All Mul strains were cultured at 32°C with shaking in Middlebrook 7H9 supplemented with 10% OADC (oleic acid, albumin, dextrose, catalase), 0.5% glycerol, and 0.02% tyloxapol or in Middlebrook 7H10 supplemented with 10% OADC and 0.5% glycerol with or without antibiotics as per requirements (hygromycin, 50 μg/ml or kanamycin, 25 μg/ml). Escherichia coli strains used for cloning were grown on LB agar or broth with or without antibiotics as per requirements (kanamycin 50 μg/ml or hygromycin, 100 μg/ml).

Antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA) unless indicated otherwise and cataloged in Table S1 . All reagents and media purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated.

Human THP-1 monocytes were maintained in supplemented Roswell Park Memorial Institute (RPMI)-1640 medium [bicarbonate buffered RPMI-1640 containing glutamine supplemented with 1% non-essential amino acids, 10% heat-inactivated fetal bovine serum (Corning, NY, USA), 1% HEPES, 1% sodium pyruvate, and 50 μM β-mercaptoethanol (RPMIc)] at 37°C with 5% CO₂. THP-1 cells were seeded in 12-well plates at 5 × 10⁵ cells/well 72 h before infection or treatment. Monocytes were derived to macrophages by adding 10 ng/ml phorbol myristate acetate (PMA) for 48 h. Adhered derived macrophages were washed once in RPMIc and rested overnight in RPMIc before infection.

Bone marrow-derived primary cells were derived according to previously published methods (32). Briefly, marrow was flushed from tibias and femurs of 6- to 8-week-old C57BL/6J mice aseptically and cultured in non-tissue-culture-treated serological plates in RPMIc supplemented with 100 U/ml penicillin and 100 μg/ml streptomycin (RPMIcAbx). For macrophage differentiation, cells were seeded in 100-mm plates at 2 × 10⁵ cells/ml and differentiated by adding 15% L929 fibroblast-conditioned media for 6 days, followed by feeding with fresh media every 2 days. On day 6, adherent cells were washed with ice-cold phosphate-buffered saline (PBS) and detached by incubation on ice for 20 to 30 min in ice-cold PBS. BMDMs were seeded in 12-well plates at 5 × 10⁵ cells/well and allowed to adhere overnight in DMEMc.

Mycobacteria were grown to an optical density (OD₆₀₀) of 0.6 to 0.8. After washing bacteria two times in PBS, bacteria were resuspended in RPMIc or DMEMc and de-clumped by centrifugation at 800 ×g for 8 min. De-clumped bacteria were used to infect macrophages at a multiplicity of infection (MOI) 10 unless otherwise stated. Infection was carried out for 4 h at 32°C, after which macrophages were washed 3 times with PBS and treated with 50 μg/ml gentamicin in complete media for 1 h to kill extracellular bacteria. Macrophages were rewashed 3 times with PBS, and assays were conducted in RPMIc or DMEMc for indicated times at 32°C with 5% CO₂.

At indicated time points, culture supernatants were collected for LDH assay and bacterial enumeration. Cells were then harvested in radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 1% NP-40 or Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris–HCl, pH 8.0, 20 mM Tris–HCl, pH 7.5) for plating for intracellular survival or immunoblot analysis. For CFU enumeration, lysates were serially diluted and plated on 7H10 with appropriate antibiotics. Alternatively, macrophages were detached by incubation in ice-cold PBS containing 5 mM EDTA for 5 min on ice for staining.

At 72 h postinfection, cell-culture supernatants were collected for LDH analysis. The BioLegend LDH-Cytox Assay Kit was used, per the manufacturer’s instructions. Briefly, 50 µl of culture supernatants was added to 50 µl PBS in a 96-well plate. 100 µl of the working solution was added to wells and incubated at room temperature for 30 min. A 50-µl stop solution was added, and each well and absorbance were read to 490 nm. Cytotoxicity was calculated by subtracting the absorbance of untreated cells normalized to the absorbance of 100% lysed cells.

Apoptosis was determined from collected cells using the GFP-certified Apoptosis/Necrosis detection kit (Enzo Life Sciences, Farmingdale, NY, USA) or CellEvent Caspase3/7 Green Detection Reagent (Invitrogen, Carlsbad, CA, USA), per manufacturers’ directions. For Apoptosis/Necrosis detection, cells were washed once in ice-cold PBS, resuspended in dual detection reagent (Annexin-V and 7-AAD), and incubated for 15 min. Cells were washed once more and resuspended in 2% PFA. For Caspase 3/7 activity, collected cells were washed once in ice-cold PBS. Macrophages were resuspended in detection reagent and incubated on ice for 30 min. Cells were washed once more in PBS and resuspended in 2% PFA. All samples were acquired on an Accuri C6 Plus Flow Cytometer (BD, Franklin Lakes, NJ, USA), and data were analyzed using FlowJo (Ashland, OR, USA).

Macrophage cellular protein was prepared in 1× RIPA buffer, and bacterial cells were lysed by bead-beating with 1-mM silica zirconium beads in 0.05 M potassium phosphate and 0.02% β-mercaptoethanol. The protein concentration was determined by bicinchoninic acid (BCA) assay (Pierce). Aliquots of lysates containing 1 to 10 μg of protein were resolved on 12% SDS-PAGE gels at 180 V for 40 min. Proteins were transferred to 0.2 μm polyvinylidene difluoride (PVDF) using a Bio-Rad Trans-Blot Turbo at 2.5 A and 25 V for 5 to 10 min depending on molecular weight. PVDF membranes were blocked in 5% non-fat dry milk in 1× Tris-buffered saline (TBS) plus 0.1% Tween 20 (TBST) or OneBlock Western-CL Blocking Buffer (Genesee Scientific, San Diego, CA, USA) for LC3B blots at room temperature for at least 1 h. Primary antibodies at 1:5,000 dilution were incubated overnight at 4°C in TBST. The anti-Rabbit IgG-horseradish peroxidase (HRP) antibody (1:10,000) was added to membranes for 45 min in TBST. Proteins of interest were revealed using Clarity ECL (Bio-Rad) according to the manufacturer’s instructions. Blots were imaged using GE Amersham Imager 600, and densitometric analysis was conducted by ImageJ software (https://imagej.nih.gov/ij/links.html). The protein of interest was normalized to β-actin or GAPDH loading control to calculate autophagy levels (33).

The pTetR-mup045 plasmid was constructed by cloning the PCR product (1,021 bp) of the mup045 gene with a C-terminal HA tag into the backbone of the Tet-based expression vector pTACT13 (Addgene # 24784, Watertown, MA, USA), which was a gift from Tanya Parish. The mup045F (CCATGGGTGATTTGGAATGACATCTACATAAGTGG) and mup045HA (TTTAAACTAGGCGTAGTCCGGCACGTCGTACGGGTACGAAGTGGAGTGTCCGGGC) primers were used to amplify mup045. Both insert and vector were digested with NcoI and DraI. Mup045 expression was induced by adding 1 mg/ml anhydrotetracycline (Tet-ON).

Mycolactone was extracted from bacterial cultures as previously described (34). Briefly, a Folch extraction was done with 0.2 volumes of bacterial culture added to 0.8 volumes of 2:1 chloroform:methanol. This extraction was incubated for 2 h at room temperature with rocking. The solvent phase of the extraction was collected and dried, which was resuspended in ice-cold acetone and incubated overnight at -20°C. The acetone-insoluble precipitate was pelleted at 2,000 ×g for 5 min, and acetone-soluble lipids were collected. These acetone soluble lipids are enriched for mycolactone. This fraction was again dried and resuspended in ethyl acetate for cytopathic assay.

L929 fibroblasts were seeded in 96-well plates at 2.5 × 10⁵ cells/ml in DMEMc and allowed to adhere overnight. Cells were treated with serial dilutions of synthetic mycolactone or extracted bacterial mycolactone for 72 h at 37°C. Synthetic mycolactone A/B were supplied by Dr. Yoshito Kishi (Harvard University), as ethanol-diluted solutions (1 mg/ml). The purity of synthetic mycolactone A/B was confirmed by 1H- and 13C-nuclear magnetic resonance and also by high-performance liquid chromatography. MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay was carried out as per the manufacturer’s instructions (Abcam, Cambridge, MA, USA). Briefly, culture supernatant was removed from cells, replaced with serum-free media containing MTT reagent, and incubated at 37°C for 3 h. MTT solvent was then added and incubated at room temperature for 15 min. Absorbance was read at 595 nm using a microplate reader (Bio-Rad, Hercules, CA, USA).

All animal studies were approved by the institutional animal care and use committee of the University of Texas Medical Branch. Female C57BL/6J mice were obtained from The Jackson Laboratory between 6 and 8 weeks of age. Mice were infected with 1 × 10⁴ CFU subcutaneously in the left hind footpad. Footpad heights were measured fortnightly for 12 weeks. At weeks 4 and 10 postinfection, mice were euthanatized, and infected footpads were collected. Footpads were disinfected by soaking for 5 min in 70% ethanol followed by washing three times in PBS. Footpads were divided medially. For histology, half of the footpad was added to 10% neutral buffered formalin. The remaining footpad was diced into 2 ml PBS and homogenized using a probe homogenizer. One volume of disinfectant solution (4% NaOH, 2.9% sodium citrate, and 1% N-acetyl-L-cysteine) was added to the homogenate and incubated at room temperature for 20 min. A further 5 ml PBS was added to homogenates, and bacteria pelleted at 2,500 × g for 10 min. Pellets were resuspended in 1 ml PBS and serially diluted for CFU enumeration by plating.

GraphPad Prism 8 was used for all analyses. ANOVA was used to determine significance with Dunnett correction for multiple comparisons unless otherwise stated. A p-value of < 0.05 was considered to be significant.

An intracellular growth phase for Mul has been previously eluded to; however, mycolactone has also previously been described as inhibiting phagocytosis in macrophages (7, 30, 31). To study the exact role of mycolactone during an intracellular growth phase, we used the previously described mycolactone-deficient strain containing a transposon insertion in mup045 (Mu::Tn118) and its parental strain (Mu1615) (7). Primary bone marrow-derived macrophages (BMDMs) from C57Bl/6J mice were infected with Mu1615 or Mu::Tn118 at a multiplicity of infection (MOI) of 10, 5, 2, and 1. At 72 h postinfection, cytotoxicity (LDH release) was measured in the macrophage culture supernatant ( Figure 1A ). Macrophages infected with Mul competent in mycolactone production (Mu1615) released increased LDH compared to macrophages infected with mycolactone-deficient Mu::Tn118 in a dose-dependent manner. While similar intracellular bacterial numbers were observed in Mu1615 and Mu::Tn118 infection 4 h postinfection (data not shown), significantly lower numbers of intracellular Mu1615 were observed at 72 h postinfection compared to Mu::Tn118 ( Figure 1B ). However, the number of Mu1615 was higher than Mu::Tn118 in the culture supernatant of these macrophage cultures. Mu1615 can egress from BMDMs more efficiently than Mu::Tn118 deficient in mycolactone synthesis. Thus, these data demonstrate that intracellular growth of Mul-producing mycolactone occurs by multiplication in individual macrophages followed by their lysis, egress of replicated bacilli.

Conversely, Mu::Tn118 induces significantly more apoptosis in BMDMs than the mycolactone-competent Mu1615 strain ( Figure 1C ). These results likely indicate that mycolactone plays a role in necrotic cell death, promoting bacterial egress from macrophages. Bright-field microscopy also shows this significantly increased lytic cell death in Mu1615-infected macrophages ( Figure 1D ). BMDMs infected with Mu1615 are more rounded and are less confluent than BMDMs infected with Mu::Tn118.

Synthetic and purified mycolactone have been previously shown to inhibit mTOR and Akt, resulting in BH3-only BCL-2-interacting mediator of cell death (Bim) protein activation and apoptosis (24). To determine if this translates to live Mul infection, we examined LC3B-II accumulation as a measure of autophagy and phospho-S6 (Ser235/236) as a measure of mTOR activation. We observed low levels of autophagy ( Figure 1E ) in Mul-infected BMDMs, like observations from other virulent mycobacteria-infected macrophages (26). As with these other well-studied mycobacterial infections, we observed significant levels of mTOR activation ( Figure 1F ). Autophagy and mTOR activation were increased in a dose-dependent manner. There was no difference in autophagy or mTOR activation between the mycolactone-competent and -deficient strains, indicating that bacterial mycolactone cannot overcome the mTOR activation by Mul.

Many studies have demonstrated a role for mycolactone in apoptosis induction during BU infection (8, 22–24, 35). While this seems likely, we hypothesized that mycolactone also induced necrosis to facilitate Mul’s escape from phagocytic cells during the early stage of infection. To facilitate the study of the bacterial egress, we determined if THP-1 human monocyte-derived macrophages behave similarly to primary murine macrophages. Monocyte-derived THP-1 macrophages were infected with Mul at an MOI of 10, 5, and 2. At 72 h postinfection, the cytotoxicity from these infected macrophages was determined ( Figure 2A ). A similar dose response in cytotoxicity to the infected BMDMs was observed, and we subsequently chose to do all the following experiments in this human monocytic cell line. As with the BMDMs, we once again demonstrated reduced cytotoxicity in Mu::Tn118-infected macrophages compared to Mu1615-infected cells. To differentiate between apoptosis and necrosis, we stained infected macrophages with Annexin-V and 7-AAD. Cells positive for only Annexin-V were determined to be apoptotic ( Figure 2B ), while cells positive for both Annexin-V and 7-AAD were determined to be necrotic, which thus have a permeable membrane allowing for bacterial egress from the macrophage. Synthetic and purified mycolactones have been previously shown to inhibit mTOR and Akt, resulting in BH3-only BCL-2-interacting mediator of cell death (Bim) protein activation and apoptosis (24). In line with results from BMDMs, these data support the proposed hypothesis that mycolactone-induced lytic cell death led to bacterial escape, which was further evidenced by the increased extracellular bacteria in the macrophage culture supernatant 72 h postinfection ( Figure 2C ).

We further demonstrated this decreased intracellular CFU by microscopically examining THP-1 monocyte-derived macrophages infected with Mu1615::GFP or Mu118::GFP at an MOI of 10, 48, and 72 h postinfection ( Figure 2D and Supplementary Figure 1 ). These cells were stained with Annexin-V to mark apoptosis and necrosis. Annexin-V staining was observed in both Mu1615- and Mu118-infected cells. However, a reduced number of intracellular bacteria were observed at 72 h postinfection compared to 48 h postinfection in Mu1615::GFP-infected cells than Mu118::GFP-infected cells, demonstrating Mu1615::GFP escape due to the increased necrosis during mycolactone competent infections ( Supplementary Figure 1 ).

During infection with mycolactone-competent Mul, we observed a significant activation of mTOR. However, it is documented that synthetic mycolactone inhibits signaling of mTORC1/2, resulting in dephosphorylation and inactivation of the Akt kinase and induction of Bim-dependent apoptosis via the mTORC2–Akt–FoxO3 axis (24). The Akt-targeted transcription factor, FoxO3, is the central transcriptional regulator of Bim-induced apoptosis in mycolactone-treated cells. To confirm that synthetic mycolactone can modulate this pathway during Mul infection similar to bacterial mycolactone modulation, we treated infected THP-1 monocyte-derived macrophages with increasing concentrations of synthetic mycolactone. Synthetic mycolactone induced high levels of cytotoxicity at all concentrations in both uninfected and infected macrophages, as expected ( Figure 3A ). These data indicated that synthetic mycolactone induces lytic cell death, as observed in mycolactone-competent Mul infections.

As all tested concentrations of mycolactone induced cytotoxicity and many other studies have used 80 to 200 nM of synthetic mycolactone, we chose to conduct the remaining studies with 100 nM of mycolactone. Unlike Mu1615 infection, the addition of synthetic mycolactone induced apoptosis and necrosis with mycolactone-competent and -deficient Mul infection ( Figure 3B ), suggesting that synthetic mycolactone may modulate apoptosis induction differently to mycolactone produced during Mul infection. As expected, we observed that intracellular bacteria are decreased in macrophages 72 h postinfection during mycolactone treatment ( Figure 3C ). These data, taken together, highlight a critical role of mycolactone in modulating host death pathways.

To confirm that necrosis induced by mycolactone was responsible for bacterial egress from macrophages, we treated infected THP-1-derived macrophages with the pan-caspase inhibitor Z-VAD-FMK. Pan caspase inhibition demonstrated reduced levels of apoptosis and necrosis during Mul infection or synthetic mycolactone treatment ( Figure 3D ). In macrophages where caspase-dependent necrosis was inhibited, we also observed increased intracellular bacteria and decreased extracellular bacteria ( Figure 3E ).

Indicative of the shortcomings of the synthetic mycolactone model to determine its role during the early stages of infection, we found significant differences in autophagy induction and mTOR modulation by synthetic mycolactone and bacterial mycolactone. The addition of synthetic mycolactone to Mul-infected macrophages resulted in increased autophagy as measured by LC3B-II accumulation ( Figures 3F, G ). As predicted, this is due to synthetic mycolactone’s ability to inhibit mTOR, as measured by p-S6 and p-Akt (Ser473) accumulation, during infection ( Figures 3F, H ). mTOR activation and autophagy activation were simultaneously increased in a dose-dependent manner and responded to the presence of synthetic mycolactone ( Supplementary Figure 2 ). These data indicated that while synthetic mycolactone is a valuable resource to help understand its ability to modulate host systems, it may overly simplify the complex infection dynamic to represent reality closely. To study the mycolactone–Mul–host infection dynamics more accurately, we have developed an inducible expression system for mycolactone.

To comprehensively study the role of mycolactone during Mul infection, we developed an inducible mup045 expression system to restore mycolactone synthesis in the Mu::Tn118 mutant strain with a transposon insertion in mup045. Utilizing a previously described tetracycline-inducible promoter (36), the mup045 gene from Mu1615 was cloned into the tetracycline-inducible vector with a 3′ hemagglutinin (HA) tag (pTetR-mup045, Figure 4A ) and transformed into Mu::Tn118, resulting in Mu::Tn118C′. The mup045-HA expression was confirmed by Western blot ( Figure 4B ), induced by anhydrous tetracycline (aTCN). The presence of the tetracycline-inducible construct (Mu::Tn118C′) did not affect the growth of Mu::Tn118 in vitro ( Figure 4C ). However, the induced expression of mup045-HA by aTCN addition (Mu::Tn118C′+) or the constitutive expression of mup045 in Mu::Tn118 (Mu::Tn118Con) decreased in vitro growth. A significant decrease in bacterial CFU from those liquid cultures at 18 and 40 days was shown ( Figure 4D ). A constitutive expression of mup045 under the HSP60 promotor is highly toxic in vitro (data not shown), underscoring an advantage of the inducible expression system described here. To confirm that mup045-HA expression resulted in mycolactone synthesis, we extracted mycolactone by Folch’s extraction method at days 3, 18, and 40 of in vitro growth. The significantly increased cytotoxicity of extractions from Mu1615 and Mu::Tn118C’ growth with aTCN to L929 fibroblasts was observed by MTT assay compared to Mu::Tn118 or the uninduced Mu::Tn118C′ ( Figure 4E ). Apoptosis and necrosis induction by the inducible mycolactone producers were also determined on L929 fibroblasts ( Supplementary Figure 3 ). The synthetic mycolactone and mycolactone produced during infection cause substantial necrosis and limited apoptosis in fibroblast. These results suggest that mycolactone is an essential substance for the induction of necrosis, and there is a difference in the activity of mycolactone on different cell types.

Because we developed this system to enable the study of the role of mycolactone during early Mul infection, we first confirmed that this inducible system behaves as Mu1615 does. Mu1615, Mu::Tn118, and Mu::Tn118C′ were used to infect the THP-1 cells at MOI 10 for 3 days. The addition of 1 µg/ml aTCN during the Mu::Tn118C′ infection successfully induced the expression of mup045-HA in this system ( Figure 5A ). Mu1615 and Mu::Tn118C′ (+ aTCN) induced high levels of necrosis, resulting in increased bacterial egress from THP-1 cells 72 h postinfection compared to Mu::Tn118 infection ( Figures 5B, D ). Conversely and as observed in Figure 2 , Mu1615 and Mu::Tn118C′ (+ aTCN) induced low levels of apoptosis and reduced intracellular bacteria in THP-1 cells 72 h postinfection compared to Mu::Tn118 infection. As expected, the induction of mup045-HA expression resulted in increased cytotoxicity in THP-1 macrophages 72 h postinfection than Mu::Tn118 infection, behaving like Mu1615 infection ( Figure 5D ). We also observed that necrosis induction by mycolactone-producing strains is significantly higher than non-producing Mul strains at 48, 72, and 96 h postinfection of THP-1 cells ( Supplementary Figure 4 ). The enhanced necrosis correlates with a higher extracellular bacterial number for Mu1615 and Mu::Tn118C’ + strains ( Supplementary Figure 5 ). Taken together, these data demonstrate that our tetracycline-inducible mycolactone synthesis system successfully complements Mu::Tn118 during macrophage infection. Perhaps even more significantly, we further demonstrated that phosphorylation and total abundance of S6 ribosomal and Akt proteins were not elevated, and autophagy inhibition is maintained in THP-1-derived macrophages infected with Mu::Tn118C′ (+aTCN), as observed in Mu1615 and Mu::Tn118 infection ( Figures 5E, F , and Supplementary Figure 6 ).

Although an in vitro system to study mycolactone function during bacterial infection is a valuable tool, studying these interactions in vivo is vital to confirming the physiological relevance of such studies. To validate if the inducible system described above works in vivo, we utilized the mouse footpad model of BU infection. C57BL/6J mice were infected in the left hind footpad with 1 × 10⁴ CFU, and footpad swelling was monitored for 14 weeks. Mice were infected with Mu1615 and Mu::Tn118C′. Mu::Tn118C’-infected groups were left untreated or treated with 1 mg/ml doxycycline in drinking water for ad libitum administration for 14 weeks of the experimental period (duration), or treatment was started 2 weeks postinfection (delayed) and maintained for the following 12 weeks. Doxycycline treatment resulted in footpad swelling in infected mice like that observed during Mu1615 infection ( Figure 6A ). Mu1615 and Mu::Tn118C′ (+doxycycline)-infected mice showed increased bacterial persistence for 4 and 10 weeks postinfection than Mu::Tn118C’-infected mice that did not receive doxycycline treatment ( Figure 6B ). This result was further confirmed by the increased presence of acid-fast bacilli as observed by histology ( Figure 6C ). Hematoxylin and eosin (H&E) staining also revealed increased cellular infiltrate during infection with Mu1615 and Mu::Tn118C′ (+doxycycline) than seen with Mu::Tn118C′ (-doxycycline). As we utilized a relatively low-dose infection model, by 14 weeks postinfection, we did not observe the establishment of an extensive necrotic infection in the footpad.

Our data taken together demonstrate an essential role for mycolactone during Mul infection. Using a tetracycline-inducible mycolactone synthesis system, we have identified an important role for mycolactone in inducing necrosis during Mul infection. Mycolactone synthesis is essential for bacterial egress from macrophages and the establishment of necrotic lesions during infection. Bacterial synthesis of mycolactone is unable to overcome mTOR induction and autophagy inhibition by Mul, unlike previous observations that mycolactone is a potent mTOR inhibitor and autophagy activator. Thus, the tetracycline-inducible mycolactone synthesis system developed in this study will help assess the mechanisms by which mycolactone induces necrosis during Mul infection.