*2.7. Measurement of Solutes Uptake*

Solutes uptake was measured as total cell-associated counts after addition of quinones on cultures of *B. subtilis* growing exponentially (OD550 ≈ 0.2). Growing cells containing half concentration of NB medium were transferred to prewarmed flasks containing celastrol (3 μg/mL) or pristimerin (10 μg/mL) and radiolabeled glucose (2 μCi/mL D-[1- 14C]-glucose) or the precursors of DNA, RNA, protein, and cell wall peptidoglycan (see concentrations in previous section). At different times (up to 30 min), 0.5 mL of samples was collected and filtered through Millipore filters of 0.45-μm pore size (Type HA, Millipore Corporation, Burlington, MA, USA). Filters were washed three times with 5 mL phosphate buffer, dried, and radioactivity was measured as above. Samples with clofoctol (5 μg/mL) or DMSO at the same concentration were included as positive and negative controls, respectively.

Furthermore, the effect of celastrol on the uptake and incorporation of radiolabeled precursors of DNA was also evaluated when the macromolecular synthesis was inhibited. *B. subtilis* cultures prepared as described above were treated for 30 min with the specific inhibitor ciprofloxacin (1.5 μg/mL) before the addition of labeled and unlabeled precursor. After 5 min, celastrol (3 μg/mL) or the same proportion of DMSO in control cultures was incorporated. At different times, precursor uptake and incorporation was measured as mentioned above. The assays were repeated at least three times.

#### *2.8. Inhibition of DNA Gyrase*

Inhibition of DNA gyrase was performed using the Gyrase Supercoiling kit 1 (#K001) (John Innes Enterprises Ltd., Norwich, UK) as described in the manufacturer's instructions. Two units of DNA gyrase were incubated with 0.5 μg relaxed plasmid pBR322 and 50 μg/mL of celastrol or pristimerin in a final volume of 30 μL. Ciprofloxacin at 25 and 50 μg/mL was used as positive control. The samples were incubated at 30 ◦C for 30 min and the products were visualized in 0.8% agarose gel containing 0.5 μg/mL ethidium bromide.

#### *2.9. Integrity of Cell Membrane*

Bacterial membrane damage was examined by determination of the release of material absorbing at 260/280 nm, detection of potassium (K+) leakage, and the BacLight Live/Dead staining method (Invitrogen Molecular Probes, Eugene, OR, US). *B. subtilis* cultures in log-phase growth (OD550 ≈ 0.8) were centrifuged at 15,000× *g* for 10 min at 4 ◦C and washed twice with saline buffer. The pellet was resuspended in the same buffer to obtain a bacterial concentration of 1–2 × 108 CFU/mL (or 1–2 × 107 CFU/mL for K<sup>+</sup> leakage experiments). Cell cultures were treated with celastrol (3 μg/mL) and incubated at 37 ◦C under shaking. Cultures with clofoctol (10 μg/mL) or DMSO at the same concentration were used as positive and negative controls, respectively. Samples were collected for quantification at different times over a 30-min period. Liberation of cytoplasmic materials was monitored by measuring the optical density (OD260 and OD280) of the supernatant after removing cells by centrifugation (at 9500× *g* for 10 min, 4 ◦C) or after membrane filtration by means of an atomic absorption spectrophotometer (Model Thermo S-Series, Thermo Electron Corporation, Cambridge, UK) for K<sup>+</sup> release. The BacLight assay was analyzed after a 20-min dark-staining period following the manufacturer's instructions. The cells were visualized at ×1000 magnification with an epifluorescence microscope (Leica DM4B, Leica Microsystems GmbH, Wetzlar, Germany) provided with a fluorescein–rhodamine dual filter.

#### *2.10. Transmission Electron Microscopy*

To further confirm the mode of action of celastrol on *B. subtilis*, transmission electron microscopy (TEM) analysis was performed. Suspensions of *B. subtilis* in log-phase growth (107 CFU/mL) were treated with celastrol (3 μg/mL) for 1 h at 37 ◦C and were harvested at 6500× *g* for 8 min at 4 ◦C. For comparative purposes, bacterial cells grown under the same conditions were also treated with pristimerin (10 μg/mL). Subsequently, bacteria were washed in fixative buffer, post-fixed in 1% osmium tetroxide in fixative buffer, and washed with distilled water. Sections (1 μm) were cut with a Reichert Ultracut ultramicrotome and stained with toluidine blue; ultra-thin sections were contrasted with uranyl acetate and lead. Preparations were observed under a Zeiss EM 912 transmission electron microscope. Images were captured with a Proscan Slow-scan CCD-Camera for TEM (Proscan, Scheuring, Germany) and Soft Imaging System software (version 5.2, Olympus Soft Imaging Solutions GmbH, Münster, Germany). Control experiments with the same proportion of DMSO were performed in parallel.

## *2.11. Oxygen Consumption*

Suspensions of *B. subtilis* and *E. coli* in the log phase of growth (OD550 ≈ 0.8) were centrifuged at 15,000× *g* for 10 min at 4 ◦C and washed twice with phosphate buffer 0.1 M (pH 7.0). Then, the pellet was suspended in the same buffer to obtain a bacterial concentration of 1–2 × <sup>10</sup><sup>7</sup> CFU/mL. Cells suspensions (2.7 mL) supplemented with 0.3 mL of 10% glucose were used to measure oxygen consumption at room temperature in a glass cell of Clark oxygen electrode equipped with a magnetic stirrer. Celastrol (at 3 μg/mL for *B. subtilis* and 20 μg/mL for *E. coli*) was added to the cell suspension, and the steadystate output of the oxygen electrode (after 4 min) was measured using a digital biological oxygen monitor (model YSI 5300, Yellow Springs, OH, USA). DMSO in the same proportion and sodium cyanate at 6.7 mM were used as negative and positive controls, respectively. Furthermore, disrupted cell preparations of *B. subtilis* and *E. coli* were used to determine the effect of celastrol on oxygen uptake using NADH (0.1 mM) as a substrate. Cultures grown in yeast extract and peptone (YP, 1% *w*/*v*) medium for 18 h at 37 ◦C under aeration were collected by centrifugation at 10,000× *g* for 10 min at 4 ◦C. Cells were washed twice in 0.1 M potassium phosphate buffer (pH 7.0), resuspended in the same buffer (5 mL of buffer per gram of cells), and sonicated (Labsonic M, Sartorius Stedim Biotech, Göttingen, Germany) for 15 min in 10-s bursts with 20-s stop intervals. Intact cells were removed by centrifugation at 4000× *g* for 10 min at 4 ◦C. The supernatant was centrifuged at 15,000× *g* for 10 min at 4 ◦C and the pellet was resuspended in the same buffer as before (5 mL). Aliquots (0.2 mL) of disrupted cells were incubated at room temperature in the same buffer (2.8 mL) containing NADH (0.1 mM).

#### *2.12. Statistical Analysis*

Three independent experiments were conducted for each evaluation and means and standard deviations (±SD) were calculated. Analysis of variance (one-way ANOVA) followed by Tukey's multiple comparison test (*p* < 0.05) to extract the specific differences between treatments were performed using R statistical software environment version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria).

#### **3. Results and Discussion**

#### *3.1. Antimicrobial Activity*

Table 1 shows the MICs and MBCs of celastrol and pristimerin against different microorganisms used in this study. Celastrol was active against all Gram-positive bacteria evaluated, with MIC values ranging between 0.16 (*B. subtilis*) and 2.5 μg/mL (*S. saprophyticus*). Compared to celastrol, pristimerin showed weaker activity on Gram-positive bacteria and no action on *S. aureus* and *M. smegmatis*. On the basis of these results, *B. subtilis* was selected to evaluate the mechanism of action of these quinones.

**Table 1.** Minimum inhibitory concentration (MIC) <sup>1</sup> and minimum bactericidal concentration (MBC) <sup>1</sup> of celastrol and pristimerin, expressed in μg/mL against different microorganisms 2.


<sup>1</sup> Values represent average obtained from a minimum of three experiments. <sup>2</sup> The quinone compounds were inactive against Gram-negative bacteria and the yeast assayed (MIC > 40 μg/mL).

Gram-negative bacteria and the yeast *C. albicans* were insensitive to the action of both compounds at the maximum concentration assayed (40 μg/mL). The inactivity of pristimerin against *S. aureus* was previously reported by our group [24], although in later works, da Cruz et al. [42] indicated MIC values of 25 μg/mL. The use of a bacterial strain of *S. aureus* less sensitive to the activity of the compound could explain this difference in the results. Gullo et al. [23] reported antimicrobial activity for pristimerin against *C. albicans* with an MIC of 250 μg/mL. In our evaluations, the maximum concentration tested was 40 μg/mL since higher concentrations of both compounds led to solubility problems.

These products differ in the functional group in ring E (Figure 1). As previously reported for other methylene quinones and phenolic nor-triterpenes compounds, replacement of the carboxylic group present in celastrol on C-29 by a methyl ester group reduces the antibacterial activity [14,43]. Celastrol has shown interesting pharmacological activities, although inconveniences related to poor water solubility, high toxicity, or poor stability have also been described [29,44]. Structural modifications in the triterpene quinones scaffolds could be of interest to obtain derivatives with improved antimicrobial activities and to overcome the pharmacokinetic limitations of these compounds.

**Figure 1.** Structure of triterpenoid methylene quinones.
