Highlights
- •Inhibitors in clinical trials or FDA-approved for cancer can be repurposed for TB.
- •MCL-1 + BCL-2 inhibitors induce apoptosis to control M.tb growth in macrophages.
- •MCL-1 + BCL-2 inhibitors reduce growth of drug resistant and susceptible M.tb.
- •MCL-1 + BCL-2 inhibitors reduce M.tb growth in a human an vitro granuloma model.
- •MCL-1 and BCL-2 are promising targets for host-directed therapy for tuberculosis.
Abstract
Tuberculosis (TB) accounts for 1.6 million deaths annually and over 25% of deaths due to antimicrobial resistance. Mycobacterium tuberculosis (M.tb) drives MCL-1 expression (family member of anti-apoptotic BCL-2 proteins) to limit apoptosis and grow intracellularly in human macrophages. The feasibility of re-purposing specific MCL-1 and BCL-2 inhibitors to limit M.tb growth, using inhibitors that are in clinical trials and FDA-approved for cancer treatment has not be tested previously. We show that specifically inhibiting MCL-1 and BCL-2 induces apoptosis of M.tb-infected macrophages, and markedly reduces M.tb growth in human and murine macrophages, and in a pre-clinical model of human granulomas. MCL-1 and BCL-2 inhibitors limit growth of drug resistant and susceptible M.tb in macrophages and act in additive fashion with the antibiotics isoniazid and rifampicin. This exciting work uncovers targeting the intrinsic apoptosis pathway as a promising approach for TB host-directed therapy. Since safety and activity studies are underway in cancer clinics for MCL-1 and BCL-2 inhibitors, we expect that re-purposing them for TB treatment should translate more readily and rapidly to the clinic. Thus, the work supports further development of this host-directed therapy approach to augment current TB treatment.
1. Introduction
Tuberculosis (TB) is the number one infectious disease killer in human history, accounting for over 1 billion deaths [1], [2]. TB is currently the second leading cause of death from a single infectious agent, second only to COVID-19 in the past two years. Efforts to control TB were stalled, and even reversed, due to the COVID-19 pandemic, with the TB incidence rate increasing by 3.6% between 2020 and 2021, reversing the average 2% decline over the past 20 years. The number of TB deaths also increased from 2019 to 2021, representing the first increase in fatality rate for TB since 2005. Without treatment, the death rate due to TB is 50%. However, TB treatment has an 85% cure rate, and requires a cocktail of rifampicin (RIF), isoniazid (INH), ethambutol, and pyrazinamide for 4–6 months, although there are efforts to shorten this to 1–6 months. Unfortunately, although cases of RIF resistant and multidrug-resistant (MDR; defined as RIF and INH R) TB increased in 2021, the number of people receiving treatment actually decreased. Treatment for drug resistant TB is even more challenging, more costly and associated with more side-effects although new regimens have reduced treatment from 20 + months to 6 months and improved success rates to 60% [1]. There is a clear need for better therapeutics for both drug resistant and susceptible TB, particularly now to help reduce the increase in TB incidence and fatality arising from COVID-19.
There has been a recent push to develop host-directed therapies (HDTs), akin to immunotherapy which has been a breakthrough in cancer treatment. HDTs are expected to work against both drug susceptible and resistant TB, provide a treatment option that the pathogen is unlikely to develop resistance to, shorten treatment duration, boost the immune response, and ameliorate pathology associated with severe disease, and so could become an important tool in the fight against infectious diseases like TB [3]. Perhaps the most studied HDT is vitamin D, which boosts macrophage responses to control M.tb growth. However, clinical trials remain inconclusive as to whether Vitamin D helps control M.tb burden in people [4]. Initial studies aimed at supplementing the cytokine response (with IL-2 or IFNγ) have had limited success, perhaps because cytokines need to be applied early in infection for maximal effect. In contrast, corticosteroids and metformin have been used to reduce pathology during TB infection [3], [5]. However, these are broad approaches and there are still no targeted HDTs in the clinic for TB.
Promising HDTs need to be active in both human macrophages (the primary niche for M.tb) and granulomas, which are the pathologic hallmark of TB. Granulomas are multicellular structures that contain bacteria, however it is hard for antibiotics to penetrate the granulomas and these structures provide a unique environment that causes some antibiotics to lose efficacy. These structures likely contribute to why TB therapy takes so long, and remain a challenge for development of new therapies. One approach for HDTs is to target cell death pathways which can impact the immune response. Historically, cell death has been mainly described as either apoptosis or necrosis. More recently, this has been expanded to include pyroptosis, necroptosis, and ferroptosis [6], [7], [8]. Apoptosis (Programmed Cell Death) has long been suggested as a mechanism to control TB, since avirulent or attenuated mycobacteria induce more apoptosis than virulent M.tb, some M.tb virulence factors have been identified that inhibit apoptosis, and induction of apoptosis leads to reduced M.tb growth in human and murine macrophages [6], [7], [9]. Although this has been contested, recent work with apoptosis deficient mice has confirmed that apoptosis contributes to M.tb control in vivo [10], confirming that M.tb dampens apoptosis as a survival strategy. Apoptosis can be induced through two main pathways: the intrinsic or mitochondrial pathway and the extrinsic or death receptor-mediated pathway. The extrinsic pathway involves death receptors like Fas and tumor necrosis factor receptor (TNFR) and their ligands (Fas ligand, TNFα) and caspase-8 activation. The intrinsic or mitochondrial pathway is regulated by the BCL-2 protein family which consists of effectors BAK and BAX, anti-apoptotic BCL-2 proteins (BCL-2, MCL-1, BCL-XL, A1, BCL-W), and pro-apoptotic BH3-only proteins (BIM, PUMA, NOXA, BID, BAD, BMF, BIK). BAK and BAX oligomerize and form pores in the mitochondrial membrane, leading to cytochrome C release; and BAK and BAX activity is regulated by the anti-apoptotic and pro-apoptotic proteins [11]. The intrinsic and extrinsic pathways merge at caspase-3 activation, leading to DNA damage and apoptosis. A hallmark of apoptosis is that cells die with intact membranes, which limits inflammatory responses and efferocytosis of these apoptotic cells contributes to antigen presentation and T cell activation during M.tb infection. In contrast, necrosis is associated with increased M.tb growth and necrotic cells have membrane damage which releases pro-inflammatory mediators into the environment [6], [7], [12]. Thus, targeting apoptosis should aid in TB control on multiple levels: limiting M.tb growth in host macrophages and increasing antigen presentation, thereby boosting both the immune and adaptive response to TB while limiting damaging inflammation associated with disease.
Although inducing apoptosis has been a focus for boosting cancer treatment and polymorphisms in MCL-1 [13] and BCL-2 [14] are associated with susceptibility to TB disease, we were the first to interrogate inducing apoptosis through targeted inhibition of the anti-apoptotic BCL-2 proteins (specifically, MCL-1) for TB treatment [15]. Our previous work showed that targeted MCL-1 inhibition reduced M.tb growth in human macrophages, but that inhibition of multiple anti-apoptotic BCL-2 proteins (with pan inhibitors that inhibited anti-apoptotic MCL-1, BCL-2, and BCL-XL) was required to reduce M.tb growth in a more complex model of human granuloma-like structures [15]. We did not think the pan inhibitors were viable HDTs for TB however, since cancer clinical trials with the pan inhibitor ABT-263/navitoclax have been hampered due to toxicity concerns related to reduced platelet counts through inhibition of BCL-XL [11]. Instead, we hypothesized that we could specifically target MCL-1 and BCL-2 (and not BCL-XL) to reduce M.tb growth in macrophages and the more complex granulomas. Since we are not inhibiting BCL-XL we expect there to be minimal to no toxicity concerns associated with platelet loss. Intriguingly, there is only one manuscript to date that interrogated chemically inducing extrinsic apoptosis in mice and found this significantly limited M.tb growth in vivo [10]. No one has investigated inducing intrinsic apoptosis as a way to limit M.tb infection.
Here we interrogated if specific inhibition of MCL-1 and BCL-2, using combinations of these inhibitors that are efficacious in mouse cancer models and have advanced to clinical trials or are FDA-approved for cancer therapy [11], [16], [17], [18], [19] would: 1) limit M.tb burden in human macrophages more than inhibition of just one target, 2) limit growth of both drug susceptible and resistant M.tb stains, 3) further reduce M.tb burden when combined with front line antibiotics INH or RIF, and 4) limit M.tb growth in a pre-clinical human in vitro granuloma model, which we previously developed [20]. In this model, the in vitro granuloma-like structures contain the predominant cells within human in vivo granulomas. The structures are up to 15 cell layers thick, and this model is used to study infectious and non-infectious granulomatous diseases, including M.tb and sarcoidosis, respectively [20], [21], [22], [23]. This model recapitulates several signatures observed in sarcoidosis, like up-regulation of Th1 immune response and alternatively activated M2-like macrophages [23], [24]. Thus, this multicellular granuloma model provides a more readily tractable, accessible, and affordable throughput bridge model for assessing HDT and antibiotic efficacy in complex granulomas than animals.
2. Material and methods
2.1. Study approval
Peripheral blood mononuclear cells (PBMCs) were isolated from human peripheral blood collected from healthy donors, following Texas Biomed approved Institutional Review Board (IRB) protocols. All donors for these studies provided informed, written consent. Texas Biomed Institutional laboratory Animal Care and Use Committee (IACUC) approved all protocols involving animal samples.
2.2. Isolation and culture of human monocyte-derived macrophages (MDMs)
MDMs were prepared as described elsewhere [25], [26]. Briefly, heparinized blood was layered on a Ficoll-Paque cushion (GE Healthcare, Uppsala, Sweden) to allow for collection of PBMCs. PBMCs were cultured in RPMI 1640 (Life Technologies, Carlsbad, CA) with 20% autologous serum in Teflon wells (Savillex, Eden Prairie, MN) for 5 days at 37 °C/5% CO2. MDMs were harvested and adhered to tissue culture dishes for 2–3 h in RPMI with 10% autologous serum, lymphocytes were washed away, and MDMs were incubated overnight in RPMI with 10% autologous serum. Such MDM monolayers are 99% pure and viable.
2.3. Isolation and culture of murine bone marrow-derived macrophages (BMDMs)
8–9-week-old female BALB/c and C57BL/6 mice (Jackson Laboratory, Bar Harbor, ME), were euthanized by CO2 asphyxiation. For BAX/BAK KO, Rosa-CreERT2 Baxf/f Bak-/- mice were treated with tamoxifen for five days to induce deletion of BAX, then mice were euthanized and tibias and femurs collected [27]. BMDMs were obtained from tibias and femurs as previously described [28]. BMDMs were frozen in DMEM supplemented with 30% L-cell conditioned media, 20% heat inactivated-FBS (HI-FBS), 10 units penicillin/streptomycin (Pen/Strep), 50.3 μM β-mercaptoethanol and 10% dimethylsulfoxide. L-cell conditioned media was obtained by collecting media from confluent NCTC clone 929 (ATCC # CCL-1) cells. The day before the experiment, BMDMs were thawed and plated in tissue culture dishes in DMEM supplemented with 10 units Pen/Strep and 10% HI-FBS. The day of the experiment, cells were washed twice with DMEM to remove the antibiotics from the culture medium.
2.4. Bacterial strains
Lyophilized M.tb H37Rv (27294) was obtained from the American Tissue Culture Collection (ATCC, Manassas, VA). M.tb H37Rv lux was created and used as described [29]. This bacterial strain contains the entire bacterial Lux operon cloned in a mycobacterial integrative expression vector. The previously validated M.tb H37Rv InhA T(−8) INH R strain was obtained from Dr. Sanjay Jain, Johns Hopkins University School of Medicine, Baltimore, MD [30]. M.tb strain HN563 (NR-18986), a clinical isolate with resistance to RIF and INH, was obtained through BEI resources, NIAID, NIH. The clinical strain of M.tb resistant to INH and EMB was obtained from the OSU Clinical Microbiology Laboratory [29]. Single cell suspensions of bacteria were prepared as previously described [31], [32]. The bacteria concentration and degree of clumping (<10%) were determined with a Petroff-Hausser Chamber. This method results in ≥ 90% viable bacteria, as determined by colony forming unit (CFU) assay.
2.5. M.tb infection of macrophages
Single cell suspensions of M.tb in RHH [for MDM infections: 10 mM HEPES (Life Technologies) and 0.1% human serum albumin (CSL Behring, King of Prussia, PA) in RPMI] or DHH [for BMDM infections: 10 mM HEPES and 0.1% human serum albumin in DMEM] were added to the macrophages at MOI 1 and cells were incubated for 2 h at 37 °C, with the first 30 min on a platform shaker. Macrophages were then washed and incubated in RPMI with 2% autologous serum (for MDMs) or DMEM with 2% HI-FBS (for BMDMs) for the indicated times. Where indicated, the MCL-1 and BCL-2 inhibitors, antibiotics, or solvent control (dimethyl sulfoxide, DMSO), were added after this wash step. The MCL-1 inhibitors S63845, MIK665, and AZD5991, and BCL-2 inhibitor ABT-199 were purchased from Selleckchem (Houston, TX), and the MCL-1 inhibitor AMG 176 was from MedChemExpress (Monmouth Junction, NJ). The antibiotics RIF and INH were purchased from Sigma (St. Louis, MO). All inhibitors and antibiotics were maintained throughout the course of infection. Cells were observed with an EVOS XL Core Imaging System to ensure monolayer integrity was maintained throughout the course of all experiments (Figs S1 and S3).
2.6. Culture, infection, staining and imaging of in vitro human granulomas
In vitro TB granulomas were generated as described elsewhere [15], [20], [21]. Briefly, human peripheral blood was collected from healthy Mantoux tuberculin skin test (TST) and/or IFNγ release assay (IGRA)-positive individuals. PBMCs were isolated as above, and were immediately infected with single cell suspensions of M.tb at MOI 1 in RPMI with 10% autologous serum, then incubated at 37 °C/5% CO2. Serum was replenished after 4 days. With this method, cells typically start to aggregate around day 4 and the granulomas are stable for up to 12 days. For CD56 staining, cells on glass coverslips were fixed with 4% paraformaldehyde (PFA) for 20 min and then washed twice with PBS and incubated in blocking buffer (5% BSA/10% HI-FBS/PBS) overnight at 4 °C. Cells were then incubated with 488-tagged antibodies against CD56 (clone HCD56, cat # 318311, lot B249761, Biolegend), for 1 h in blocking buffer, washed, stained with DAPI and mounted with ProLong Gold antifade reagent. Cells were imaged with a Zeiss LSM 800 confocal microscope. Live cell imaging was performed using the Cytation 5 microscope paired with BioSpa (Agilent, Santa Clara, CA) to maintain cells at 37 °C/5% CO2 for the duration of the movie. Images were acquired every 6 h (0–138 h time period) with a 4x objective. See [20] for details regarding CD11b, CD3 and M.tb staining in granulomas. For trypan blue staining, equal volume of 0.4% trypan blue was added to structures at days 4 or 7 post infection. Cells were immediately centrifuged at 100 g for 5 min (to ensure granuloma structures remained attached to the tissue culture wells), washed with PBS and then immediately imaged with a 20x objective on an EVOS XL Core Imaging System. As a positive control, structures were treated with 0.1% Triton-X 100 10 min prior to trypan blue staining, this resulted in disruption of granuloma structures and > 99% cells becoming trypan blue-positive. For CFU experiments antibiotics, MCL-1 and BCL-2 inhibitors, or solvent controls were added as described in the figure legends. The antibiotic moxifloxacin (MX) was purchased from Selleckchem, and linezolid (LNZ) and cycloserine (CS) were purchased from Sigma.
2.7. M.tb growth assays
2.7.1. Intracellular growth
Intracellular growth was assayed with two approaches. For CFU assays, infected macrophages and granulomas were lysed as described previously [20], [33]. Lysates were diluted, and plated on 7H11 agar (Remel, San Diego, CA). The number of CFUs was enumerated after growth for at least 3 weeks at 37 °C. For luciferase growth assays, macrophages were infected with M.tb-lux, and bacterial bioluminescence was measured in relative luminescence units (RLUs) every 24 h for up to 7 days with a GloMax Microplate Reader (Promega, Madison, WI) [29].
2.7.2. Broth cultures
M.tb was cultured in 7H9 broth (BD, Franklin Lakes, NJ) with S63845 + ABT-199 for 4 days, followed by dilutions in 7H9 and plating on 7H11 agar. The number of CFUs was enumerated after growth for at least 3 weeks at 37 °C.
2.8. Cell death assays
Macrophages in 96 well plates were infected with M.tb at MOI 1 and caspase activity and cell death was assayed in triplicate wells with Caspase-3/7 or CellTiter Glo Assays (Promega), respectively, following the manufacturer’s instructions.
2.9. Western blotting
Cells were washed with PBS, then lysed with TN1 lysis buffer (125 mM NaCl, 50 mM Tris, 10 mM EDTA, 1% Triton X-100, 10 mM Na4PO7, 10 mM NaF with 10 mM Na3VO4, 10 μg/ml aprotinin, and 10 μg/ml leupeptin) at 4 °C. Lysates were centrifuged (10,000 g, 4 °C, 10 min) to remove cell debris, then a Pierce BCA assay (Thermo Scientific, Waltham, MA) was performed to determine protein concentration. Equivalent amounts of denatured and reduced protein were separated by SDS-PAGE and analyzed by Western blot using antibodies against caspse-3 (cat# 9662, lot 19), poly ADP-ribose polymerase (PARP; clone 46D11, cat# 9532, lot 10), and β-actin (clone 13E5, cat# 5125, lot 7), obtained from Cell Signaling (Danvers, MA). The membranes were developed using clarity ECL reagent on a UVP ChemStudio 815 system (Analytik Jena US LLC, Upland CA).
2.10. Statistics
A minimum of at least three independent experiments was performed, with cells from at least three different human donors, unless indicated otherwise. Although the trend was the same for each donor, the magnitude of change differed among donors. Consequently, results from each experiment were normalized to an internal control and an unpaired one-tailed Student’s t-test (if comparing two groups) or ANOVA (when comparing more than 2 groups) were performed on the normalized data, with P < 0.05 considered significant.
3. Results
3.1. Combination inhibitors against MCL-1 and BCL-2 reduce M.tb growth in human and murine macrophages
We have previously shown that MCL-1 inhibitors reduce M.tb growth in human macrophages, but that inhibition of multiple proteins in the MCL-1/BCL-2 family was more effective and required to reduce M.tb growth in more complex human in vitro granuloma structures [15]. Since this work, newer more potent MCL-1 inhibitors have been reported, including ones with efficacy in mouse cancer models. We were interested in determining if a new MCL-1 inhibitor, in combination with a BCL-2 inhibitor, would limit M.tb growth to a similar extent as the pan BCL-2 family inhibitors. To assess this, we infected human monocyte-derived macrophages (MDMs) with M.tb, then treated with MCL-1 or BCL-2 inhibitors alone and together. We tested the MCL-1 inhibitor S63845 and the BCL-2 inhibitor ABT-199 at concentrations similar to those used in the literature that specifically induce apoptosis in a BAX/BAK and/or caspase-dependent manner [19], [34], [35], [36], [37]. Of note, 10 μM S63845 alone reduced M.tb burden in macrophages by 27.25 ± 7.47% (Fig. 1A), indicating this is more potent than the previous MCL-1 inhibitors, which required a higher concentration to reduce M.tb growth [15]. 10 μM ABT-199 alone reduced M.tb burden by 33.49 ± 3.81%, to a similar extent as the MCL-1 inhibitor S63845. In contrast to single inhibitors, combination therapy was significantly more potent at reducing M.tb growth in macrophages, and this occurred in a concentration dependent manner (Fig. 1A). At 10 μM S63845 + 1 μM ABT-199, intracellular growth was reduced by 62.77 ± 7.74%, N = 10, and at 10 μM S63845 + 10 μM ABT-199, intracellular growth was reduced by 85.76 ± 3.27%, N = 9. This marked level of growth reduction is particularly notable with use of primary human macrophages.
