M. tuberculosis H37Rv and M. canettii CIPT 140010059 were used in this study. Mycobacteria were cultivated in Middelbrook 7H10 (Becton Dickinson, Le Pont de Claix, France) supplemented with 10% oleic acid-albumin-dextrose-catalase (OADC) (Becton Dickinson). Fluorescent M. canettii CIPT 140010059 mCherry (CSURP3621) was constructed by transforming the pMV261 mCherry vector (gift from L. Kremer, CPBS, Montpellier, France) into the M. canettii strain (Alibaud et al., 2011). Fluorescent strains were cultured in Middlebrook 7H10 broth supplemented with 10% OADC and 50 μg/mL kanamycin (Sigma). Prior to infection, mycobacteria were resuspended in phosphate buffered saline (PBS), shaken on a vortex mixer for 10 min with glass ball to disperse clumps and centrifuged for 1 min at 300 g to remove residual clumps. The supernatant was then dispersed by expelling the suspension 10 times through a sterile 25-gauge needle attached to a 1-mL syringe. Calibration was then performed according to Mcfarland standard confirmed by counting mycobacteria after Ziehl-Neelsen staining (RAL diagnostics, Martillac, France). All experiments using these mycobacteria were performed in a biosafety level 3 laboratory of the Faculté de Médecine, Aix-Marseille Université, France.

The experiments conducted on mice was approved by the Institutional Animal Care and Use Committee of Aix-Marseille University “C2EA-14,” France and registered by the “Ministère de l'Enseignement Supérieur et de la Recherche” under reference n° 2015092415474605. Mice were handled according to the rules of Décret N° 2013–118, Février 7, 2013, France. Inguinal WAT and inter-scapular BAT were collected from ten 6-week-old Balb/cByj mice. Pre-adipocytes cells were isolated from tissues as described elsewhere (Aune et al., 2013) and then cultured in complete culture medium Dulbecco's Modified Eagle Medium DMEM/F12 (Invitrogen, France) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and were plated at a concentration of 5 × 10⁴ cells/ml/well in 12-well plates and incubated at 37°C under a 5% CO₂ atmosphere.

Murine embryonic fibroblasts 3T3-L1 (ATCC, lot 62158491) were cultured in DMEM (Invitrogen, France) supplemented with 10% heat-inactivated FBS at 37°C under a 5% CO₂ atmosphere. Prior to confluence, cells were plated at 5 × 10⁴ cells/well in a 12-well sterile plates and incubated as described above. Six days after plating, pre-adipocytes reach confluency and stop dividing. Adipocyte differentiation was initiated upon confluency by adding 1 μg/mL insulin (Sigma), 0.5 mM isobuthyl-methylxanthine (Sigma), and 1 μM dexamethasone (sigma) into the culture medium. After 48 h of induction, the medium was replaced by DMEM-10% FBS containing 1 μg/mL of insulin and maintained each 2 days in order to achieve a full differentiation as previously described (Neyrolles et al., 2006). Adipocyte differentiation was checked by BODIPY 493/503 staining (Sigma) of lipid droplets over the 2-week experiments.

Pre-adipocytes and mature adipocytes were infected with a multiplicity of infection (MOI) of 1 per cell at day 6 after plating and at day 10 post-induction respectively. Accordingly, 1 × 10⁵ cells per well were infected with 1 × 10⁵ mycobacteria. After 4-h incubation, cells were washed 3 times with serum-free DMEM to remove all extracellular bacteria. Only mature adipocytes were then incubated with 200 μg/mL amikacin for 2 h and then washed thrice. The last wash was plated to validate the correct elimination of extracellular mycobacteria. The inoculated cultures were incubated with fresh complete medium up to 7 days for primary cultures and 12 days for 3T3-L1 cell lines at 37°C under a 5% CO₂ atmosphere. At various time points, infected cells were lysed in PBS containing 0.1% Triton X-100 and the lysate was plated at 10-fold dilutions from 10⁻¹ to 10⁻⁴ onto Middlebrook 7H10 agar plates incubated at 37°C under a 5% CO₂ atmosphere. The number of colonies enumerated after 20-day of incubation was used to estimate the number of colony-forming units (CFUs) per 10⁵ cells. All experiments were conducted in triplicate: three independent plates the same day for primary cell infection experiments and three replicates at different days with different bacterial cultures for 3T3-L1 cells infection experiments.

Ziehl-Neelsen staining (RAL diagnostics, Martillac, France) was performed on infected cells (MOI 1:1 and 5:1) at day 3 p.i. Observations were acquired using the slide scanner Axio Scan.Z1 (Zeiss) (Magnification 20X) with a color Hitachi tri-CCD (1,800 × 1,200 pixels) camera. The number of intracellular mycobacteria per one cell was recorded on image acquisition by ImageJ software.

M. canettii mCherry was used to localize intracellular bacteria using confocal microscopy. Infected adipocytes (MOI 10) were fixed using 4% formaldehyde at day 3 p.i. Lipid droplets were stained with BODIPY 493/503 (Sigma). The images were acquired using a confocal microscope (Zeiss, Spinning disk with a Yokogawa head and an emCCD camera 512 × 512 pixels) at magnification 63X/1.4 oil objective.

Cells were fixed at room temperature with 2.5% glutaraldehyde in Na-cacodylate buffer 0.1 M (pH 7.2) containing 0.1 M sucrose, 5 mM CaCl₂, and MgCl₂ 5 mM, washed with complete cacodylate buffer and postfixed for 1 h at room temperature with 1% osmium tetroxide in the same buffer without sucrose (de Chastellier, 2008). They were washed with buffer, scratched gently, concentrated in agar up to 2% with cacodylate buffer and processed for 1 h at room temperature with 1% uranyl acetate in maleate buffer. The samples were dehydrated in a graded series of ethanol solutions and gradually incorporated in Spurr resin. Thin sections (80 nm thick) were stained with 1% uranyl acetate in distilled water and then with lead citrate before being observed by electron microscopy.

Adipocytes were infected with M. canettii at MOI 5:1 on 12-well plate. At day 7 p.i. all wells were lysed in 500 μL water containing 0.1% Triton X-100. The pooled lysate was sonicated and centrifuged at 9,000 g for 25 min. The M. canettii cells were washed thrice with 0.1% Triton X-100 and the pellet was resuspended in 500 μL of PBS containing 0.05% Tween 80. Nile red staining was performed as previously described (Christensen et al., 1999). The stained smears were subjected to high resolution confocal (Zeiss AiryScan head, magnification 63X, Numerical aperture = 1.4), with higher resolution (1.6X) than conventional confocal.

LDH, a soluble cytoplasmic enzyme released into extracellular space when the cell membrane is damaged, was used as an estimator of cell lysis (Korzeniewski and Callewaert, 1983). Culture supernatants of uninfected or infected mature adipocytes with M. canettii or M. tuberculosis (MOI 1) were collected at day 3, 7, and 12 post-inoculation. The release of LDH from negative controls and infected cells was measured spectrophotometrically at 340 nm on a Cobas 8000 (Roche, Meylan, France).

Statistics were performed using the SigmaPlot13 software. The distribution of the variables was assessed statistically with the Shapiro-Wilk test. The equality of the variance was checked using the Brown-Forsythe test. All data were normally distributed and were expressed using means ± standard deviation. The statistical significance was performed using the Student T-test. The Chi-square test was used to compare rates and proportions with Yates correction. A p <0.05 was regarded as significant.

Pre-adipocytes purified from murine inguinal WAT and inter-scapular BAT were ex vivo inoculated with M. canettii or M. tuberculosis for 4 h p.i. For all experiments, extracellular mycobacteria were removed by extensive washing with serum-free medium, 4 h p.i. At day 3 p.i. Ziehl-Neelsen staining was performed to assess the ability of mycobacteria to infect primary murine pre-adipocytes. While negative control cells remained free of detectable mycobacteria, microscopic observations clearly showed intracellular M. canettii and M. tuberculosis in white and brown pre-adipocytes (Figure 1).

The ratio of infected cells relative to the total number of observed cells as well as the average number of intracellular bacilli per one cell were measured on a random selection of 10 microscope fields on stained infected cells at day 3 p.i. (see Supplementary Data S1). White pre-adipocytes infected at MOI 1:1 yielded 19% and 18% of infected cells with 9 ± 6 M. canettii bacilli per cell and 7 ± 5 M. tuberculosis organisms per cell, respectively. Likewise, 22 and 19% of brown pre-adipocytes were infected with 16 ± 7 M. canettii organisms/cell and 11 ± 4 M. tuberculosis organisms/cell, respectively. The ratio of infection did not significantly differ between brown and white pre-adipocytes (p = 0.7 for M. canettii and p = 0.8 for M. tuberculosis, χ² test) but the number of bacilli in brown pre-adipocytes was significantly higher than in white pre-adipocytes (p = 0.02 for M. canettii and p = 0.03 for M. tuberculosis).

In order to confirm these observations and to study the viability of bacilli into the infected cells, the number of intracellular mycobacteria was quantified by scoring CFUs/10⁵ cells after plating infected cell lysates with M. canettii or M. tuberculosis at MOI 1:1 at different time intervals (Figure 2) (see Supplementary Data S1). At day 0, white pre-adipocytes had internalized 15% (1.5 × 10⁴ ± 9 × 10³ CFUs) and 10% (9 × 10³ ± 4 × 10³ CFUs) of the initial 1 × 10⁵ inoculum of M. canettii and M. tuberculosis, respectively. As for brown pre-adipocytes, 45% (4.5 × 10⁴ ± 2.3 × 10³ CFUs) of the M. canettii inoculum and 25% (2.47 × 10⁴ ± 4 × 10³ CFUs) of the M. tuberculosis inoculum was phagocytized. By scoring CFUs, intracellular M. canettii and M. tuberculosis organisms were more abundant in primary brown pre-adipocytes than in white pre-adipocytes (p = 0.01 for M. canettii and p = 0.008 for M. tuberculosis) which is consistent with our microscopic observations.

A similar intracellular behavior was observed for M. canettii and M. tuberculosis during the 7-day experiment in primary white and brown pre-adipocytes: both mycobacteria survived and multiplied with approximately one log gain.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.