/SIV co-infection. Notably, no increase in macrophage turnover was observed (Fig. 2H, I). In fact, significantly lower (P < 0.05) macrophage turnover was noted in the lungs of cART+3HP-treated RMs compared to those treated with cART alone or untreated RMs (Fig. 2I). A higher number of BrDU⁺ nuclei (green) within macrophages (red), indicated by white arrows, was observed in the lungs of cART-naive and cART-treated RMs, but absent in the lungs of cART+3HP-treated RMs (Fig. 2H). In conclusion, these findings suggest that cART+3HP treatment leads to a significant reduction in macrophage turnover in the lungs of rhesus macaques, indicating that this combined therapy may impair the typical immune cell dynamics observed in LTBI. The absence of BrDU+ macrophages in the lungs of cART+3HP-treated RMs further emphasizes the potential influence of this treatment on macrophage proliferation and overall immune function in the respiratory system.

We further studied the impact of cART+3HP on T_H17 and T_H1* phenotypes in the pulmonary compartment of Mtb/SIV co-infected RMs. A significantly higher percentage of CD4⁺ T cells expressing CCR6, a regulator of migration and function of T_H17 cells was observed in BAL cells of cART and cART+3HP-treated RMs (Fig. 2J) compared to LTBI and cART naive RMs at necropsy. Similarly, we observed a significantly higher percentage of CD4⁺ T cells co-expressing CXCR3 and CCR6 in cART and cART+3HP-treated RMs compared to LTBI and cART naive RMs, in both, BAL and peripheral blood cells (Fig. 2K, L). Additionally, cART+3HP-treated RMs harbored a significantly higher percentage of CXCR3⁺CCR6⁺CD4⁺ T cells (T_H1*) in local and peripheral compartments compared to cART-treated RMs (Fig. 2K, L). These findings align with our previous observation that higher frequencies of CD4⁺ T cells co-expressing CXCR3 and CCR6 associate with bacterial control in Mtb/SIV co-infection²⁴. The T_H1* subset has been previously reported to be the most frequent Mtb-specific T-cell subset in the lungs of latent TB donors and their numbers increase when compared to healthy subjects²⁵. The higher percentage of CXCR3⁺CCR6⁺CD4⁺ T cells in local and peripheral compartments could also be attributed to cART mediated control of viral replication as CXCR3⁺CCR6⁺ cells are known to be preferential targets of HIV/SIV infection^25,26. Further, a reduction in this subset could be attributed to higher rates of LTBI reactivation. Thus, treatment of Mtb/SIV co-infected RMs with cART+3HP increases migration of T_H17 and T_H1* cells into pulmonary compartment compared to cART naive RMs.

Poor recovery of effector memory T cells by cART + 3HP in Mtb/SIV co-infected RMs

To investigate the impact of cART+3HP on functional immune reconstitution in the pulmonary compartment of Mtb/SIV co-infected RMs, we further immunophenotyped the partially replenished CD4⁺ T cells into central memory (CD28⁺/CD95⁺) (CD4⁺T_CM) and effector memory (CD28^-/CD95⁺) T cells (CD4⁺T_EM) (Supplementary Fig. 2). SIV co-infection of latent Mtb infection caused a significant increase in percentage of CD4⁺T_CM in BAL at week 11 (peak viremia prior to cART+3HP treatment) (P < 0.0001) (Fig. 3A; Supplementary Fig. 3A). The increased percentage of CD4⁺T_CM persisted during and till end of the 12-week-long concurrent cART+3HP treatment. On the contrary, a significant decline occurred in the frequency of CD4⁺T_EM in BAL at the time of peak viremia. While the frequency of this subset marginally increased at end of the treatment (Fig. 3B; Supplementary Fig. 3A), it was still significantly lower than that seen in LTBI phase of the study (week 3 post Mtb-infection) (P = 0.002). These findings align with our previous observation that cART treatment fails to fully replenish the depleted CD4⁺T_EM in BAL and lung of Mtb/SIV co-infected RMs¹⁵. Immunophenotyping of BAL CD8⁺ T cells into CD8⁺T_CM and CD8⁺T_EM showed a significant increase (P = 0.01) in percentage of CD4⁺T_CM at peak viremia (week 11 post-Mtb infection or 2 weeks post SIV co-infection). This increase was mitigated by cART+3HP as seen by marginally reduced percentage at necropsy (P = 0.01) (Fig. 3C; Supplementary Fig. 3B). No significant change was observed in percentage of CD8⁺T_EM in BAL at weeks 3, 11 and 24 (Fig. 3D; Supplementary Fig. 3B) (P = 0.2). Thus, cART+3HP expands the CD4⁺ and CD8 + T_CM but is unable to replenish CD4⁺T_EM in pulmonary compartment of Mtb/SIV co-infected RMs.

**Fig. 3: Poor recovery of effector memory T cells in RMs treated with cART or cART + 3HP.**

We further compared the restoration of CD4⁺T_CM and CD4⁺T_EM in BAL and lung of Mtb/SIV co-infected RMs treated with cART or cART+3HP (Fig. 3E–L). Despite a similar percentage of CD4⁺ T cells in BAL at necropsy, there was a significantly higher percentage (P < 0.0001) of CD4⁺T_CM in cART+3HP-treated RMs compared to cART-treated RMs (Fig. 3E). No significant difference was observed in lung CD4⁺T_CM (Fig. 3F), BAL CD4⁺T_EM (Fig. 3G) and lung CD4⁺ T_EM (Fig. 3H) between cART and cART+3HP-treated RMs. Similar to CD4⁺T_CM, cART+3HP RMs exhibited significantly higher (P = 0.009) percentage of CD8⁺T_CM in BAL (Fig. 3I) with a concurrent decrease in CD8⁺T_EM (P < 0.0001) (Fig. 3K) compared to cART-treated RMs. However, there was no significant difference between lung CD8⁺T_CM (Fig. 3J) and CD8⁺T_EM (Fig. 3L) in cART and cART+3HP-treated RMs. Overall, there were dynamic changes in the memory phenotype of CD4⁺ and CD8⁺ T cells in BAL compared to lung in cART and cART+3HP-treated RMs. BAL is a critical resource to study longitudinal changes in pulmonary immune response and has been shown to be useful to evaluate local response to therapy^27,28.

cART + 3HP increases Mtb-specific T_H1/T_H17 response in pulmonary compartment of Mtb/SIV co-infected RMs

BAL samples were collected from study RMs at weeks 5, 11 and necropsy post Mtb infection using standard operating procedures by the veterinarian. Single-cell suspensions were prepared as per the standardized lab protocol²⁹. BAL cells were stimulated ex vivo with Mtb-specific antigens, ESAT-6/CFP-10 and Mtb Cell Wall Fraction (Mtb CW) for 16 h and stained with flow cytometry antibodies to detect IFNγ, TNFα, and IL-17, as described earlier^30,31 (Supplementary Fig. 4). All Mtb-specific responses were background corrected (Supplementary Fig. 5). A significantly higher percentage of IFNγ expressing Mtb-specific CD4⁺ T cells was observed in BAL of cART+3HP-treated RMs at end of treatment when stimulated with ESAT-6/CFP-10 (Fig. 4A) (P = 0.04) and Mtb CW (Fig. 4B) (P = 0.03) compared to cART-treated RMs. We hypothesize that cART+3HP treatment effectively control bacteria thus enhancing production of protective IFNγ by Mtb-specific CD4⁺ T cells in pulmonary compartment of Mtb/SIV co-infected RMs³². In contrast to IFNγ, cART+3HP treatment resulted in a significantly lower percentage of Mtb-specific CD4⁺ T cells to produce TNFα after stimulation with either ESAT-6/CFP-10 (Fig. 4C) (P = 0.03) or Mtb CW (Fig. 4D) (P = 0.009) compared to cART-treated RMs. During Mtb infection, both IFNγ- and TNFα-producing T cells have crucial but distinct protective functions. While the former population is critical for macrophage activation and the subsequent killing of Mtb³³, the latter contributes to immunocyte recruitment and granuloma formation^33,34. It has also been reported previously that T-cell derived TNFα can promote proliferation of effector T cells resulting in increased immunogenicity^35,36. Thus, while the IFNγ producing T-cell population is restored by combinatorial treatment, the TNFα producting population is not. The antigen-specific expression of TNFα in the absence of IFNγ on CD4⁺ T cells in Mtb-infected patients has been demonstrated to strongly correlate with the potential to develop active TB, while the opposite phenotype (absence of TNFα in the presence of IFNγ on CD4⁺ T cells) promotes latent infection^37,38. The lack of increased Mtb-specific TNFα during concurrent cART+3HP treatment could however have a detrimental impact on effector T-cell function needed for sustained protection. Similar to IFNγ, a significant increase in IL-17⁺CD4⁺ T cells was also observed in BAL of cART+3HP-treated RMs upon ESAT-6/CFP-10 (Fig. 4E) (P = 0.01) or Mtb CW (Fig. 4F) (P = 0.005) stimulation, compared to cART-treated RMs. The trends were similar in lung with significantly higher percentage of CD4⁺ T cells expressing IFNγ (P = 0.04) and IL-17 (P = 0.01) when stimulated with ESAT-6/CFP-10 (Fig. 4G) or Mtb CW (Fig. 4H) compared to cART-treated RMs. While the role of T_H1 cells is clearly associated with protection in Mtb infection through IFNγ production, the role of T_H17 cells is complex and is associated with tissue damage on one hand and anti-inflammatory responses on the other. However, our findings align with the recent studies that show that Mtb-responsive IL-17- producing CD4⁺ T cells are preserved in humans with LTBI with HIV-ART and that IL-17 producing CD4⁺ T cells constitute the dominant response to Mtb antigen³⁹. Moreover, we did not observe an increase in the levels of pro-inflammatory cytokines, IL-6 and IP-10 in cART+3HP-treated RMs compared to cART-treated RMs (Fig. 4I). Overall, there is an increased T_H1/T_H17 Mtb-specific response in cART+3HP-treated RMs that associates with protection.

To better understand immune responses after concurrent cART+3HP treatment, we assessed transcriptional profiles of lung cells collected at necropsy from Mtb/SIV co-infected, cART+3HP-treated RMs by RNA sequencing (Fig. 4J). An increased expression of Type I IFN genes, such as IFNA2, IFNA1/IFNA13 was seen in cART+3HP-treated RMs (Fig. 4J). 20 out of 38 IFN associated genes from the cART+3HP dataset were upregulated (Supplementary Data 3). The role of Type I IFN in TB is ambiguous. Both human and animal studies show evidence for the role of Type I IFN in Mtb expansion and TB pathogenesis^40,41. Murine data particularly suggests that Type I IFN signaling promotes TB progression. Our own data from RMs suggests that pDC expressing Type I IFN associate with TB progression^42,43. A human blood transcriptional signature also largely comprised of Type I IFN response genes was described in TB patients⁴⁴ and validated in macaques with TB⁴⁵. Together, these results suggest a pathological role for Type I IFN in TB. Type I IFN signatures are however enriched among the lymphoid cell clusters from the lungs of Mtb-infected mice⁴⁶ and in the BAL T cells of macaques vaccinated with a leading preclinical vaccine candidate where it is associated with protection from TB³¹. Our finding of an increased Type I IFN signature therefore aligns with previously reported transcriptional signatures in humans and NHPs^42,44 and could be predictive of better lymphoid responses engendered after combinatorial treatment relative to cART alone. Our results however also suggest that while clinical disease is controlled by concurrent therapy, these animals continue to harbor molecular signatures that can be associated with immune activation and lung pathology⁴⁷.

Single-cell transcriptomic signature in pulmonary compartment of Mtb/SIV co-infected RMs

To better understand lung-specific immune responses to combinatorial treatment in Mtb/SIV co-infected macaques, we further investigated the transcriptional changes at single cell level in these macaques. We collected BAL at four critical time points from the same RMs during the study period; week 5 (represents the asymptomatic phase of Mtb infection), week 11 (represents 2 weeks post-SIV co-infection), week 13 (represents post-SIV co-infection and 4 weeks of cART treatment) and necropsy (study endpoint after 12 weeks of cART+3HP treatment) (Fig. 5A). Using this experimental design, we were able to track the early transcriptomic changes in defined populations of cells at four different stages of Mtb/SIV co-infection (Supplementary Fig. 6A, B). This negates the need for LTBI- and cART-naive controls since they are represented by week 5 and week 11 timepoints in the same animals in this study. All samples passed quality control in terms of cell quality (fraction reads in cells) and sequencing after which they were run on 10x chromium controller and the combined data filtered by removing cells with 1) > 8000 detected genes or 2) > 15% mitochondrial content (Supplementary Data 4; Supplementary Figs. 7–11). t-distributed Stochastic Neighbor Embedding (tSNE) clustering identified eight transcriptionally distinct cell clusters across all samples that could be broadly classified into lymphoid and myeloid cell subsets (Fig. 5B; Supplementary Data 5): macrophages, monocytes, CD4⁺ or CD8⁺ T cells, B cells, basophils, NK/NKT cells and mast cells (Fig. 5C). The total number of transcripts (nFeature_RNA) and molecules (nCount_RNA) detected within each cell increased in early phase of SIV co-infection compared to LTBI phase (Supplementary Fig. 7). Cells were filtered to detect genes within the range of 10-8000 to remove extreme outliers. scRNA-seq data unequivocally demonstrate a significant reduction in CD4⁺ T (Fig. 5D) (P = 0.02) or NK/NKT (Fig. 5E) (P = 0.004) cell counts and proportions following SIV co-infection (Supplementary Fig. 12A–F), which cART+3HP fails to restore to levels observed in LTBI. Contrarily, cell counts and proportions of CD8⁺ T-cell cluster (Fig. 5F) (P = 0.002), macrophage (Fig. 5G) (P = 0.008) and monocyte (Fig. 5H) (P = 0.02) significantly increased post SIV co-infection (Supplementary Fig. 12A–F).

We further studied the distributions of 42 genes across cell clusters (color-coded) when comparing three time points (week 5, week 11, and necropsy). These genes were selected from the DGE list based on their association with T_H1 (IFNG, IFNGR1, IFNGR2, TBX21, STAT1, TNFAIP2, TNFAIP3, TNFAIP8, TNFAIP8L1, TNFAIP8L3, IL2, IL2RA, and BHLHE40), T_H17 (IL20, IL20RA, IL20RB, IL21, IL21R, IL22, IL22RA1, IL22RA2, IL23A, IL23R, RORA, and STAT3), and immune activation (IDO, IRF4, CD38, CD69, LAG3, ICOS, CTLA4, IL13, CXCL10, CXCR1, CXCL12, CXCL13, IL27, KLRD1, IL6, MMP9, and FCGR1A) (Supplementary Fig. 13). A significant difference was observed in the 42 genes when compared between time points in each cell cluster (Fig. 6A–C). Module scores were computed and differences between time points were assessed for T_H1 (Fig. 6A), T_H17 (Fig. 6B) and Immune activation (Fig. 6C)—associated genes (Supplementary Data 6). Significant differential expression of the T_H1 module was observed across multiple immune cell clusters (Fig. 6D), with the most pronounced changes occurring in macrophages and monocytes (FDR < 5 × 10⁻²⁹), particularly between week 11 and necropsy conditions (e.g., macrophages: FDR = 4.83 × 10⁻¹⁷⁵; monocytes: FDR = 6.61 × 10⁻¹¹⁵), while CD4⁺ T cells, CD8⁺ T cells, and mast cells also showed significant, though comparatively moderate, changes (FDR range: 1.03 × 10⁻⁷ to 4.15 × 10⁻⁶). T_H17 module expression (Fig. 6E) was significantly altered across multiple immune cell clusters, with the most striking differences observed in macrophages, particularly between week 11 and the necropsy (FDR = 3.35 × 10⁻³⁰²), as well as other timepoints (e.g., wk5 vs N: FDR = 1.16 × 10⁻⁴⁶; wk13 vs N: FDR = 2.57 × 10⁻¹⁹). Monocytes also exhibited great T_H17-associated changes between week 11 and necropsy conditions (FDR = 1.11 × 10⁻³⁴). CD4⁺ and CD8⁺ T cells showed robust differential expression, especially at week 11 vs necropsy (e.g., CD4⁺: FDR = 8.19 × 10⁻⁸¹; CD8⁺: FDR = 2.87 × 10⁻⁶³), suggesting broad activation of the T_H17 program across lymphoid populations. Significant but more modest changes were also detected in NK/NKT cells (FDR = 0.0042) and mast cells (FDR = 2.21 × 10⁻⁶). The immune activation module (Fig. 6F) showed significant differential expression across a wide range of immune cell clusters. Macrophages exhibited the most substantial changes, particularly between wk5 and wk11 (FDR = 1.43 × 10⁻⁵⁷). Monocytes also displayed robust modulation of the immune activation module across all timepoints, with especially strong effects in the wk5 vs wk13 contrast (FDR = 5.58 × 10⁻⁷⁷) and wk11 vs necropsy (FDR = 6.40 × 10⁻⁴⁶). CD4⁺ and CD8⁺ T cells demonstrated significant immune activation module activity, notably in wk5 vs wk13 (CD4⁺: FDR = 1.08 × 10⁻¹⁷; CD8⁺: FDR = 1.99 × 10⁻⁶), while NK/NKT cells and mast cells showed modest but statistically significant changes (FDR = 0.03 and 0.032, respectively).

Fig. 6: Distribution of TH1, TH17, and immune activation-associated genes across cell clusters and time points. — **Fig. 6: Distribution of T_H1, T_H17, and immune activation-associated genes across cell clusters and time points.**

Analysis of gene expression across timepoints within individual animals suggested that most animals may follow similar temporal expression trends (Supplementary Figs. 14–21). Week 5 captures the LTBI phase prior to viral challenge, while week 11 represents the early stage of SIV co-infection—before the initiation of cART+3HP—allowing us to isolate the impact of viral infection on innate immune dynamics in the lung. This comparison is critical for understanding how the host’s early response to SIV shapes subsequent immune control and informs the effectiveness of cART+3HP. Based on findings from Fig. 6F, DGEs were identified from timecourse contrasts between week 5 versus week 11 in the macrophage cluster (Supplementary Fig. 22). Pathway analysis of DGEs in the macrophage cluster between wk5 and wk 11 revealed significant enrichment for innate immune and inflammatory responses. The GO enriched pathways include response to type I IFN, IFN-mediated signaling pathway, cellular response to type I IFN and activation of the innate immune response (Supplementary Fig. 22A). GSEA enriched pathways in cART+3HP RMs include cytokine mediated signaling pathway, response to type I IFN and innate immune response (Supplementary Fig. 22B), coupled with increased expression of immune activation—associated CD69 (Supplementary Fig. 23). This data corroborates our findings in bulkRNAseq (Fig. 4J) wherein, we observed predicted activation of IFN associated genes as well as upregulation of 20 out of 38 IFN associated genes in the cART+3HP dataset (Supplementary Data 3).

In conclusion, the focus on macrophages in this comparison allows us to highlight early immune perturbations that are highly relevant to the host-pathogen interaction in TB/HIV co-infection. The observed enrichment of type I IFN and innate immune pathways suggests a central role for these mechanisms in shaping the trajectory of infection and treatment response. Moreover, these findings complement our broader analysis of differential gene expression across T_H1, T_H17, and immune activation-associated genes, underscoring the complexity of the immune response during co-infection. Understanding how these innate responses evolve in the context of cART+3HP can provide important insights for optimizing therapeutic strategies aimed at achieving durable immune control and TB sterilization in HIV-positive individuals.

Discussion

This study reports the impact of the WHO-recommended cART+3HP treatment regimen on LTBI reactivation in rhesus macaques co-infected with Mtb and SIV. The findings offer insights into the host response to co-infection and concurrent treatment. 3HP, a once-weekly 12-week oral regimen combining high-dose isoniazid and 3HP, has been associated with significantly lower hepatotoxicity and higher completion rates than traditional isoniazid preventive treatment in humans. Notably, 3HP is a recommended regimen for treating LTBI and preventing TB in PLHIV. Recent clinical trials demonstrated that combining dolutegravir-based cART with 3HP for TB prevention is safe and effective, achieving high rates of viral suppression⁴⁸. Our previous studies on the effects of 3HP in rhesus macaques showed that while it reduced persistent Mtb infection, it did not completely sterilize the bacteria, indicating a potential risk for LTBI reactivation in TB-endemic areas¹³. It is therefore critical to model concurrent cART/ATT treatments (e.g., 3HP), in the setting of Mtb/HIV co-infection.

The NHP model is attractive for studying human Mtb infection and for performing preclinical studies on treatment regimens, as it recapitulates key aspects of human Mtb infection states and TB disease^49,50,51,52. Modeling cART and 3HP in Mtb and HIV co-infection in NHPs is crucial from a human perspective because it provides a more accurate representation of the complex interactions between the two pathogens and host responses, during co-infection, which is challenging to fully replicate in human clinical studies. NHPs, particularly rhesus macaques, offer a unique opportunity to study the immune and microbiological responses to co-infection in a controlled environment. This model allows for the evaluation of therapeutic strategies such as the simultaneous initiation of cART and anti-TB treatment, which is currently being tested in clinical trials for individuals with latent TB and HIV co-infection. Understanding the pulmonary immune responses pre-and post-therapy, and potential drug interactions in NHPs provides critical insights into the unique challenges posed by HIV co-infection. Moreover, NHP models are essential for exploring the safety, efficacy, and timing of combined therapies, ultimately leading to better clinical outcomes in people living with both HIV and TB. Although the WHO recommends concurrent 3HP and cART in TB-endemic settings, uptake remains poor and the immune mechanisms underlying the benefits of concurrent cART and 3HP have not been defined. Identifying the components of TB immunity that remain impaired after cART, versus those that are restored by concurrent cART and 3HP is an innovative approach to identify correlates of immune control versus reactivation. Such studies will also aid in identifying potential targets for host-directed therapies that further improve treatment regimens for TB/HIV and provide insights into designing therapeutic vaccines that prevent reactivation of LTBI.

Our group has previously shown that earlier initiation of cART suppresses the virus, partially reconstitutes CD4⁺ T cells, but fails to control inflammation and immune activation^15,20. In this study, we sought to determine if concurrent cART+3HP therapy initiated at early stages of co-infection better controls immune activation and dysfunction in lungs compared to cART. We utilized historical controls to minimize animal usage, adhering to standard practices in non-human primate (NHP) research that strongly encourage the use of archived samples or control cohorts to reduce the number of animals needed for repeated measures of the same outcomes. Our objective was to compare immune activation markers and T_H1, T_H17 response reconstituted in the cART and cART+3HP groups against those in the cART-naive group, and to assess whether the reconstitution was comparable to the immune responses seen in LTBI. To this end, we selected previously completed cohorts for comparison. Our findings indicate that concurrent treatment with cART and 3HP in Mtb/SIV co-infected rhesus macaques did not better restore immune function—disrupted as early as two weeks into SIV co-infection—compared to cART alone. Specifically, cART+3HP treatment failed to control virus-driven chronic immune activation in the pulmonary compartment (HLA-DR⁺CD4⁺ T and CD69⁺CD4⁺ T cells) or in the periphery (CXCR3⁺CCR6⁺CD4⁺ T cells). Moreover, the global percentage of CD4⁺ T cells and Mtb-specific CD4⁺ T_EM cells remained significantly lower at the conclusion of both cART and cART+3HP treatments, compared to LTBI-only macaques, highlighting the failure of the treatments to restore functional immune responses in the pulmonary compartment. These findings, derived from a biologically relevant preclinical animal model, hold clinical implications for human patients receiving these treatments. Administration of concurrent cART+3HP improved the clinical and microbiological attributes of Mtb/SIV co-infection compared to cART-naive or cART-treated RMs. RMs were trained to take 3HP orally, mimicking humans. As seen in the DOLPHIN study⁴⁸, our model demonstrated that co-administration of dolutegravir with 3HP was safe, well-tolerated and did not require any dose-adjustment of dolutegravir. Initiation of cART and 3HP at 2 weeks post Mtb/SIV co-infection sterilized bacterial burden in the lung of five out of six RMs and completely prevented dissemination to extra-pulmonary organs in all six RMs. The observed differences in bacterial burden and immune responses can be explained by the mechanisms underlying each group’s treatment protocol. As expected, the bacterial burden in the LTBI and cART+3HP groups were comparable, given that both rely on distinct mechanisms for bacterial control (host control in the LTBI group and drug control in the cART+3HP group). In contrast, the cART naive animals demonstrated a higher bacterial burden with extrapulmonary spread, consistent with LTBI reactivation and dissemination upon SIV co-infection. The bacterial burden in the cART-only group is also as expected, given the absence of anti-TB therapy.

There was a significant reduction in the percentage of lung involvement in pathology in cART+3HP-treated RMs with visibly fewer granulomas compared to cART-naive or cART-treated RMs. The few granulomas observed at the end of cART+3HP treatment were characterized as an equal mix of non-necrotizing and caseous type. 18F-FDG PET/CT scans revealed a significant reduction in the number of lesions post-treatment with cART+3HP but not in the uptake of 18F-FDG in the few lesions that remained at the end of treatment. Taken together, cART+3HP treatment exerts bacterial and viral control, thereby improving the health status of Mtb/SIV co-infected RMs during the study period. However, cART+3HP-treated RMs continued to harbor granulomas that have the potential to release infectious bacilli, and exhibit increased 18F-FDG uptake associated with inflammation. The main finding of this study is the incomplete immune reconstitution and persistent immune-mediated inflammation, even with LTBI treatment and concomitant cART. This is a crucial finding, and we believe it underscores the complexity of TB and HIV co-infection, particularly in the context of treatment. While cART+3HP clearly provides benefits, our findings highlight the persistent challenges in eliminating inflammation, which may continue to hinder TB responses despite the intervention.

We have previously shown that cART fails to reconstitute CD4⁺ T cells in the lung tissue to the levels seen in LTBI and that the reconstituted CD4⁺ T cells are dysfunctional for Mtb-specific responses^15,20. Concurrent administration of cART and 3HP did not further improve the frequency of reconstituted CD4⁺ T cells in the lung of Mtb/SIV co-infected RMs compared to cART-only-treated RMs. The reconstituted CD4⁺ T cells in BAL and lung of cART+3HP-treated RMs exhibited an increased frequency of activated and inflamed phenotype compared to LTBI RMs. Such CD4⁺ T-cell phenotypes correlated with high risk for TB progression⁵³. Our model, therefore, demonstrates that SIV-induced activation of pulmonary CD4+ T cells is not ameliorated by cART+3HP. A majority of reconstituted CD4⁺ T cells were central memory phenotype. On the contrary, there was a significant reduction in the CD4⁺ T_EM cell population in the pulmonary compartment post SIV co-infection that cART+3HP did not alleviate, as was also seen in cART-treated RMs. CD4⁺ T_EM cells are critical for host protection to subsequent antigen encounter. The CD4⁺ T_EM cells can produce early effector cytokines such as IFNγ and TNFα that help activate other cell types, such as CD8⁺ T cells, or they can directly kill the infected cells. It is feasible that reduced bacterial burden results in reduced antigen presentation, which can cause a reduced frequency of CD4⁺ T_EM cells in cART+3HP-treated RMs. However, chronic Mtb infection, such as a latent TB infection, is known to elicit effector memory phenotype in CD4⁺ and CD8⁺ T cells⁵⁴. Our model recapitulates this phenotype as is seen by >10% CD4⁺ T_EM in BAL collected from the same RM during the LTBI phase that reduces to less than 3% post SIV co-infection. Clearly, the presence of CD4⁺ T_EM associates with an immune balance seen in LTBI in our model and a decrease in the frequency of this cell type contributes to immune dysfunction that cART+3HP fails to mitigate.

We next determined the frequency and functionality of Mtb-specific CD4⁺ T cells in the pulmonary compartment of Mtb/SIV co-infected RMs treated with cART+3HP compared to cART. We performed ex vivo stimulation of BAL cells isolated at week 5 (this time-point represents the asymptomatic phase of Mtb infection), week 11 (this time-point represents 2 weeks post-SIV co-infection), and necropsy (after 12 weeks of cART+3HP treatment) with ESAT-6/CFP-10 and Mtb CW. Upon 12 weeks of cART+3HP treatment, an increased percentage of IFNγ and IL-17 producing Mtb-specific CD4⁺ T cells was seen in BAL and lung. Similar to what has been reported in humans, it is feasible that a majority of these T_H1/T_H17 cytokine-producing cells in BAL and lung are of central memory phenotype since CD4⁺ T_CM were the dominant cell type observed in the pulmonary compartment at the end of cART+3HP treatment⁵⁵. On the contrary, a lesser percentage of TNFα-producing Mtb-specific CD4⁺ T cells was observed at the end of cART+3HP treatment compared to cART-treated RMs. TNFα is required for granuloma organization, and inhibition of TNFα through TNFα inhibitors, result in TB reactivation⁵⁶. Hence, the skewed reconstitution of Mtb-specific response, consisting of an increased IFNγ and IL-17 response but a defective TNFα response, could prove detrimental in long-term protection, altered granuloma formation and dissemination of disease.

RNA sequencing of lung tissue collected at necropsy from cART+3HP-treated RMs showed increased type I IFN response-associated genes; “Interferon signaling”, “IFNA2”, “IFNA1/IFNA13”, “ifnar”, “interferon alpha”, “IRF9”, “IRF1” and apoptosis genes; “Apoptosis”, “Apoptosis of epithelial cells”, “cell death of progenitor cells”, “cell death of germ cells”, “Apoptosis of hematopoietic cells” compared to cART-treated RMs. Type I IFNs are critical in host defense to viruses. There is a growing body of literature that describes the detrimental impact of type I IFN in Mtb infection^57,58, including in humans^59,60. Some reports, however, implicate type I IFN in protection from TB, especially in responses engendered by protective vaccination³¹. In this regard, recently, type I IFN was shown to play a role in Mtb-induced macrophage cell death that leads to the release of bacilli from dead macrophages and dissemination. Previously, it was shown that the signaling pathways involved with type I IFN are involved in apoptosis^61,62 that could explain the concomitant increase in expression of genes associated with apoptosis in cART+3HP-treated RMs. Overall, RMs treated with cART+3HP present a distinct transcriptomic signature that associates with immune activation. A deeper analysis of immunological recovery at the single-cell level confirmed increased expression of genes associated with innate immune activation type I IFN pathways, along with cytokine-signaling cellular responses in macrophages. Concurrent with the flow cytometry data, scRNA-seq showed an increased expression of certain T_H1 and T_H17-associated genes in CD8⁺ T clusters at the end of cART+3HP treatment. CD8⁺ T-cell cluster, characterized by an activated signature, can lead to higher cytotoxic function-associated gene expression compared to CD4⁺ memory T cells, NK and B cells. One possibility could be that this increased cytotoxic gene signature in CD8⁺ T-cell cluster associates with the increased apoptotic signature seen in bulk RNA-seq since release of cytotoxic molecules by CD8⁺ T cells is known to cause apoptosis of target cells⁶³. In humans on cART, increased expression of immune activation markers on CD8⁺ T cells during chronic HIV infection associates with the inability to proliferate and increased exhaustion. The sustained dysregulation of IDO expression and downstream signaling pathways may contribute to the suppression of T_H1-mediated immune responses. Notably, cART does not appear to restore normal IDO function, suggesting that this immunoregulatory imbalance persists despite viral suppression. Overall, it is important to note that while cART+3HP effectively controls the virus and the bacilli, there is disproportionate reconstitution of memory subsets, levels of activation and exhaustion markers, as well as their functional capacity.

There are some limitations to this study. Since functional restoration of CD4⁺ and CD8⁺ T cells is a gradual process in humans, our study, with a window of ~3 months post-treatment, may not recapitulate these settings exactly. We necropsied the RMs at the end of 12-week cART+3HP treatment to match time points with previous cohorts. To study long-term immune reconstitution by cART+3HP, we are now planning future studies with extended time to necropsy post-treatment completion. Another caveat is that the model may not provide a full physiological recapitulation of human Mtb/HIV co-infection, because RMs are exposed to a supraphysiological dose of SIV. Not all humans on cART are likely to exhibit treatment failure and progression to TB reactivation. However, Mtb/HIV co-infected individuals on cART remain ~10-fold more likely to reactivate than HIV-naive people with LTBI^64,65. Humans likely develop LTBI with a substantially lower infectious dose of Mtb (1–2 CFU) than we use to infect RMs (~10–15 CFU Mtb CDC1551). RMs infected with the CDC1551 dose/strain combination exhibit control of Mtb infection akin to human LTBI, yet the dose is higher than the physiologically relevant human infectious dose. Hence, our results are indicative of the worst outcomes in co-infected humans. We infect the RMs through aerosol, the natural route of infection, mimicking humans. Mtb strain, CDC1551, allows for the development of a human TB model resulting in a latent to chronic rather than active TB disease⁴⁹. CDC1551 has also been shown to induce a protective immune response despite being similar in virulence to other lab strains⁶⁶. Thus, our model allows for an in-depth analysis of the clinical and immunological response in the lung to cART+3HP, which is possible only in a handful of research institutions worldwide. To this end, we plan to perform experiments with samples from human cohorts to compare the NHP findings. We acknowledge that the distribution of sex across the different experimental groups is imbalanced. However, given that the primary focus of this study is not specifically sex differences, we believe that while sex may influence certain aspects of the immune response, the main conclusions of the study remain valid. Nonetheless, we emphasize the need for further research to explore how sex differences interact with treatment regimens in similar models. We acknowledge the limitations inherent in animal models of LTBI, and we agree that it is crucial to highlight the distinctions between macaque models and human LTBI to properly contextualize our findings. While macaques exhibit a form of LTBI that shares key characteristics with human infection—such as the persistence of Mtb in a non-replicating state—there are notable differences. For example, macaques tend to develop a different immune response compared to humans, potentially leading to differences in the dynamics of infection and latency^67,68. Furthermore, the frequency of multiple, chronic exposures in endemic human populations may differ from the controlled exposures in macaque models, which could affect the progression and persistence of the infection. While the macaque model provides valuable insights, we recognize the need for careful extrapolation when translating these findings to human LTBI, particularly in highly endemic regions. We recognize the limitation of the number of macaques for the imaging component in this study (n = 3). Given the high cost and logistical challenges associated with conducting imaging studies in non-human primates, including the time, resources, and specialized expertise required for such experiments, we were constrained by the number of macaques that could be included. However, despite this limitation, we believe the findings provide valuable preliminary insights into the impact of cART+3HP on LTBI reactivation, and we hope that this work will lay the foundation for future studies with larger sample sizes.

In conclusion, while concurrent cART and 3HP effectively suppress the virus and bacteria, the quality of immune reconstitution in the pulmonary compartment remains significantly sub-optimal. cART+3HP treatment increases the T_H1/T_H17 response in lung, but there is incomplete restoration of protective, CD4⁺ T_EM, and replenished Mtb-specific CD4⁺ T cells are skewed in their ability to produce TNFα. Though concurrent therapy improves pathological burden, there is increased 18F-FDG uptake in the few lesions that remain despite treatment. Further, transcript analysis of lung and BAL showed an increased expression of CD38, an immune activation marker on CD8⁺ T cells, as well as of apoptotic signature characteristic of cell death. Our results clearly show that despite the mitigation of co-infection, chronic immune activation persists in the lungs of concurrently treated NHPs. Targeting the host immune response via a host-directed immunotherapy provides an opportunity to augment immunity during the short window of acute HIV-1 co-infection of Mtb. Future studies should perform testing of the safety and efficacy of host-directed therapies such as IL-21-IgFc fusion protein administration or the use of IDO-1 inhibitors concurrent with standardized therapies in tissues and organs like the lung, which are impossible to access in humans. This is critical for the development of an immune-based intervention along with cART and anti-TB therapy to control dysregulated immune responses generated during early events of HIV co-infection of LTBI and provide long-term immune reconstitution.

Methods

Study approval

All infected animals were housed under Animal Biosafety Level 3 facilities at the Southwest National Primate Research Center, where they were treated according to the standards recommended by AAALAC International and the NIH guide for the Care and Use of Laboratory Animals. The study procedures were approved by the Animal Care and Use Committee of the Texas Biomedical Research Institute.

Animal infection

This study included macaque data from completed studies^15,19,69. A total of 22 specific pathogen-free Indian-origin rhesus macaques (Macaca mulatta) (4 females and 18 males) were infected with a low dose of approximately 10 CFU M. tuberculosis CDC1551 (BEI Resources, catalog NR13649) via aerosol as described before^29,70,71,72. All the animals were housed in ABSL3 containment during the entirety of the experiment. Tuberculin Skin Test was performed at weeks 3 and 5 post TB infection to confirm infection. All the RMs were monitored for CRP, percent body weight and body temperature weekly through the study period. 18 of the LTBI RMs were then co-infected with 300 TCID₅₀ SIVmac₂₃₉ via the intravenous route 9 weeks post-TB infection^15,17,19,69 (provided by the Preston Marx Laboratory, TNPRC, Covington, Louisiana, USA). All the procedures were conducted by a board-certified veterinary clinician. The remaining four RMs served as LTBI controls for the study. The viral infection was confirmed through plasma viral loads via reverse transcription quantitative PCR (RT-qPCR). Upon confirmation of SIV infection, the 18 RMs were then divided into three groups: the first group of 8 RMs served as co-infected controls with no cART administration; the second group of 4 RMs was started on cART at 2 weeks post-SIV co-infection or 11 weeks post TB infection (cART at peak viremia), and the third group of 6 RMs started cART+3HP at 2 weeks post-SIV co-infection or 11 weeks post TB infection. 3HP was administered once weekly for 12 weeks, while cART was administered daily for the 12 weeks. The RMs in LTBI control group were euthanized at week 24 post Mtb infection. All the RMs in cART-naive group were euthanized between 17-24 weeks post Mtb infection due to clinical signs of LTBI reactivation. The RMs in the cART group were euthanized after 9 weeks of cART treatment or at week 20 post Mtb infection while the RMs in cART+3HP group were euthanized at end of 12-week treatment or at week 24 post Mtb infection. Overall, all the RMs were euthanized by week 24, which was the predetermined study endpoint, or at earlier time points if they became clinically unwell and/or showed signs of TB reactivation, as assessed by veterinarians in accordance with humane endpoint criteria.

cART + 3HP regimen

Co-infected RMs received a drug regimen consisting of 20 mg/kg of (R)-9-(2-phosphonylmethoxypropyl) adenine (PMPA, tenofovir, Gilead Sciences), 30 mg/kg of 2’, 3’-dideoxy-5-fluoro-3’-thiacytidine (FTC, emtricitabine, Gilead Sciences) and 2.5 mg/mL of the integrase inhibitor, DTG, Dolutegravir (ViiV Healthcare). The drugs were administered daily via subcutaneous injection of a cocktail of these three drugs in the vehicle kleptose at previously published doses¹⁹. Co-infected RMs also received a weekly oral dose of 15 mg/kg isoniazid and 15 mg/kg 3HP for 12 weeks beginning week 12 after aerosol infection up to week 23 post-TB infection. Oral intake was monitored by veterinary staff to ensure consumption.

PET/CT imaging

Longitudinal CT and PET/CT scans were performed using MEDISO’s LFER150 PET-CT scanner at 3–6 week intervals, starting from week 6 post-Mtb infection, with the last scan prior to necropsy⁷³. Briefly, we performed 18F-fluorodeoxyglucose (FDG) PET/CT scans for each anesthetized RM using the breath-hold technique. RMs were anesthetized and intubated under the supervision of a board-certified veterinarian as per approved IACUC protocols. All the RMs received an intravenous injection of 1 mCi per kg of body weight dose of 18F-FDG⁷⁴, procured from Cardinal Health radiopharmacy. The single field of view (FOV) and/or double FOV lung CT scans were performed using breath-hold as described⁷⁵. PET scans were acquired after completion of the 40–50 min FDG uptake period. Images were visualized using Interview Fusion 3.03 (Mediso) and reconstructed using Nucline NanoScan LFER 1.07 (Mediso) with parameters as described⁷⁶. The lung segmentation, volumetric and SUV analysis was performed using Vivoquant 4.0 (Invicro, USA)⁷³.

Viral load and bacterial burden measurement

Bacterial burden in BAL was measured throughout the study period as previously described¹⁷. Viable Mtb burden was also measured at necropsy in BAL, lung, spleen, bronchial lymph node and individual granulomas collected at necropsy^17,69. Viral loads in acellular BAL supernatant and plasma were determined by RT-qPCR at peak viremia (2 weeks post-SIV or 11 weeks post TB-infection), week 13, week 15 post-Mtb infection and at necropsy. The measurements were performed by NIAID, DAIDS, NHP Core Virology Laboratory for AIDS Vaccine Research and Development). Qiagen QISymphony DSP Virus/Pathogen Midi Kit (96)/QIAgility Applied Biosystems StepOne Plus Quantitative Real-Time PCR was used to measure the viral loads. A lower limit of 100 copies/ sample was set for quantification of SIV copies in this assay.

High parameter flow cytometry

High parameter flow cytometry was performed on BAL cells at pre-infection, pre-SIV (wk 3, 5), post-SIV, pre-cART (wk 11), post-cART (wk 20 or necropsy) and post-cART+3HP (wk 24 or necropsy). Lung, bronchial lymph nodes and granulomas were harvested at necropsy and processed as described earlier^15,17,69. Briefly, the tissues and granulomas were dematricized, filtered, RBC lysis was performed, and cell counts were obtained using an automated cell counter. The lung tissue was digested with Liberase TM/DL (Sigma-Aldrich), lung lysate was filtered, RBC lysed and cells were counted using an automated cell counter. The BAL was filtered through a 100 μm filter, RBCs were lysed, and cell counts were also measured with an automated cell counter. A total of 1 × 10⁶ single cells were stained with surface markers to analyze different T-cell phenotypes (Supplementary Data 7). The single-cell suspension from tissues and BAL was incubated with the antibody cocktail for 25 minutes, washed, and fixed in BD Fixation and Stabilization Buffer. The samples were then run on the BD Symphony. To study the effect of treatment on T-cell function, the freshly collected BAL cells were stimulated ex vivo with Mtb-specific antigens, ESAT-6/CFP-10 and Mtb Cell Wall Fraction (BEI Resources, 10 μg/mL) for a total of 16 h. Brefeldin A (0.5 μg/mL, SIGMA) was added 2 h after the onset of stimulation. After stimulation, the cells were stained with LIVE/DEAD fixable Near-IR stain (ThermoFisher) and stained subsequently with the surface antibodies: CD4-PerCP-Cy5.5 (BD Biosciences), CD8-APC (BD Biosciences), CD3-AlexaFlour 700 (BD Biosciences), CD95-BV421 (BD Biosciences), CD28-PECy7 (BD Biosciences) and CD45-BUV395 (BD Biosciences). Cells were then fixed, permeabilized and stained with intracellular antibodies: IFNγ - APC-Cy7 (Biolegend), IL-17-BV605 (Biolegend) and TNFα - BV650 (Biolegend). Cells were washed, suspended in BD stabilizing fixative buffer and acquired on BD FACS Symphony flow cytometer. Analysis was performed using FlowJo (v10.6.1) using previously published gating strategy^15,17,19,72.

Gross pathology

The animals were euthanized for necropsy, and lung lobes, spleen, liver, and bronchial lymph nodes were collected. All the tissues were weighed at the time of collection. Tissues were fixed in 10% neutral-buffered formalin, paraffin-embedded, sectioned at 5 μm thickness and stained with hematoxylin and eosin (H&E) using standard methods. Lung tissues were collected stereologically at necropsy by a board-certified veterinary pathologist. The H&E-stained slides were scanned in Zeiss Axio Scan Z1, and the images were analyzed using HALO 4.0 software. HALO scores served as an indication for the true percentage of lung affected (primary and secondary lesions from TB). The lesions in each lung lobe were scored for pleural thickening, intralobular septae inflammation, perivasculitis, pneumocyte hyperplasia and lymphadenitis. The score was then utilized to annotate the different disease types: active non-necrotizing, active suppurative, active caseous, latent sclerotic, and latent fibrocalcific.

Immunohistochemistry staining

Fluorescent immunohistochemistry was performed on formalin-fixed, paraffin-embedded lung and bronchial lymph node tissues as previously described^{15,16,19,69,77}. The stained slides were scanned in the Axio Scan Z1, and the images were analyzed using HALO software. Immunohistochemistry was performed to study the impact of cART or cART+3HP on macrophage proliferation by staining BrDU⁺CD163⁺CD68⁺ macrophages in the lungs of the cART-naive, cART and cART+3HP groups (Fig. 2K). The presence of BrDU⁺ nuclei (green) within macrophages (red), as indicated by the white arrows, was observed in the lungs of cART naive and cART only treated macaques.

BAL cell isolation

BAL was performed with 80 mL of sterile saline in BSL3 on NHPs at weeks 5, 11, 15 and necropsy after Mtb infection. The BAL was filtered, centrifuged 400 × g, 10 min at 4 °C), lysed for red blood cells (RBCs) using ACK lysis buffer (Gibco^TM) and resuspended in 2 mL final volume of RPMI 1640 media (Gibco^TM). The cells were counted and frozen in Cryostor in internally threaded cryovials in liquid nitrogen for downstream scRNA-seq processing using 10× Genomics platform. The BAL cells were frozen in Cryostor (Sigma-Aldrich). Cryostor is formulated to contain 2% dimethyl sulfoxide (DMSO). Cryostor is designed to preserve cell viability and functionality through modulation of cellular biochemical response to the cryopreservation process. Our lab routinely utilizes this reagent to preserve freshly isolated BAL cells.

Quality control for frozen BAL cells

Prior to running the BAL cells on 10× Genomics platform, the cells were analyzed for viability using i) automated cell counters, ii) manual counts using Trypan Blue and iii) microscopic evaluation. Briefly, cells were thawed on ice. 100 μL of cells was washed once in 1 mL warmed 1× phosphate-buffered saline (PBS) (Gibco), centrifuged, and resuspended in 1 mL of 1× PBS. Cells were mixed in a 1:1 ratio with Trypan blue and counted in an automated Countess as well as by a hemocytometer. Cellular morphology, including shape and size, was determined using a standard bright-field light microscope. Institutionally approved protocols were applied when removing samples from BSL3.

Single-cell RNA Library generation and sequencing

BAL cell suspensions were loaded onto Chromium instrument (10x Genomics) to generate single-cell beads in emulsion. Single-cell RNA-seq libraries were then prepared using Single-Cell 3’ Gel bead and library kit version 3.1 (10× Genomics). Single-cell barcoded cDNA libraries were quantified and sequenced on an Illumina NovaSeq 6000. Read lengths were 28bp for read 1, 10 bp for index 1, 10 bp for index 2, and 100 bp for read 2. Cells were sequenced to about 50,000 reads per cell.

Single-cell data analysis

Cell Ranger Single-Cell Software suite (V7.0.1) from 10x was used to perform sample demultiplexing and generate fastq files. The resulting fastq files were aligned against reference genome mmul10 (Genebank, https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_003339765.1/) with cellranger count. The targeted cell recovery per sample was set to 10,000 cells. The cellranger counting results for 24 samples were further integrated and analyzed by R software with package Seurat (V5.0.1). The data matrix for each sample was read by Read10X. The Seurat objects of all 24 sample data were created and merged. The combined data were filtered by removing cells that had >8000 detected genes or more than 15% mitochondrial contents. Then the data were normalized by method “LogNormalize” with scale.factor=1e6, and the data were further filtered by keeping cells with CD45(PTPRC) expression as the final dataset for single cell clustering analysis. The most variable genes were detected by the FindVariableFeatures function with nfeatures=2000, then data were scaled by ScaleData, and principal component analysis was performed by RunPCA with npcs=50. The merged and filtered data were integrated by the function ‘IntegrateLayers’ with method “CCAIntegration” and parameters orig.reduction = “pca”, k.weight = 50. Data clustering was run by FindNeighbors (“integrated.cca”, dims = 1:32) and FindClusters (resolution=0.155) to pre-select 15 clusters. To visualize the data, the RunTSNE dimensionality reduction was performed by reduction = “integrated.cca”, dims = 1:32. To annotate the clusters, the package clustermole was applied and combined with the generally known immune makers. Differential gene expression between different timeCourse for each cluster were firstly identified using FindAllMarkers function in Seurat R package by (only.pos = FALSE, min.pct = 0.0, logfc.threshold = 0.0), then the marker genes were further studied and evaluated by different cut-off. GO enrichment analyses for all clusters for week 5 versus week 11 are provided in Supplementary Data 8–15. Differential gene expression for each cluster is provided in Supplementary Data 16–23. Only genes with a raw p value < 0.05 and a log2 fold change >1 were retained for subsequent Gene Ontology (GO) and KEGG pathway enrichment analyses. In these analyses, enriched terms were defined as those with an FDR (Benjamini-Hochberg) corrected p value < 0.05 and a Storey’s q value < 0.05. All enrichment analyses were performed using the Bioconductor package clusterProfiler.

Bulk RNA sequencing

Lung tissue pieces from RMs were weighed for genomic DNA extraction, transferred to a petri dish, and tissue dissociation was performed using a tissue homogenizer. The Chelex method was used for Mtb DNA extraction from macaque lung tissue. Post dissociation, the lung samples were resuspended in 2 mL tubes containing lysing matrix B and Chelex. The bacteria were lysed mechanically using a bead beater for 30s at 3800 rpm for 3 cycles. The tubes were transferred to a heating block and boiled at 100 degrees Celsius for 15 mins. This was followed by water bath sonication for 15 mins. The prepared samples were then used for stranded mRNA-Seq library prep, and 75 bp PE sequencing on Illumina NextSeq 500 Mid Output Platform. RNA quality control, library, and Agilent quality control were also performed. Qiagen’s Ingenuity Pathway Analysis (IPA) software was used to analyze the RNA-seq data to identify significant pathways. Raw, demultiplexed.fastq sequence files were uploaded to Partek Flow analysis software for standard RNA-seq analysis, including differential gene expression analysis. Briefly, bases with a Phred score less than Q30 were trimmed, aligned to RefSeq Mmul10 reference assembly (GCF_003339765.1) with STAR aligner v2.7.8a using default settings, and quantified using the Quantify to annotation model (Partek E/M) tool with the RefSeq Mmul10 annotation release as the transcript model. Raw gene counts were normalized by median ratio (transformed on samples), and subsequent DESeq2 differential expression analysis was carried out by comparing cART-treated vs. cART+3HP-treated samples with the cART+3HP-treated samples as reference levels. The details of the analysis are now provided in Supplementary Data 24. QC results are provided in Supplementary Data 25. Differentially expressed gene lists and gene set analysis results are now included in Supplementary Data 26. A list of enriched canonical pathways, along with associated p values for the cART+3HP samples, is provided in Supplementary Data 27. Accession for these SRA data: PRJNA1327945.

Statistics and reproducibility

All statistical analyses were performed using appropriate methods as indicated in the figure legends. Sample sizes were based on prior experience with rhesus macaque studies of similar design and were sufficient to detect biologically meaningful differences in immune responses. No data were excluded from the analyses. The experiments were not randomized. The investigators were not blinded to group allocation during the experiments or outcome assessments. All attempts at replication were successful, and consistent results were obtained across biological replicates. Data and analyses have been verified for accuracy and reproducibility using independent computational pipelines.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The scRNA-seq data generated in this study have been deposited in the NCBI Gene Expression Omnibus database under accession code GSE293173. The scRNA-seq data generated in this study are provided in the Supplementary Information/Source Data file. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE293173. Accession for Bulk RNA-seq SRA data: PRJNA1327945. https://trace.ncbi.nlm.nih.gov/Traces/study/?acc=PRJNA1327945%20&o=acc_s%3Aa Source data are provided with this paper.

References

Wong, N. S. et al. A longitudinal study on latent TB infection screening and its association with TB incidence in HIV patients. Sci. Rep. 9, 10093 (2019).
Article ADS PubMed PubMed Central Google Scholar
Dravid, A. et al. Incidence of tuberculosis among HIV infected individuals on long term antiretroviral therapy in private healthcare sector in Pune, Western India. BMC Infect. Dis. 19, 714 (2019).
Article ADS PubMed PubMed Central Google Scholar
Ahmed, A., Mekonnen, D., Shiferaw, A. M., Belayneh, F. & Yenit, M. K. Incidence and determinants of tuberculosis infection among adult patients with HIV attending HIV care in north-east Ethiopia: a retrospective cohort study. BMJ Open 8, e016961 (2018).
Article PubMed PubMed Central Google Scholar
Liu, E. et al. Tuberculosis incidence rate and risk factors among HIV-infected adults with access to antiretroviral therapy. AIDS 29, 1391–1399 (2015).
Article CAS PubMed Google Scholar
Lawn, S. D., Myer, L., Bekker, L. G. & Wood, R. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control. AIDS 20, 1605–1612 (2006).
Article PubMed Google Scholar
Suthar, A. B. et al. Antiretroviral therapy for prevention of tuberculosis in adults with HIV: a systematic review and meta-analysis. PLoS Med. 9, e1001270 (2012).
Article CAS PubMed PubMed Central Google Scholar
Adhikari, N. et al. Prevalence and associated risk factors for tuberculosis among people living with HIV in Nepal. PLoS One 17, e0262720 (2022).
Article CAS PubMed PubMed Central Google Scholar
Danel, C. et al. A trial of early antiretrovirals and isoniazid preventive therapy in Africa. N. Engl. J. Med. 373, 808–822 (2015).
Article CAS PubMed Google Scholar
Badje, A. et al. Effect of isoniazid preventive therapy on risk of death in west African, HIV-infected adults with high CD4 cell counts: long-term follow-up of the Temprano ANRS 12136 trial. Lancet Glob. Health 5, e1080–e1089 (2017).
Article PubMed Google Scholar
Rangaka, M. X. et al. Isoniazid plus antiretroviral therapy to prevent tuberculosis: a randomised double-blind, placebo-controlled trial. Lancet 384, 682–690 (2014).
Article CAS PubMed PubMed Central Google Scholar
Semitala, F. C. et al. Completion of isoniazid-rifapentine (3HP) for tuberculosis prevention among people living with HIV: Interim analysis of a hybrid type 3 effectiveness-implementation randomized trial. PLoS Med. 18, e1003875 (2021).
Article CAS PubMed PubMed Central Google Scholar
Chaisson, L. H. et al. Viral suppression among adults with HIV receiving routine dolutegravir-based antiretroviral therapy and 3 months weekly isoniazid-rifapentine. AIDS 37, 1097–1101 (2023).
Article CAS PubMed Google Scholar
Sharan, R. et al. Isoniazid and rifapentine treatment effectively reduces persistent M. tuberculosis infection in macaque lungs. J. Clin. Invest. 132, e161564 (2022).
Bucsan, A. N. et al. Mechanisms of reactivation of latent tuberculosis infection due to SIV co-infection. J. Clin. Invest. 129, 5254–5260 (2019).
Sharan, R. et al. Antiretroviral therapy timing impacts latent tuberculosis infection reactivation in a Mycobacterium tuberculosis/SIV coinfection model. J. Clin. Invest. 132, e153090 (2022).
Kuroda, M. J. et al. High turnover of tissue macrophages contributes to tuberculosis reactivation in simian immunodeficiency virus-infected Rhesus Macaques. J. Infect. Dis. 217, 1865–1874 (2018).
Article CAS PubMed PubMed Central Google Scholar
Foreman, T. W. et al. CD4+ T-cell-independent mechanisms suppress reactivation of latent tuberculosis in a macaque model of HIV coinfection. Proc. Natl. Acad. Sci. USA 113, E5636–E5644 (2016).
Article CAS PubMed PubMed Central Google Scholar
Sharan, R. et al. Chronic immune activation in TB/HIV co-infection. Trends Microbiol. 28, 619–632 (2020).
Article CAS PubMed PubMed Central Google Scholar
Ganatra, S. R. et al. Anti-retroviral therapy does not reduce tuberculosis reactivation in a tuberculosis-HIV co-infection model. J. Clin. Invest. 130, 5171–5179 (2020).
Ganatra, S. R. et al. Antiretroviral therapy does not reduce tuberculosis reactivation in a tuberculosis-HIV coinfection model. J. Clin. Invest. 130, 5171–5179 (2020).
Article CAS PubMed PubMed Central Google Scholar
Sharan, R. et al. Antiretroviral therapy timing impacts latent tuberculosis infection reactivation in a tuberculosis/simian immunodeficiency virus coinfection model. J. Clin. Invest. 132, e153090 (2021).
Singh, B. et al. Inhibition of indoleamine dioxygenase leads to better control of tuberculosis adjunctive to chemotherapy. JCI Insight. 8, e163101 (2023).
Singh, B. et al. Indoleamine-2,3-dioxygenase inhibition improves immunity and is safe for concurrent use with cART during Mtb/SIV coinfection. JCI Insight. 9, e179317 (2024).
Shanmugasundaram, U. et al. Pulmonary Mycobacterium tuberculosis control associates with CXCR3- and CCR6-expressing antigen-specific Th1 and Th17 cell recruitment. JCI Insight. 5, e137858 (2020).
Arlehamn, C. L. et al. Transcriptional profile of tuberculosis antigen-specific T cells reveals novel multifunctional features. J. Immunol. 193, 2931–2940 (2014).
Article CAS PubMed Google Scholar
Gosselin, A. et al. Peripheral blood CCR4+CCR6+ and CXCR3+CCR6+CD4+ T cells are highly permissive to HIV-1 infection. J. Immunol. 184, 1604–1616 (2010).
Article CAS PubMed Google Scholar
Dawson, R. et al. Immunomodulation with recombinant interferon-gamma1b in pulmonary tuberculosis. PLoS ONE 4, e6984 (2009).
Article ADS PubMed PubMed Central Google Scholar
Schwander, S. & Dheda, K. Human lung immunity against Mycobacterium tuberculosis: insights into pathogenesis and protection. Am. J. Respir. Crit. Care Med. 183, 696–707 (2011).
Article CAS PubMed Google Scholar
Mehra, S. et al. Reactivation of latent tuberculosis in rhesus macaques by coinfection with simian immunodeficiency virus. J. Med. Primatol. 40, 233–243 (2011).
Article PubMed PubMed Central Google Scholar
Sharan, R., Singh, D. K., Rengarajan, J. & Kaushal, D. Characterizing early T cell responses in nonhuman primate model of tuberculosis. Front. Immunol. 12, 706723 (2021).
Article CAS PubMed PubMed Central Google Scholar
Singh, D. K. et al. Prevention of tuberculosis in cynomolgus macaques by an attenuated Mycobacterium tuberculosis vaccine candidate. Nat. Commun. 16, 1957 (2025).
Article ADS CAS PubMed PubMed Central Google Scholar
Li, G. et al. Anti-tuberculosis (TB) chemotherapy dynamically rescues Th1 and CD8+ T effector levels in Han Chinese pulmonary TB patients. Microbes Infect. 22, 119–126 (2020).
Article PubMed Google Scholar
Cavalcanti, Y. V., Brelaz, M. C., Neves, J. K., Ferraz, J. C. & Pereira, V. R. Role of TNF-alpha, IFN-gamma, and IL-10 in the development of pulmonary tuberculosis. Pulm. Med. 2012, 745483 (2012).
Article PubMed PubMed Central Google Scholar
Allie, N. et al. Prominent role for T cell-derived tumour necrosis factor for sustained control of Mycobacterium tuberculosis infection. Sci. Rep. 3, 1809 (2013).
Article ADS PubMed PubMed Central Google Scholar
Olsen, A. et al. Targeting Mycobacterium tuberculosis tumor necrosis factor alpha-downregulating genes for the development of antituberculous vaccines. mBio 7, e01023-15 (2016).
Mehta, A. K., Gracias, D. T. & Croft, M. TNF activity and T cells. Cytokine 101, 14–18 (2018).
Article CAS PubMed Google Scholar
Pollock, K. M. et al. T-cell immunophenotyping distinguishes active from latent tuberculosis. J. Infect. Dis. 208, 952–968 (2013).
Article PubMed PubMed Central Google Scholar
Harari, A. et al. Dominant TNF-α+ Mycobacterium tuberculosis-specific CD4+ T cell responses discriminate between latent infection and active disease. Nat. Med. 17, 372–376 (2011).
Article CAS PubMed PubMed Central Google Scholar
Ogongo, P. et al. High-parameter phenotypic characterization reveals a subset of human Th17 cells that preferentially produce IL17 against M. tuberculosis antigen. Front. Immunol. 15, 1378040 (2024).
Moreira-Teixeira, L., Mayer-Barber, K., Sher, A. & O’Garra, A. Type I interferons in tuberculosis: Foe and occasionally friend. J. Exp. Med. 215, 1273–1285 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kotov, D. I. et al. Early cellular mechanisms of type I interferon-driven susceptibility to tuberculosis. Cell 186, 5536–5553 e5522 (2023).
Article CAS PubMed PubMed Central Google Scholar
Esaulova, E. et al. The immune landscape in tuberculosis reveals populations linked to disease and latency. Cell Host Microbe 29, 165-178.e168 (2020).
Scott, N. R. et al. S100A8/A9 regulates CD11b expression and neutrophil recruitment during chronic tuberculosis. J. Clin. Invest. 130, 3098–3112 (2020).
Article CAS PubMed PubMed Central Google Scholar
Berry, M. P. et al. An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature 466, 973–977 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Ahmed, M. et al. Immune correlates of tuberculosis disease and risk translate across species. Sci. Transl. Med. 12, eaay0233 (2020).
Akter, S. et al. Mycobacterium tuberculosis infection drives a type I IFN signature in lung lymphocytes. Cell Rep. 39, 110983 (2022).
Article CAS PubMed PubMed Central Google Scholar
Chowdhury, C. S. et al. Type I IFN-mediated NET release promotes Mycobacterium tuberculosis replication and is associated with granuloma caseation. Cell Host Microbe 32, 2092–2111 e2097 (2024).
Article CAS PubMed PubMed Central Google Scholar
Dooley, K. E. et al. Once-weekly rifapentine and isoniazid for tuberculosis prevention in patients with HIV taking dolutegravir-based antiretroviral therapy: a phase 1/2 trial. Lancet HIV 7, e401–e409 (2020).
Article PubMed Google Scholar
Kaushal, D., Mehra, S., Didier, P. J. & Lackner, A. A. The non-human primate model of tuberculosis. J. Med Primatol. 41, 191–201 (2012).
Article CAS PubMed PubMed Central Google Scholar
Scanga, C. A. & Flynn, J. L. Modeling tuberculosis in nonhuman primates. Cold Spring Harb. Perspect. Med 4, a018564 (2014).
Article PubMed PubMed Central Google Scholar
Flynn, J. L., Gideon, H. P., Mattila, J. T. & Lin, P. L. Immunology studies in non-human primate models of tuberculosis. Immunol. Rev. 264, 60–73 (2015).
Article CAS PubMed PubMed Central Google Scholar
Gideon, H. P. et al. Multimodal profiling of lung granulomas in macaques reveals cellular correlates of tuberculosis control. Immunity 55, 827–846.e810 (2022).
Article CAS PubMed PubMed Central Google Scholar
Mpande, C. A. M. et al. Immune profiling of Mycobacterium tuberculosis-specific T cells in recent and remote infection. EBioMedicine 64, 103233 (2021).
Article CAS PubMed PubMed Central Google Scholar
Counoupas, C. & Triccas, J. A. The generation of T-cell memory to protect against tuberculosis. Immunol. Cell Biol. 97, 656–663 (2019).
Article CAS PubMed Google Scholar
Gehad, A. et al. A primary role for human central memory cells in tissue immunosurveillance. Blood Adv. 2, 292–298 (2018).
Article CAS PubMed PubMed Central Google Scholar
Robert, M. & Miossec, P. Reactivation of latent tuberculosis with TNF inhibitors: critical role of the beta 2 chain of the IL-12 receptor. Cell Mol. Immunol. 18, 1644–1651 (2021).
Article CAS PubMed PubMed Central Google Scholar
Mundra, A. et al. Pathogenicity of type I interferons in Mycobacterium tuberculosis. Int. J. Mol. Sci. 24, 3919 (2023).
McNab, F., Mayer-Barber, K., Sher, A., Wack, A. & O’Garra, A. Type I interferons in infectious disease. Nat. Rev. Immunol. 15, 87–103 (2015).
Article CAS PubMed PubMed Central Google Scholar
Mayer-Barber, K. D. & Yan, B. Clash of the cytokine titans: counter-regulation of interleukin-1 and type I interferon-mediated inflammatory responses. Cell Mol. Immunol. 14, 22–35 (2017).
Article CAS PubMed Google Scholar
Moreira-Teixeira, L. et al. Type I IFN exacerbates disease in tuberculosis-susceptible mice by inducing neutrophil-mediated lung inflammation and NETosis. Nat. Commun. 11, 5566 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Xu, G., Wang, J., Gao, G. F. & Liu, C. H. Insights into battles between Mycobacterium tuberculosis and macrophages. Protein Cell 5, 728–736 (2014).
Article CAS PubMed PubMed Central Google Scholar
Apelbaum, A., Yarden, G., Warszawski, S., Harari, D. & Schreiber, G. Type I interferons induce apoptosis by balancing cFLIP and caspase-8 independent of death ligands. Mol. Cell Biol. 33, 800–814 (2013).
Article CAS PubMed PubMed Central Google Scholar
Prezzemolo, T. et al. Functional signatures of human CD4 and CD8 T cell responses to mycobacterium tuberculosis. Front Immunol. 5, 180 (2014).
Article PubMed PubMed Central Google Scholar
Lawn, S. D., Gupta, A. & Wood, R. Assessing the impact of prevalent tuberculosis on mortality among antiretroviral treatment initiators: accurate tuberculosis diagnosis is essential. AIDS 26, 1730–1731 (2012).
Article PubMed Google Scholar
Gupta, A., Wood, R., Kaplan, R., Bekker, L. G. & Lawn, S. D. Tuberculosis incidence rates during 8 years of follow-up of an antiretroviral treatment cohort in South Africa: comparison with rates in the community. PloS ONE 7, e34156 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Manca, C. et al. Mycobacterium tuberculosis CDC1551 induces a more vigorous host response in vivo and in vitro, but is not more virulent than other clinical isolates. J. Immunol. 162, 6740–6746 (1999).
Article CAS PubMed Google Scholar
Dijkman, K. et al. Disparate tuberculosis disease development in macaque species is associated with innate immunity. Front. Immunol. 10, 2479 (2019).
Article CAS PubMed PubMed Central Google Scholar
Barreiro, L. B., Marioni, J. C., Blekhman, R., Stephens, M. & Gilad, Y. Functional comparison of innate immune signaling pathways in primates. PLoS Genet. 6, e1001249 (2010).
Article CAS PubMed PubMed Central Google Scholar
Bucşan, A. N. et al. Mechanisms of reactivation of latent tuberculosis infection due to SIV coinfection. J. Clin. Invest. 129, 5254–5260 (2019).
Article PubMed PubMed Central Google Scholar
Mehra, S. et al. The Mycobacterium tuberculosis stress response factor SigH is required for bacterial burden as well as immunopathology in primate lungs. J. Infect. Dis. 205, 1203–1213 (2012).
Article PubMed PubMed Central Google Scholar
Mehra, S. et al. Granuloma correlates of protection against tuberculosis and mechanisms of immune modulation by Mycobacterium tuberculosis. J. Infect. Dis. 207, 1115–1127 (2013).
Article CAS PubMed Google Scholar
Kaushal, D. et al. Mucosal vaccination with attenuated Mycobacterium tuberculosis induces strong central memory responses and protects against tuberculosis. Nat. Commun. 6, 8533 (2015).
Article ADS CAS PubMed Google Scholar
Singh, D. K. et al. Responses to acute infection with SARS-CoV-2 in the lungs of rhesus macaques, baboons and marmosets. Nat. Microbiol. 6, 73–86 (2021).
Article CAS PubMed Google Scholar
Stammes, M. A. et al. Recommendations for standardizing thorax PET-CT in non-human primates by recent experience from macaque studies. Animals 11, 204 (2021).
Mattila, J. T. et al. Positron emission tomography imaging of macaques with tuberculosis identifies temporal changes in granuloma glucose metabolism and integrin α4β1-expressing immune cells. J. Immunol. 199, 806–815 (2017).
Article CAS PubMed Google Scholar
Sakai, S. et al. Functional inactivation of pulmonary MAIT cells following 5-OP-RU treatment of non-human primates. Mucosal Immunol. 14, 1055–1066 (2021).
Article CAS PubMed PubMed Central Google Scholar
Li, Q., Skinner, P. J., Duan, L. & Haase, A. T. A technique to simultaneously visualize virus-specific CD8+ T cells and virus-infected cells in situ. J. Vis. Exp. 1561 (2009).

Download references

Acknowledgements

This work was supported by National Institutes of Health investigator- AI111943 (awarded to D.K.), AI123047 (to D.K.), OD031898 (to D.K.), AI170148 (to D.K.), R21AI170148 (to R.S.), K01OD031898 (to R.S.), institutional- grants OD010442, AI168439, AI161943, OD028732, OD032443 and CPRIT Core Facility Award (RP220662). We also acknowledge the role of institutional grants U42OD10442 (to D.K.), P30AI168439 (Texas D-CFAR, co-director D.K., and P30AI161943 (IN-TRAC; both R.S. and D.K. are members). SIVmac₂₃₉ was graciously provided by Drs. Preston Marx and Nick Manness, Tulane National Primate Research Center. SIV viral load assays were performed by the NHP Core Virology Laboratory for AIDS Research and Development, Division of AIDS, NIAID. PMPA and FTC were provided by Gilead Sciences, and DTG was provided by Viiv Healthcare. Data was generated in the Genome Sequencing Facility, which is supported by UT Health San Antonio, NIH Shared Instrument grant S10OD030311 (S10 grant to NovaSeq 6000 System), S10OD028732 (to SNPRC).

Author information

Bindu Singh
Present address: Department of Biological Sciences, Indian Institute of Science Education and Research, Bhopal, MP, India

Authors and Affiliations

Southwest National Primate Research Center, Texas Biomedical Research Institute, San Antonio, TX, USA
Riti Sharan, Bindu Singh, Vinay Shivanna, Edward J. Dick Jr., Shannan Hall-Ursone, Xavier Alvarez, Smriti Mehra & Deepak Kaushal
Greehey Children’s Cancer Research Institute, Long School of Medicine, UT Health, San Antonio, TX, USA
Yi Zou
UTHealth Houston School of Public Health, Houston, TX, USA
Xi Luo, Guanlin Guo & Hulin Wu
Bernard and Betty Roizman Professor and Chair, University of Chicago Department of Microbiology and Immunology, Chicago, IL, USA
Shabaana A. Khader
Emory Vaccine Center and Emory National Primate Research Center, Emory University School of Medicine, Atlanta, GA, USA
Jyothi Rengarajan
Greehey Children’s Cancer Research Institute, Department of Molecular Medicine, Long School of Medicine, UT Health, San Antonio, TX, USA
Zhao Lai

Authors

Riti Sharan
View author publications
Search author on:PubMed Google Scholar
Yi Zou
View author publications
Search author on:PubMed Google Scholar
Bindu Singh
View author publications
Search author on:PubMed Google Scholar
Vinay Shivanna
View author publications
Search author on:PubMed Google Scholar
Edward J. Dick Jr.
View author publications
Search author on:PubMed Google Scholar
Shannan Hall-Ursone
View author publications
Search author on:PubMed Google Scholar
Xi Luo
View author publications
Search author on:PubMed Google Scholar
Guanlin Guo
View author publications
Search author on:PubMed Google Scholar
Shabaana A. Khader
View author publications
Search author on:PubMed Google Scholar
Xavier Alvarez
View author publications
Search author on:PubMed Google Scholar
Jyothi Rengarajan
View author publications
Search author on:PubMed Google Scholar
Zhao Lai
View author publications
Search author on:PubMed Google Scholar
Smriti Mehra
View author publications
Search author on:PubMed Google Scholar
Hulin Wu
View author publications
Search author on:PubMed Google Scholar
Deepak Kaushal
View author publications
Search author on:PubMed Google Scholar

Contributions

R.S., J.R., and D.K. designed the study. R.S. and B.S. performed sample collection and processing. V.S. and E.J. performed macaque necropsies and pathology studies. S.H.U. was the attending veterinarian for the study. X.A. performed the PET/CT and related analysis. R.S., Z.L., Y.Z., X.L., G.G., H.W., and D.K. performed the data analysis. Z.L. performed quality control of BAL cells, 10x scRNA-seq and NGS workflow. S.A.K., J.R., and S.M. helped R.S. and D.K. in writing the manuscript.

Corresponding authors

Correspondence to Riti Sharan or Deepak Kaushal.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Description of Additional Supplementary Files (download PDF )

Supplementary Data 1 (download TIF )

Supplementary Data 2 (download TIF )

Supplementary Data 3 (download XLSX )

Supplementary Data 4 (download PDF )

Supplementary Data 5 (download PDF )

Supplementary Data 6 (download PDF )

Supplementary Data 7 (download TIF )

Supplementary Data 8 (download PDF )

Supplementary Data 9 (download PDF )

Supplementary Data 10 (download PDF )

Supplementary Data 11 (download PDF )

Supplementary Data 12 (download PDF )

Supplementary Data 13 (download PDF )

Supplementary Data 14 (download PDF )

Supplementary Data 15 (download PDF )

Supplementary Data 16 (download XLSX )

Supplementary Data 17 (download XLSX )

Supplementary Data 18 (download XLSX )

Supplementary Data 19 (download XLSX )

Supplementary Data 20 (download XLSX )

Supplementary Data 21 (download XLSX )

Supplementary Data 22 (download PDF )

Supplementary Data 23 (download XLSX )

Supplementary Data 24 (download PDF )

Supplementary Data 25 (download XLSX )

Supplementary Data 26 (download XLSX )

Supplementary Data 27 (download XLSX )

Reporting Summary (download PDF )

Transparent Peer Review file (download PDF )

Source data

Source Data (download XLSX )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Sharan, R., Zou, Y., Singh, B. et al. Concurrent TB and HIV therapies control TB reactivation during co-infection but not chronic immune activation. Nat Commun 17, 499 (2026). https://doi.org/10.1038/s41467-025-67188-4

Download citation

Received: 16 August 2024
Accepted: 24 November 2025
Published: 12 December 2025
Version of record: 14 January 2026
DOI: https://doi.org/10.1038/s41467-025-67188-4