3-Chloro-5(S)-[(1R,2S,5R)-2-isopropyl-5-methylcyclohexyloxy]-4-[4-methylphenylsulfonyl]-2(5H)-furanone (F105) was synthesized in three steps. For the detailed description of compounds preparation and characterization, we refer to the Supplementary Data, available online. The stock solutions of F105 were prepared by diluting powders in pure DMSO (Sigma–Aldrich, Saint-Quentin Fallavier, France) at the concentration of 20 g/L. To solubilize the furanone at high concentrations in medium, pluoronic acid F-127 (Sigma–Aldrich) (10% stock solution in DMSO) was added to the final concentration of 0.1%. Working solutions were prepared in bacterial growth medium such that the final concentrations of DMSO 5% have been obtained, which was next verified to be non-toxic for the bacterial strains tested. All conventional antibiotics were purchased from Sigma.

A methicillin sensitive Staphylococcus aureus ATCC^®29213 (MSSA) and a methicillin resistant S. aureus strain ATCC^®43300 (MRSA) (both laboratory strains) as well as 10 clinical MRSA isolates provided by the Republican Clinical Hospital, Laboratory of Clinical Bacteriology in Kazan were used in this study (see Table 1). The bacterial strains were stored in 10% (V/V) glycerol stocks at -80°C and freshly streaked on blood agar plates (BD Diagnostics) following by their overnight growth at 35°C before use. Fresh colony material was used to adjust an optical density to 0.5 McFarland (equivalent to 10⁸ cells/mL) in 0.9% NaCl solution that was used as a working suspension.

The MIC of furanone F105 was determined by the broth microdilution method in 96-well microtiter plates (Eppendorf) according to the EUCAST rules for antimicrobial susceptibility testing (Leclercq et al., 2013) with minor modifications to account for the decreased solubility of the compound. Briefly, the 10⁸ cells/mL bacterial suspension was subsequently diluted 1:300 with Mueller-Hinton broth (MH) (Carl Roth GmbH, Germany), cation-adjusted with 20 mg/L Ca²⁺ and 10 mg/L Mg²⁺ and supplemented with various concentrations of F105 in microwell plates to obtain a 3 × 10⁵ cells/mL suspension. To solubilize the furanone at high concentrations in medium, pluoronic acid F-127 (Sigma–Aldrich) (10% stock solution in DMSO) was added to the final concentration of 0.1%. The final concentration of DMSO was adjusted to 5% in all bacterial cultures. The concentrations of F105 ranged from 1.25 to 160 mg/L. Besides the usual double dilutions, additional concentrations were included in between. The cultures were incubated at 35°C for 24 h. The MIC was determined as the lowest concentration of furanone for which no visible bacterial growth could be observed after 24 h of incubation.

To determine the MBC, the CFU/mL were further evaluated in culture liquid from wells without visible growth. The F105 concentration reducing the number of viable cells by at least three orders of magnitude was considered as MBC according to the recommendation of the European Committee for Antimicrobial Susceptibility Testing (EUCAST) of the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) (2000).

The checkerboard assay was performed similarly to the MIC testing in 96-well microtiter plates (Eppendorf). Each plate contained serial dilutions of F105 in 0.1% pluoronic acid F-127 and different antibiotics in a checkerboard fashion as described previously (Eliopoulos and Moellering, 1996). Briefly, the final concentrations of both compounds ranged from 1/16× to 4× MIC for F105 and from 1/256× to 4× MIC for the antibiotics. In total, 11 dilution steps of antibiotics and 7 dilution steps of F105 were analyzed. The microwell plates were incubated at 35°C for 24 h. Each test was performed in triplicate and included a growth control with neither antibiotic nor F105 addition. The fractional inhibitory concentration index (FICI) for each double combination was calculated as

The FICIs were counted from the concentrations in the first non-turbid well found in each row and column along the turbidity/non-turbidity interface and the lowest FICI value was used to characterize the synergy. For the FICI interpretation we refer to den Hollander et al. (1998) and Odds (2003): FICI < 0.5 corresponds to synergy, 0.5 < FICI < 4 corresponds to either additive effects or indifference, while FICI > 4 corresponds to antagonism.

To determine the BPC of F105, two methods have been applied, including the modified crystal-violet staining (Merritt et al., 2005) and the drop plate approach (Herigstad et al., 2001). Briefly, the bacterial culture was adjusted to 5 × 10⁵ cells/mL in the MH broth and seeded into 24-well polystyrene culture plates (Eppendorf). F105 was added in serial dilutions to the final concentrations between 1.25 and 160 mg/L following by the cells growth under static conditions for 24 h at 35°C. For the crystal violet staining, the liquid culture was removed after 24 h of incubation and the plates were washed twice with PBS (pH 7.4) and dried. Then 1 ml of the 0.5% crystal violet solution (Sigma) in 96% ethanol was added per well, followed by 20 min incubation. Next the crystal violet solution was removed and the plate was washed 3 times with PBS. After 30 min air drying, 1 ml of 96% ethanol was added to resolubilize the bound crystal violet, and the absorbance was measured at 570 nm with the microplate reader Infinite 200 Pro (Tecan). Cell-free wells incubated with pure medium subjected to all staining manipulations were used as control.

Alternatively, the viability of cells was evaluated by the drop plate approach with minor modifications (Herigstad et al., 2001). Serial 10-fold dilutions from each well were prepared and 5 ml of suspension was dropped onto LB agar plates in five repeats. CFU/mL were counted and averaged from those drops containing 5–30 colonies. To evaluate the viability of biofilm-embedded cells, wells were washed several times with phosphate-buffered saline (PBS) to remove both non-adherent and detached cells. The washed biofilms were suspended in PBS by scratching the well bottoms with following treatment in a sonicator bath for 2 min at 20 kHz to favor the disintegration of bacterial clumps, and viable cells were counted by the drop plate method as described above.

Cytotoxicity of F105 was determined using the CellTiter 96^® Aqueous Non-Radioactive Cell Proliferation Assay (Promega) using MCF-7 cells. The cells were cultured in DMEM – Dulbecco’s Modified Eagle’s Medium (Sigma–Aldrich) supplemented with 10% FBS, 2 mM L-glutamine, 100 mg/L penicillin and 100 mg/L streptomycin. Cells were seeded in 96-well plates with the density of 3000 cells per well and left overnight to allow for the attachment. Cells were next cultured at 37°C and 5% CO₂ in the presence of F105 at various concentrations from 1.25 to 160 mg/L. After 24 h of cultivation the cells were subjected to MTS-assay based on the cellular reduction of MTS (3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) by the mitochondrial dehydroxygenase using phenazinemethosulfate (PMS) as the electron coupling reagent. The MTS tetrazolium compound was bioreduced by viable cells into a colored formazan product, which was measured on Tecan Infinite 200Pro at 550 nm. The concentration required to inhibit cellular dehydrogenase activity by 50% (CC₅₀ value) was calculated as recommended by the manufacturer.

The mutagenicity of F105 was evaluated in the Ames test with Salmonella typhimurium TA98, TA100 and TA102 strains as described in McCann and Ames (1976). The tested compound was considered to be mutagenic if the number of revertant colonies in the experiment was more than two times higher than that in the negative control (DMSO) and increased at higher F105 concentrations (OECD, 1997).

All experiments were performed in biological triplicates with three repeats in each run. The data were analyzed and graphically visualized using GraphPad Prism version 6.00 for Windows (GraphPad Software, United States, www.graphpad.com). In each experiment, comparison against negative control has been performed using the non-parametric Kruskal–Wallis one-way analysis of variance test. Significant differences against respective controls were considered at p < 0.05 and are specified in the corresponding figure captions. Additionally, statistical significance of time-kill curves and dose-response curves have been assessed by linear regression analysis applied in log scales where appropriate. For the regression analysis data, shown confidence intervals also correspond to p < 0.05.

The 2(5H)-furanone derivative F105 (4) was synthesized in three steps from commercially available mucochloric acid 1 (see Figure 1). In the first stage, a mixture of diastereomers 2a + 2b was obtained in the reaction of mucochloric acid 1 with l-menthol in the presence of catalytic amounts of concentrated sulfuric acid as described previously (Fenske and Merzweiler, 1989). The pure stereoisomer 2a with S-configuration of the carbon atom C⁵ of the γ-lactone ring was isolated in 52% yield after two recrystallizations from hexane (Figure 1).

Next a thiolation reaction of the isolated stereoisomer 2a under basic catalysis was carried out. It is well-known that in the presence of triethylamine reactions of 5-alkoxy-2(5H)-furanones with thiols proceed with the regioselective substitution of the chlorine atom in the fourth position of the lactone ring (Kurbangalieva et al., 2007; Latypova et al., 2014). The reaction was performed in diethyl ether at room temperature with the equimolar ratio of furanone 2a, p-thiocresol and triethylamine resulting in the novel optically pure thioether 3 with 85% yield. Thioether 3 was further subjected to oxidation to the corresponding sulfone 4 using a recently developed selective method (Latypova et al., 2014). The compound 3 was treated with 10-fold excess of 33% hydrogen peroxide in acetic acid at room temperature and the novel optically pure sulfonyl derivative of 2(5H)-furanone 4 (studied compound F105, Figure 1) was isolated in the form of colorless crystals in 70% yield. The structure of compounds 2–4 was characterized in detail by IR and NMR spectroscopy (Supplementary Figures S1–S5).

The antimicrobial properties of F105 were determined on both methycillin-susceptible (MSSA) and -resistant (MRSA) S. aureus strains ATCC29213 and ATCC43300, respectively. The minimal inhibitory concentration (MIC) of F105 for MSSA was found to be 10 mg/L (25 μM), and 20 mg/L (50 μM) for MRSA (see Table 1). The minimal bactericidal concentration (MBC) value of F105 was found to be 40 mg/L in MSSA and 80 mg/L in MRSA. The time-kill curves revealed that all cells of MSSA exposed to F105 at concentration of 2× MBC were killed within 8 h of treatment (Figure 2A). Alternatively, 1× MBC of F105 led to the reduction in the number of viable cells by three orders of magnitude within 12 h. The dose-response curves (Figure 2B) confirm the concentration-dependence of F105 effect on cells viability explicitly suggesting that F105 exhibits biocidal activity. Interestingly, our earlier results indicate that the analogs of F105 lacking either sulfonyl group (compound 3 on the Figure 1) or l-menthol moiety (F70, 3-chloro-5-hydroxy-4-[(4-methylphenylsulfonyl)]-2(5H)-furanone) exhibited no activity against S. aureus (Latypova et al., 2014; Kayumov et al., 2015a) (data not shown) suggesting the requirement of both functional groups for bactericidal activity, while in other works the presence of either sulfonyl group or l-menthol moiety increased antibacterial effects (Meadows and Gervay-Hague, 2006; Kudryavtsev et al., 2009; Rogers et al., 2010; Low et al., 2011).

We additionally assessed both MIC and MBC in 10 clinical isolates exhibiting a MRSA phenotype. Both values ranged within the MIC and MBC found in the MSSA laboratory standard strain with all MICs at 10 mg/L, and MBC at 20 mg/L (n = 5) or 40 mg/L (n = 5) (Table 1).

The synergism of F105 in combination with various antibiotics on the bacterial growth was analyzed by the checkerboard assay (Table 2 and Supplementary Table S1). Thereby synergism of F105 was observed when combined with aminoglycosides. Particularly, the FICI values for F105 were determined to be 0.33 ± 0.04 in combination with amikacin, 0.33 ± 0.16 with gentamicin and 0.44 ± 0.17 with kanamycin. Besides aminoglycosides, strong synergy has been observed also for benzalkonium chloride with FICI of 0.29 ± 0.09. Indifferent effects, but with relatively low FICI (between 0.51 and 0.75) indicating enhancement of antimicrobial activity were observed for F105 in combination with daptomycin, rifampicin, ciprofloxacin, tetracycline, and linezolid; whereas the only additive effects were observed for combinations of F105 with erythromycin and vancomycin. For deeper analysis of F105 synergy with antimicrobials, the combined MICs of the antibiotic/F105-mixtures were plotted (as isoboles) and the effective concentrations of F105 (EC₅₀) leading to the twofold reduction of antibiotic’s MIC were calculated (Table 2 and Supplementary Figure S6). The EC₅₀ values of F105 in combinations with amikacin, gentamycin, and kanamycin were determined to be 0.7, 0.7, and 1.3 mg/L, respectively. Only 0.5 mg/L of F105 was required to reduce the MIC of benzalkonium chloride twofold. For other studied antibiotics EC₅₀ values of F105 were in the range of 4.3 and 9.5 mg/L.

Since the anti-biofilm activity of furanones was reported for some compounds (Ren et al., 2001; Hentzer et al., 2002; Lonn-Stensrud et al., 2009; Fedorova et al., 2013), the biofilm preventing concentration (BPC) of F105 for the MSSA laboratory standard strain was assessed first by two different methods. The crystal violet staining method revealed that F105 completely inhibited the biofilm formation of the MSSA at the concentration of 20 mg/L (Figure 3A); whereas, applying direct counting of viable cell in the biofilm the BPC was found to be 40 mg/L (Figure 3B and Table 1). Since the amount of viable planktonic cells (swimming cells) also decreased by three orders of magnitude at 20 mg/L of F105, we suggested that the biofilm suppression was rather the consequence of cell growth repression. The CLSM analysis confirmed the BPC of F105 of 40 mg/L for MSSA (Figure 4), thus we estimated that these direct methods are more adequate compared to the indirect crystal violet staining. The determined BPC of MRSA was 80 mg/L according to the CFU/ml counting and 40 mg/L according to the CLSM (Table 1 and Figure 4). The BPC was measured by the drop plate method in the clinical MRSA isolates resulting in 20 mg/L in nine and 40 mg/L in one isolate (Table 1).

To further analyze, whether F105 not only inhibits the biofilm formation but also exhibits activity against already established biofilms, 24 h-old S. aureus biofilms of the MSSA and MRSA laboratory standard strains were treated with different concentrations of F105. The CLSM analysis of the MBEC indicated that the treatment with F105 did not lead to any visibly remarkable decrease of the biofilm thickness (Figure 5), while the ratio of dead/viable cell increased significantly in the concentration dependent manner. Biofilm-embedded cells of MRSA were nearly completely killed at 80 mg/L of F105, whereas MSSA cells were killed at 160 mg/L. However, the CFU/mL of the planktonic cells and the adherent cells of MSSA biofilm cultures revealed a reduction in the number of viable cells by three orders of magnitude at 80 mg/L of F105 for both attached and detached cells (Figure 6) indicating that CLSM results vary depending on the screened section and the CFU/mL values are more reliable.

To compare the biofilm activity of F105 against other antimicrobials, similar experiments were performed with ampicillin, gentamicin, and benzalkonium chloride (see Supplementary Table S2 for MBC values). A fourfold MBC of benzalkonium chloride was required to reduce the number of viable biofilm-embedded cells by three orders of magnitude, while even 16-fold MBC of gentamicin or ampicillin led to only 10-fold decrease of viable cells in the biofilm (Figure 7). In marked contrast, twofold MBC of F105 reduced the viable cells of the MSSA biofilm by more than four orders of magnitude (Figure 7 and Supplementary Figure S7).

No mutagenicity of F105 was detected in the Ames test using Salmonella typhimurium TA98, TA100, and TA102 strains at the furanone concentrations up to 80 mg/L without (Supplementary Table S3) and with (Supplementary Table S4) metabolic activation. The cytotoxicity test provided with CC₅₀ values of 40.3 mg/L for MCF-7 cells (a human breast adenocarcinoma cell line) but only 8.1 mg/L for skin fibroblast suggesting that the current structure of F105 might be too toxic for direct use for systemic therapeutic in humans and requires further modification of the structure in order to reduce its toxicity and improve its solubility.