Reagents for the synthesis of SeNPs: chitosan, bovine serum albumin, sodium selenite, ascorbic, and acetic acid were purchased from Sigma-Aldrich. Reagents for electron microscopy: glutaraldehyde, cacodylate and formaldehyde were purchased from Sigma. Uranyl formate was purchased from Electron Microscopy Sciences (Hatfield, PA, United States).

Chitosan-stabilized SeNPs (Ch-SeNPs) were synthesized following the procedure described by Bai et al. (2008). Briefly, 10 mL of an aqueous chitosan polysaccharide solution (0.5% w/v) were mixed with 7.5 mL of ascorbic acid 0.23 M and 5 mL of acetic acid 2.4 M. Then, 0.25 mL of sodium selenite 0.51 M were slowly added to the previous solution. SeNPs formation was recognized by observing the change of the solution from colorless to red as the reaction progressed. After the synthesis, the colloidal suspension was diluted to 50 mL with distilled water, resulting in final concentrations of 200 mg L^–1 of Se and 0.1% of chitosan. Finally, the colloidal suspension was dialyzed for 2 h at room temperature in a ratio of 10 mL against 2 L of distilled water and using a 12 kDa of MWCO membrane.

For control experiments, BSA-SeNPs were also prepared following the same procedure but using an aqueous bovine serum albumin solution (0.5% w/v) instead of the chitosan solution.

Transmission electron microscopy (TEM) was performed using a JEOL JEM 1400 instrument operated at 120 kV equipped with a CCD camera (KeenView Camera) and an energy dispersive X-ray spectroscopy (EDS) analyzer. Sample preparation was performed by placing one or two drops of the SeNPs colloidal suspension onto carbon-coated copper grids.

Electrophoretic mobility measurements of the materials suspended in water were used to calculate the zeta potential (ζ) values of the nanoparticles. Measurements were performed in a Zetasizer Nano ZS (Malvern Instruments Ltd., United Kingdom) equipped with a 633 nm “red” laser. The hydrodynamic size of nanoparticles was measured by dynamic light scattering (DLS) with the same Malvern instrument.

Mycobacterium smegmatis mc²155 (Msm) and Mycobacterium tuberculosis H37Rv (Mtb) strains (obtained from the ATCC) were grown in Middlebrook 7H9 medium supplemented with 10% (v/v) OADC supplement (NaCl 8.5 g/L, BSA fraction V 50 g/L, dextrose 20 g/L, 5% (v/v) oleic acid solution 1%, 40 mg/L catalase), 0.5% (v/v) glycerol and 0.05% Tyloxapol (v/v; Sigma). Cultures were grown at 37°C in static standing 25 cm² flasks with vented caps. Msm-Lux was generated by transforming Msm with the plasmid construct pMV306hsp + Lux (which contains the entire bacterial Lux operon cloned in a mycobacterial integrative expression vector) (Andreu et al., 2010).

Minimum inhibitory concentration (MIC) assay was carried out in 96-well plates diluting exponentially growing Msm or Mtb at an initial density of 1 × 10⁵ bacteria/mL in the presence of two-fold serial dilutions of Ch-SeNPs or BSA-SeNPs starting at 50 μg/mL and including one control sample with no NPs. Plates were incubated at 37°C for 7 days (Msm) or for 20 days (Mtb). Bacterial growth was monitored every day for Msm and every week for Mtb and was determined by measuring optical density at 570 nm. The assay was performed in triplicate. MIC values were selected as the minimum concentration able to suppress mycobacterial growth. Alternatively, Msm-lux was submitted to MIC assay and cell viability was assessed by measuring relative luminescence units (RLUs) at day 2. Correlation between colony forming units (CFUs) was performed according to a previously performed in vitro RLUs vs. CFUs curve.

Mycobacterium smegmatis cells were fixed with 2% glutaraldehyde in 0.1 M cacodylate at room temperature for 2 h, and then incubated overnight in 4% formaldehyde, 1% glutaraldehyde, and 0.1% PBS. For negative staining of the samples, a drop of the fixed bacterial suspension was applied directly onto a glow-discharged EM grid (QUANTIFOIL. Formvar/Carbon. Cu 400 mesh grids). The sample was allowed to be adsorbed and then blotted with filter paper (Whatman grade No. 1). The grid was washed by touching the surface with two consecutive drops of 0.75% (w/v) uranyl formate, blotting each time, and stained for 1 min with one more drop of the same staining agent. Negative stained samples were examined in a JEOL JEM-1230 (accelerating voltage 100 kV) electron microscope, and images were recorded with a CCD camera ORIUS SC100 (4 × 2.7 k pixel).

Mycobacterium smegmatis and Mtb cultures were fixed as above prior to vitrification. Grids were prepared following standard procedures and observed at liquid nitrogen temperatures in a JEM-2200FS/CR transmission electron microscope (JEOL Europe, Croissy-sur-Seine, France) operated at 200 kV. An in-column omega energy filter helped to record images with improved signal/noise ratio by zero-loss filtering. The energy selecting slit width was set at 9 eV. Digital images were recorded on an UltraScan4000 CCD camera under low-dose conditions at a magnification of 55,058 obtaining a final pixel size of 2.7 Å/pixel.

Selenium nanoparticles were prepared via a redox system in the presence of a biomacromolecule as soft template to control the nucleation and growth of the inorganic nanoselenium. The chemical reduction of selenite with ascorbic acid in the presence of a polysaccharide such as chitosan as stabilizer and capping agent afforded red elemental selenium in colloidal state (Zhang et al., 2010). Based on our previous work, 0.1% of chitosan concentration was selected to prepare the SeNPs (Estevez et al., 2014). Under these conditions, the nanoparticles proved to be colloidally stable. Two months after their synthesis, no flocculated material was present, DLS measurements showed no displacement of the hydrodynamic size distribution and no significant differences in the shape or size of the SeNPs were observed by TEM.

Transmission electron microscopy images show well dispersed Ch-SeNPs that exhibit spherical morphology and homogeneous sizes of around 60–80 nm (Figure 1A). The energy dispersive spectroscopy analysis (Figure 1B) indicates a composition in selenium for the Ch-SeNPs, also being observed in the spectrum the signals for C and O from chitosan and the carbon coated copper grid.

The hydrodynamic size distribution of the Ch-SeNPs measured in the aqueous colloidal suspension (Figure 1C) is monomodal and reasonably narrow, with a maximum centered at 105.7 ± 2.5 nm of diameter, which is in concordance with the smaller size of the inorganic nanoparticle determined in the TEM study. The ζ-potential of the Ch-SeNPs in water is in the zone of colloidal stability with a value of +66.6 ± 4.7 mV. This highly positive value is due to the surface stabilization with chitosan, which possesses a equilibrium for the protonation of the amino groups in water. Therefore, chitosan allows the SeNPs to form stable colloidal suspensions due to both electrostatic as well as steric stabilization.

We evaluated the ability of SeNPs to inhibit the growth of two different species of mycobacteria: the fast-growing Msm and the slow-growing Mtb. SeNPs were found to be effective in inhibiting mycobacterial growth in a dose-dependent manner, showing MIC values of 0.400 μg/mL for Msm (Supplementary Figure S1A) and 0.195 μg/mL for Mtb (Figure 2A). To determine whether this inhibition was concomitant to bacterial cell death we generated a luciferase-expressing reporter strain of Msm (Msm-Lux) (Andreu et al., 2010). Correlation between relative fluorescence units (RLUs) and CFUs was previously determined (Supplementary Figure S2A) and used to calculate viable bacteria after treatment of Msm-Lux with SeNPs (Supplementary Figure S2B). We observed a reduction in CFUs as concentration of NPs increased, indicating the bactericidal effect of SeNPs. By this method MIC was similar to that of previously determined by measuring optical density.

To gain insight into the nature of the inhibitory effect of SeNPs, we performed ultrastructural analysis of Msm submitted to 0.400 μg/mL of SeNPs. Transmission electron micrographs of negative stained samples suggested that Msm exposed to SeNPs manifested reduced integrity of cell envelope leading to extrusion of cytoplasmic material, relative to untreated Msm (Supplementary Figures S1B,C). To capture the antimycobacterial effect of SeNPs in a close-to-native state, treated (0.195 μg/mL) and untreated Msm and Mtb cells were submitted to cryo-EM. This technique revealed more in detail the membrane damage, both in Msm (Supplementary Figures S1D,E) and Mtb (Figures 2B–F). In addition, SeNPs could be spotted in direct contact with the mycobacterial cell wall, suggesting a direct connection between the effect of SeNPs and the reduction of the integrity of the mycobacterial cell envelope.