Mycobacterium tuberculosis reactivates latent HIV-1 in T cells in vitro

Following proviral integration into the host cell genome and establishment of a latent state, the human immunodeficiency virus type 1 (HIV-1) can reenter a productive life cycle in response to various stimuli. HIV-1 reactivation occurs when transcription factors, such as nuclear factor-κB (NF-κB), nuclear factor of activated T cells (NFAT), and activator protein -1 (AP-1), bind cognate sites within the long terminal repeat (LTR) region of the HIV-1 provirus to promote transcription. Interestingly, pattern recognition receptors (PRRs) that recognize pathogen-associated molecular patterns (PAMPs) can reactivate latent HIV-1 through activation of the transcription factor NF-κB. Some PRRs are expressed on central memory CD4+ T cells (TCM), which in HIV-1 patients constitute the main reservoir of latent HIV-1. Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB), interacts with PRRs through membrane components. However, the ability of Mtb to reactivate latent HIV-1 has not been extensively studied. Here we show that phosphatidylinositol mannoside 6 (PIM6), a component of the Mtb membrane, in addition to whole bacteria in co-culture, can reactivate HIV-1 in a primary TCM cell model of latency. Using a JLAT model of HIV-1 latency, we found this interaction to be mediated through Toll-like receptor-2 (TLR-2). Thus, we describe a mechanism by which Mtb can exacerbate HIV-1 infection. We hypothesize that chronic Mtb infection can drive HIV-1 reactivation. The phenomenon described here could explain, in part, the poor prognosis that characterizes HIV-1/Mtb co-infection.


Introduction
Tuberculosis is the leading cause of death for individuals living with human immunodeficiency virus type-1 (HIV-1) [1][2][3][4]. In 2015 alone, it was estimated that one in every three deaths among HIV-1-infected individuals was due to TB [2]. HIV-1-infected individuals arẽ 20-30 times more likely to contract TB compared to uninfected individuals [5]. HIV-1/Mtb co-infected patients exhibit accelerated HIV-1 disease and shorter overall survival [6]. In addition, the risk for active TB infection increases from around 10% in a lifetime to 10% per year for patients that are co-infected with HIV-1 [3]. This accelerated disease progression indicates an interaction between these two pathogens. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 detected with the test agents (H37Ra or Mycobacterium smegmatis in co-culture, H37Rv lysate, PIM6, and LAM). It has been observed that Jurkat cells have low levels of TLR-2 expression and are insensitive to TLR-2 agonists [25,30]. Therefore, in order to more efficiently test the specific role of TLR-2 in HIV-1 reactivation we stably transduced TLR-2 with a lentiviral vector to express high levels of TLR-2 (JLAT-TLR2) (S1 Fig). PIM6 and H37Rv lysate reactivated latent HIV-1 in the JLAT-TLR2 cells (Fig 1B). Whereas H37Ra, M. smegmatis and LAM did not cause significant reactivation in the JLAT-TLR2 cells. To confirm that the reactivation seen was indeed dependent on TLR-2 signaling, we tested the same conditions in the presence of a TLR-2 neutralizing antibody (PAb-hTLR2). Incubation with the TLR-2 neutralizing antibody significantly attenuated induction of GFP by PIM6 and H37Rv lysate (Fig 1C). A downstream mediator of TLR-2 activation is the transcription factor, NF-κB. Intriguingly, the HIV-  16 hours and GFP expression was measured by flow cytometry. Data represent mean ± SD of four independent experiments run in triplicate. *p<0.05 compared to PBS control. B) JLAT-TLR2 cells were incubated for 16 hours and GFP expression was measured by flow cytometry. Data represent mean ± SD of ten independent experiments run in triplicate. *p<0.05 compared to PBS control. C) JLAT-TLR2 cells were pre-incubated with the TLR-2 neutralizing antibody, PAb-hTLR2, for 30 minutes prior to addition of test conditions. Cells were subsequently incubated for 16 hours and GFP expression was measured by flow cytometry. Data represent mean ± SD. *p<0.05 test condition in the absence of TLR-2 neutralizing antibody compared to test condition in the presence of TLR-2 neutralizing antibody. D) JLAT-TLR2 cells pre-incubated with BAY 11-7082 for 30 minutes prior to addition of conditions. Cells were subsequently incubated for 16 hours and GFP expression was measured by flow cytometry. Data represent mean ± SD. *p<0.05 test condition in the absence of BAY 11-7082 compared to test condition in the presence of BAY 11-7082.
1 LTR contains two NF-κB consensus binding sites within the enhancer region and NF-κB has been shown to be necessary for HIV-1 reactivation [31]. Thus, we tested the NF-κB antagonist, BAY 11-7082, for the ability block viral reactivation mediated by TLR-2. As expected, BAY 11-7082 attenuated viral reactivation mediated by PIM6 and H37Rv lysate ( Fig 1D).
PIM6, H37Rv lysate, and whole bacteria co-culture reactivate HIV-1 in a T CM model of latency We sought to confirm the observations obtained with the JLAT model in a primary cell model of latency that uses cultured T CM cells as targets for replication-competent HIV-1 [29]. For the work presented here, a slight modification was made to the published model; we utilized a replication-competent HIV-1 construct bearing the secNLuc gene within the nef locus of HIV-1 NL4-3 . Cells latently infected with this modified, replication-competent HIV-1 secrete Nano-Luc1 luciferase enzyme into cell culture supernatant when treated with latency reversal agents. This allows for the ability to quantify HIV-1 reactivation from the supernatant of test cultures, providing timely detection of HIV-1 reactivation combined with the sensitivity of luminescence. We also measured intracellular p24 Gag production by flow cytometry. By measuring p24 Gag production, the results documented a statistically significant increase in reactivation in the whole bacteria co-culture conditions (H37Ra and M. smegmatis) and H37Rv lysate, and a trending increase PIM6 (due to inter-individual variation; Fig 2A). However, when luciferase activity was used to measure HIV-1 reactivation in these experiments, statistically significant increases in reactivation was observed with all the tested stimuli except LAM ( Fig 2B). H37Rv lysate was observed to suppress luminescence even though latency reversal was indicated by increased p24 Gag (S3A & S3B Fig). We tested if this observation could be the result of H37Rv lysate interactions with the NanoLuc1 enzyme or inhibition of enzyme secretion. Incubation of H37Rv lysate with supernatant from SupT1 cells infected with the modified HIV-1 showed attenuated luminescence (S3C Fig). Thus, we concluded that the observed suppression of luminescence is due to an undefined interaction with the NanoLuc1 enzyme.

Discussion
A major concern of HIV-1/Mtb co-infection in patients is accelerated HIV-1 infection [6]. While many studies provide evidence for increases in viral spread, mechanisms underlying the impact of Mtb on latent HIV-1 reservoirs have not been investigated. In our study, we found that whole Mycobacteria as well as their membrane component PIM6 and whole H37Rv lysate, which is composed of a mixture of proteins, lipids and carbohydrates from the bacterial cell, Mycobacteria and some of their components were shown early on to induce HIV-1 LTR expression in human monocytes and other cell lines [20,[37][38][39][40][41]. Work in 1994 by Shattock and colleagues concluded that "phagocytosis of M. tuberculosis by monocytes expressing latent . Several groups showed this induction to be dependent on TLR-2 and to be mediated, at least in part, by NF-κB binding to the HIV-1 LTR region [15,25,41,43].
Phagocytosis of mycobacteria by macrophages results in exosome release of lipids and other pro-inflammatory bacterial components that are capable of binding PRRs and activating T cells [52][53][54][55][56]. PIM6 is an example of such a mediator. PIM6 is a powerful TLR-2 agonist shown by Rodriguez and colleagues to up-regulate HIV-1 replication in productively infected and CD3 co-stimulated T cells [20]. Here we report that PIM6-mediated activation of TLR-2 can also lead to HIV-1 latency reversal.
Using the intracellular p24 Gag endpoint to quantify HIV-1 reactivation in the T CM model, we were able to show statistically significant latency reversal in cultures co-incubated with H37Ra, M. smegmatis, and H37Rv lysate. This confirmed the basic hypothesis that the presence of Mtb in proximity to infected T cells is sufficient to reactivate latent HIV-1. While the other stimuli, such as PIM6, showed a trend in the induction of p24 Gag production, the variability observed using latently infected primary cell culture prevented further conclusions regarding its effects. However, when luciferase activity, a more sensitive endpoint than p24 expression, was used it became clear that, in addition to co-culture with whole Mycobacteria and H37Rv lysate, PIM6 significantly reversed HIV-1 latency in this model. These data support the idea that mycobacteria and mycobacterial membrane components can drive HIV-1 activation in latently infected T cells.
Experiments in JLAT cells interrogated the hypothesized major pathway of HIV-1 reactivation by Mtb. JLAT cells, while responding robustly to the positive control PMA activation, did not show statistically significant GFP induction with any of the test conditions. However, when TLR-2 was expressed ectopically, both PIM6 and H37Rv lysate were found to induce significant GFP expression. We found that TLR-2 neutralizing antibody, while not affecting the response to PMA, attenuated the induction responses of PIM6 and H37Rv lysate. Furthermore, use of the NF-κB antagonist, BAY 11-7082, also attenuated the GFP induction by PIM6 and H37Rv lysate. These data suggest that HIV-1 reactivation by PIM6 and H37Rv lysate in The horizontal line within the box represents the median, the boundaries of the box represent the 25 th -and 75 th -percentile, and the whiskers represent the maximum and minimum values. Significance for intracellular p24 Gag was determined using a 2-tailed, paired Student's t-test versus PBS (*p<0.05). (B) Relative luminescence was measured from supernatant of cultured T CM cells following 72-hour incubation with conditions or co-stimulation with αCD3/αCD28. The horizontal line within the box represents the median, the boundaries of the box represent the 25 th -and 75 th -percentile, and the whiskers represent the maximum and minimum values. Significance was determined using a 2-tailed, paired Student's t-test versus PBS (*p<0.05). Significance of individual test conditions are as follows: αCD3/αCD28 (p 0.01), PIM6 (p<0.05), H37Ra (p 0.01), and M. smegmatis (p 0.01). the T CM model occurs through activation of the TLR-2 pathway. Although whole bacteria coculture (H37Ra and M. smegmatis) did not significantly reactivate latent HIV-1 in the JLAT or JLAT-TLR2 studies, they were significant in our primary T CM HIV-1 latency model. This indicates that pathways activated in our T CM cells may vary in expression from those in the JLATs.
Interactions between Mtb and latent HIV-1 infected cells have remained largely undefined, yet HIV-1/Mtb co-infection is one of the world's most lethal conditions. Our study suggests that chronic Mtb infection could drive latency reversal in T CM cells, which may contribute to higher viral load, increased T cell loss and thereby T cell exhaustion. Even periodic infections by other microbial pathogens that activate TLR-2 or converging on pro-inflammatory signaling pathways (e.g. NF-κB) could conceivably contribute to transiently elevated levels of viremia in HIV-1 patients [25]. In poorly managed or non-compliant patients this could have negative health outcomes [57,58]. While our report provides the initial observation, further work is needed to fully describe mechanisms and pathways by which mycobacteria might affect HIV-1 latency reversal.

Materials & methods Reagents
The following reagents were obtained through BEI Resources, NIAID, NIH:

Bacterial culture
H37Ra and M. smegmatis were grown to mid-to-late-logarithmic phase in Difco™ Middlebrook 7H9 Broth (Becton, Dickinson and Company; Cat. #271310) supplemented with ADC enrichment media (Remel™, ThermoFisher Scientific, Cat. #R450592), 0.2% glycerol and 0.05% Tween 80 at 37˚C. For assays, bacterial cultures were adjusted to a total OD 600 of 1.2 reconstituted in 1 mL of phosphate buffered saline (PBS) prior to addition to wells.

Flow cytometry
TLR-2 surface expression was determined using anti-human CD282-APC (and isotype control (Biolegend, San Diego, CA)) Expression was measured in a BD FACSCanto II flow cytometer.

Nanoluciferase assay
Ten microliter aliquots of the supernatant from T CM experiment wells were mixed with 0.1% bovine serum albumin (BSA) in PBS. Nano-Glo1 Luciferase Assay Substrate (Promega) was diluted 1/50 in Nano-Glo1 Luciferase Assay Buffer (Promega). The mixture was incubated at room temperature for 5 min and luminescence was quantified using a Biotek Synergy 2 microplate reader.

Experimental and statistical analysis
Cell-line based assays (JLAT & JLAT-TLR2) were conducted in triplicate and mean values were calculated across four and ten independent experiments, respectively. Significance was determined using an unpaired, two-tailed Student's t-test versus PBS. For inhibitor-based assays, three independent experiments ran in triplicate were performed and mean values were calculated. An unpaired, two-tailed Student's t-test comparing test conditions in the absence and presence of inhibitor was used to determine significance. The T CM model was run in triplicate across eight donors with the exception of PIM6 (n = 7) and H37Rv lysate (n = 5). Analysis of flow cytometry data for the T CM model was conducted for the mean percent p24 positive values for eight donors. Significance was determined using a paired, two-tailed Student's t-test versus PBS. For the NanoLuc1 luciferase assay, the mean relative luminescence units for each donor was calculated and significance was determined using a paired, two-tailed Student's ttest versus PBS. All statistical analyses were performed using GraphPad Prism (GraphPad Software, Inc., La Jolla, CA).

Ethics statement
PBMCs were collected from unidentified, healthy donors following protocols outlined in IRB #67637 (University of Utah Institutional Review Board approved). All donors provided written informed consent prior to entry to the study. paired Student's t-test versus PBS ( Ã p<0.05). (B) Relative luminescence was measured from supernatant of cultured T CM cells following 72-hour incubation with conditions or co-stimulation with αCD3/αCD28. The horizontal line within the box represents the media, the boundaries of the box represent the 25 th -and 75 th -percentile, and the whiskers represent the maximum and minimum values. Significance was determined using a 2-tailed, paired Student's t-test versus PBS ( Ã p<0.05). (C) Supernatant from SupT1 cells infected with HIV Nluc2A (MOI 0.1) or supernatant from uninfected SupT1 cells (Mock) was incubated with H37Rv lysate (100 μg/mL) for 72 hours at 37˚C after which point luminescence was measured. Significance was determined using a one-tailed, unpaired Student's t-test (p<0.05). (TIF) S4 Fig. Plasmid map of pNL43-