Anthra[1,2-d][1,2,3]triazine-4,7,12(3H)-triones as a New Class of Antistaphylococcal Agents: Synthesis and Biological Evaluation

The development and spread of resistance of human pathogenic bacteria to the action of commonly used antibacterial drugs is one of the key problems in modern medicine. One of the especially dangerous and easily developing antibiotic resistant bacterial species is Staphylococcus aureus. Anthra[1,2-d][1,2,3]triazine-4,7,12(3H)-triones 22–38 have been developed as novel effective antistaphylococcal agents. These compounds have been obtained by sequential conversion of 1-amino-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid (1) and 1-amino-4-bromo-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid (2) into the corresponding amides 5–21, followed by subsequent endo-cyclization under the influence of sodium nitrite in acetic acid. Evaluation of the antimicrobial activity of the synthesized compounds against selected species of Gram-positive and Gram-negative bacteria as well as pathogenic yeasts of the Candida genus has been carried out by the serial dilution method. It has been established that anthra[1,2-d][1,2,3]triazine-4,7,12(3H)-triones exhibit selective antibacterial activity against Gram-positive bacteria. Eight, six and seven, out of seventeen compounds tested, effectively inhibited the growth of S. aureus ATCC 25923, S. aureus ATCC 29213 and S. epidermidis ATCC12228, respectively, at a concentration equal to 1 µg/mL or lower. The high antistaphylococcal potential of the most active compounds has been also confirmed against clinical isolates of S. aureus, including the MRSA strains. However, bacteria of the Staphylococcus genus have demonstrated apparent resistance to the novel compounds when grown as a biofilm. None of the four selected compounds 3234 and 36 at a concentration of 64 µg/mL (128 or 256 × MIC—against planktonic cells) has caused any decrease in the metabolic activity of the staphylococcal cells forming the biofilm. The kinetic time–kill assay revealed some important differences in the activity of these substances. Compound 33 is bacteriostatic, while the other three demonstrate bactericidal activity.


Introduction
The resistance of human pathogenic bacteria and yeasts to the well-known antibiotics and synthetic chemotherapeutics is becoming a major global health challenge [1][2][3]. Among the array of various antibiotic-resistant pathogenic bacteria, staphylococci belong to the most common and dangerous.
They are able to produce a wide array of virulence factors and have evolved mechanisms of resistance to a plethora of antibiotics currently in use for human and animal therapies. As a consequence, bacteria of the Staphylococcus genus, particularly S. aureus, are responsible for a broad spectrum of difficult to treat diseases including skin and ocular infections, foodborne illness, pneumonia, meningitis, endocarditis, and osteomyelitis [4]. Over the past decades, this bacterium has become the main cause of bloodstream infections (BSI) in the world, with a mortality level of 20-40% [5]. Recent advances in the developments of antistaphylococcal agents are presented in works on search of new non-antibiotic agents: peptidoglycan hydrolases compounds of plant origin (plant extracts, essential oils and their components) [6][7][8][9], vaccines development [8], silver nanoparticles [10], bacteriocins [11], and bacteriophages [12]. Currently, the most reasonable approach for understanding the crucial cellular pathways by which unknown to date (or not used as antimicrobial agents) molecules affect pathogenic microorganisms at a molecular level seems to be transcriptome analysis. Recently, Chauhan and coworkers [13] successfully applied RNA sequencing for the determination of antistaphylococcal modes of action of novel discovered benzimidazole molecules. Comparative analysis of differentially expressed genes between control and treated S. aureus cells showed selective hits. Therefore, the search and development for new antibacterial agents effective against these pathogenic bacteria remain an urgent task of the present.
Derivatives of the 9,10-anthraquinone of both natural and synthetic origin are well known for their wide range of biological effects. Of particular interest, compounds with an antibacterial effect are known among them [14][15][16][17]. The 1,2,3-triazinone ring is an important biophore fragment for the creation of new promising antimicrobial agents based on the 9,10-anthraquinone platform. This pharmacophore is a structural component of many natural and synthetic biologically active compounds with various therapeutic effects [18,19]. The compounds with a sedative, diuretic, anesthetic, antiarthritic, antitumor, antibacterial, antifungal, antituberculosis, antiviral, antiinflammatory, antihistamine activity were found among 1,2,3-triazine derivatives connected with various carbo-and heterocyclic fragments [20][21][22][23].
It is known that diazotization of the corresponding ortho-amino amides with NaNO 2 or t-BuONO under slightly acidic conditions is one of the most common methods for the formation of the 1,2,3-triazin-4-one ring [15]. Despite the widespread use of this method, condensed carbocyclic 1,2,3-triazin-4-one derivatives are known only for the benzene and naphthalene systems [24]. In the context of the above, herein we report a method for the synthesis of new triazine-containing anthraquinone structures and the study their antibacterial effect for identifying potential antimicrobial agents.
The target compounds 22-38 were synthesized via an intramolecular cyclization reaction of amides 5-21 with nitrous acid generated in situ in acetic acid at room temperature in 63-90% yields (Scheme 2). It should be noted that the use of t-BuONO as a diazotizing agent in the reactions at 20-65 • C proved to be ineffective and gave a low conversion of amides 5-21. Moreover, the presence of a bromine atom at the position 4 of the anthracenedione ring did not provide for a significant impact on the cyclization process.
Molecules 2020, 25 The formation of amides 6,7,9,10,15,16,18 was confirmed by the presence of a broad singlet of proton of the NH group of the amide fragment in the range of 10.03-10.72 ppm in the 1 H-NMR spectra. Also, the signal of protons of the NH 2 group falls into the region of aromatic anthracenedione and benzene protons. The amide protons in N-cyclohexyl-substituted amides 8 and 17 resonated in the aromatic proton region of the quinoid ring.
The carbon atom signal of the carbonyl group of amide residue of 5-21 was observed within the range of 165. .76 ppm in the 13 C-NMR spectra. In the case of fluorine-containing amides 9,10,15,18, the characteristic doublet signals attributable to the C-F bond in the meta or para position of the benzene residue were presented in the range from 158.19 to 163.45 ppm with a coupling constant of J = 241.1 Hz.
The characteristic broad singlet signal of NH proton at 15.37 and 15.45 ppm was presented in the 1 H-NMR spectra of anthratriazinones 22 and 31. In the 13 C-NMR spectra of compounds 22,31, the signal of the carbon atom of the carbonyl group in the triazine ring was shifted by 16 ppm compared to the C=O group in the corresponding amides 5 and 14. Additionally, the structure of anthra [1,2-d] [1,2,3]triazine-4,7,12(3H)-trione 22 was confirmed by X-ray diffraction analysis ( Figure 1). The formation of amides 6,7,9,10,15,16,18 was confirmed by the presence of a broad singlet of proton of the NH group of the amide fragment in the range of 10.03-10.72 ppm in the 1 H-NMR spectra. Also, the signal of protons of the NH2 group falls into the region of aromatic anthracenedione and benzene protons. The amide protons in N-cyclohexyl-substituted amides 8 and 17 resonated in the aromatic proton region of the quinoid ring.
The carbon atom signal of the carbonyl group of amide residue of 5-21 was observed within the range of 165. .76 ppm in the 13 C-NMR spectra. In the case of fluorine-containing amides 9,10,15,18, the characteristic doublet signals attributable to the C-F bond in the meta or para position of the benzene residue were presented in the range from 158.19 to 163.45 ppm with a coupling constant of J = 241.1 Hz.
The characteristic broad singlet signal of NH proton at 15.37 and 15.45 ppm was presented in the 1 H-NMR spectra of anthratriazinones 22 and 31. In the 13 C-NMR spectra of compounds 22,31, the signal of the carbon atom of the carbonyl group in the triazine ring was shifted by 16 ppm compared to the C=O group in the corresponding amides 5 and 14. Additionally, the structure of anthra [1,2d] [1,2,3]triazine-4,7,12(3H)-trione 22 was confirmed by X-ray diffraction analysis ( Figure 1). In structure of compound 22 ( Figure 2) the central tetracyclic system is strongly conjugated and planar (Rms deviation of fitted atoms 0.0399), while the atoms O2 and O3 deviate out of this plane for 0.200(3) and 0.154(3) Å respectively. In crystal molecules packed in columns directed along a crystallographic axis.  The formation of amides 6,7,9,10,15,16,18 was confirmed by the presence of a broad singlet of proton of the NH group of the amide fragment in the range of 10.03-10.72 ppm in the 1 H-NMR spectra. Also, the signal of protons of the NH2 group falls into the region of aromatic anthracenedione and benzene protons. The amide protons in N-cyclohexyl-substituted amides 8 and 17 resonated in the aromatic proton region of the quinoid ring.
The carbon atom signal of the carbonyl group of amide residue of 5-21 was observed within the range of 165. .76 ppm in the 13 C-NMR spectra. In the case of fluorine-containing amides 9,10,15,18, the characteristic doublet signals attributable to the C-F bond in the meta or para position of the benzene residue were presented in the range from 158.19 to 163.45 ppm with a coupling constant of J = 241.1 Hz.
The characteristic broad singlet signal of NH proton at 15.37 and 15.45 ppm was presented in the 1 H-NMR spectra of anthratriazinones 22 and 31. In the 13 C-NMR spectra of compounds 22,31, the signal of the carbon atom of the carbonyl group in the triazine ring was shifted by 16 ppm compared to the C=O group in the corresponding amides 5 and 14. Additionally, the structure of anthra [1,2d] [1,2,3]triazine-4,7,12(3H)-trione 22 was confirmed by X-ray diffraction analysis ( Figure 1). In structure of compound 22 ( Figure 2) the central tetracyclic system is strongly conjugated and planar (Rms deviation of fitted atoms 0.0399), while the atoms O2 and O3 deviate out of this plane for 0.200(3) and 0.154(3) Å respectively. In crystal molecules packed in columns directed along a crystallographic axis.  The shortest intermolecular distances between neighboring molecules in columns are 3.342 Å and 3.424 Å for C···N and C···C contacts correspondingly, thus π-π stacking in columns is observed. Solvate water molecules occupy volume between the columns and bonding together neighboring columns by the system of hydrogen bonds. The parameters of H-bonds in a crystal are shown in Table 1. The combination of π-π stacking in columns along axes a and 2D hydrogen bonds across b and c direction giving 3D net and densely packed crystal with density 1.59 g/cm 3 .  (Table 2). Eight, six and seven out of seventeenths compounds tested inhibited the growth of S. aureus ATCC 25923, S. aureus ATCC 29,213 and S. epidermidis respectively at a concentration equal to 1 µg/mL or lower. This level of activity is comparable to the activity of many antibacterial antibiotics (Table 2) [7,31,32]. The tested antibiotics exhibited activity with the range of MIC from 0.125 to 16.0 µg/mL in the case of staphylococci.
Up to the concentration of 32 µg/mL, no activity was observed against both investigated reference strains of Gram-negative bacteria, namely E. coli ATCC 25,922 and P. aeruginosa ATCC 27,853 nor the human pathogenic yeasts of the Candida genus. Our research revealed also lower activity of investigated antibiotics against Gram-negative bacteria. Six out of seven tested antibiotics did not affect E. coli and P. aeruginosa up to the highest tested concentration-32.0 µg/mL. The high antistaphylococcal potential of selected agents was additionally confirmed against clinical isolates of S. aureus. Interestingly, no difference was observed in the susceptibility of MRSA and MSSA strains (Table 3), which is very important from the clinical point of view. Five of these substances inhibited the growth of all strains tested at the concentration not higher than 1 µg/mL. The four most active compounds, namely 32-34, and 36 were selected for further tests.
Several research groups also synthesized heterocyclic compounds of similar structures. However, according to our best knowledge, none of those compounds exhibited comparable antistaphylococcal activity (MIC values below 1 µg/mL) [33][34][35][36]. Moreover, the results presented by other authors did not indicate similar selectivity of antimicrobial properties (high activity against Gram-positive bacteria and no activity against both Gram-negative bacteria and human pathogenic yeasts) [33][34][35]. In some cases, it is difficult to compare results, because different method of determination of antimicrobial activity (disc diffusion) was used by the authors [33,34].

Bactericidal Effect of Selected Compounds
The structures of four compounds selected for further analysis, namely 32, 33, 34 and 36, are similar. Despite this fact, the kinetic time-kill assay revealed some important differences in the activity of these substances. The respective curves are shown in Figure 3. Compound 33 was apparently bacteriostatic only and did not exhibit any bactericidal effect, while the other three demonstrated a clear such effect, i.e., their action resulted in some reduction of the cell number, even in the case of its concentrations 2 or 4 times higher than MIC. Undoubtedly, compound 34 was the most effective in this respect. A substantial reduction of the log 10 CFU value (from 6 to 0) of the bacterial cell suspension was observed at 4 × MIC after 1 h-long treatment with this compound and at 2 × MIC after 3 h-long treatment. In the case of compound 36, such reduction of the cell number was achieved only at 4 × MIC after 1 h-long treatment. The action of compound 32 resulted in a maximal reduction of log 10 CFU from 6 to 2 after 6 h-long treatment at 4 × MIC. Therefore, the bactericidal potential of compounds tested ranks them in the 34 > 36 > 32 order, whereas compound 33 was not bactericidal at all. For all four compounds, the re-growth of bacterial cells was observed after 24 h-long treatment. This observation indicates that even in the cases when the reduction of log 10 CFU from 6 to 0 was noted after shorter treatments, some residual live cells must have been preserved.

Bactericidal Effect of Selected Compounds
The structures of four compounds selected for further analysis, namely 32, 33, 34 and 36, are similar. Despite this fact, the kinetic time-kill assay revealed some important differences in the activity of these substances. The respective curves are shown in Figure 3. Compound 33 was apparently bacteriostatic only and did not exhibit any bactericidal effect, while the other three demonstrated a clear such effect, i.e., their action resulted in some reduction of the cell number, even in the case of its concentrations 2 or 4 times higher than MIC. Undoubtedly, compound 34 was the most effective in this respect. A substantial reduction of the log10 CFU value (from 6 to 0) of the bacterial cell suspension was observed at 4 × MIC after 1 h-long treatment with this compound and at 2 × MIC after 3 h-long treatment. In the case of compound 36, such reduction of the cell number was achieved only at 4 × MIC after 1 h-long treatment. The action of compound 32 resulted in a maximal reduction of log10 CFU from 6 to 2 after 6 h-long treatment at 4 × MIC. Therefore, the bactericidal potential of compounds tested ranks them in the 34 > 36 > 32 order, whereas compound 33 was not bactericidal at all. For all four compounds, the re-growth of bacterial cells was observed after 24 h-long treatment. This observation indicates that even in the cases when the reduction of log10 CFU from 6 to 0 was noted after shorter treatments, some residual live cells must have been preserved.

Activity of the Selected Compounds against Staphylococcal Biofilm
When the S. aureus cells were grown in the biofilm mode, the antistaphylococcal potential of compounds 32-34 and 36 appeared much lower, if any, than that found against the cells grown in the planktonic mode. None of these compounds caused any decrease in the metabolic activity of the cells forming the biofilm, as measured with the MTT assay, at concentrations as high as 64 µg/mL (128 or 256 × MIC). It is not especially surprising, since many earlier reports indicated that for different antibacterial agents, their effective concentrations inhibiting metabolic activity of staphylococci grown in the biofilm mode can be 10-1000 times higher than those inhibiting the planktonic growth [37]. This is because the embedded bacterial cells are getting an optimal defense mechanism against the adverse effects of antibiotics and the immune system of the host [37]. Thus in our opinion, the evidentially lower activity against biofilm should not be considered as a crucial disadvantage of the synthesized compounds.

Influence of the Selected Compounds on Metabolic Activity
When a semi-quantitative API ZYM micro-method was applied for the determination of influence of the selected compounds on activity of 19 hydrolases of S. aureus, it appeared that the S. aureus ATCC 29,213 reference strain exhibited activity of 5 enzymes, namely alkaline phosphatase, esterase (C4), esterase lipase (C8), acid phosphatase and naphthol-AS-BI-phosphohydrolase. Analysis of the enzymatic activity of bacterial cells treated with the selected agents at ½ × MIC, results of which are shown in Table 4, revealed a statistically significant decrease in the activity of alkaline phosphatase, esterase (C4) and esterase lipase (C8). The activity of two other enzymes was not affected. In order to explain whether the observed reduction of activity of some enzymes could be a consequence of the killing of bacterial cells, the determination of CFU number in treated and control samples was performed. No difference was observed (no killing effect in the presence of the agents at a concentration of ½ MIC during 1 h incubation was observed). It clearly indicates that treatment of the cells with the selected substances (at a concentration lower than MIC) importantly affected the activity of some of the important enzymes, however without any lethal effect.

Activity of the Selected Compounds against Staphylococcal Biofilm
When the S. aureus cells were grown in the biofilm mode, the antistaphylococcal potential of compounds 32-34 and 36 appeared much lower, if any, than that found against the cells grown in the planktonic mode. None of these compounds caused any decrease in the metabolic activity of the cells forming the biofilm, as measured with the MTT assay, at concentrations as high as 64 µg/mL (128 or 256 × MIC). It is not especially surprising, since many earlier reports indicated that for different antibacterial agents, their effective concentrations inhibiting metabolic activity of staphylococci grown in the biofilm mode can be 10-1000 times higher than those inhibiting the planktonic growth [37]. This is because the embedded bacterial cells are getting an optimal defense mechanism against the adverse effects of antibiotics and the immune system of the host [37]. Thus in our opinion, the evidentially lower activity against biofilm should not be considered as a crucial disadvantage of the synthesized compounds.

Influence of the Selected Compounds on Metabolic Activity
When a semi-quantitative API ZYM micro-method was applied for the determination of influence of the selected compounds on activity of 19 hydrolases of S. aureus, it appeared that the S. aureus ATCC 29,213 reference strain exhibited activity of 5 enzymes, namely alkaline phosphatase, esterase (C4), esterase lipase (C8), acid phosphatase and naphthol-AS-BI-phosphohydrolase. Analysis of the enzymatic activity of bacterial cells treated with the selected agents at 1 2 × MIC, results of which are shown in Table 4, revealed a statistically significant decrease in the activity of alkaline phosphatase, esterase (C4) and esterase lipase (C8). The activity of two other enzymes was not affected. In order to explain whether the observed reduction of activity of some enzymes could be a consequence of the killing of bacterial cells, the determination of CFU number in treated and control samples was performed. No difference was observed (no killing effect in the presence of the agents at a concentration of 1 2 MIC during 1 h incubation was observed). It clearly indicates that treatment of the cells with the selected substances (at a concentration lower than MIC) importantly affected the activity of some of the important enzymes, however without any lethal effect. Therefore, the outcomes of our biological investigation revealed the very high antistaphylococcal activity against S. aureus ATCC 25923, S. aureus ATCC 29,213 and S. epidermidis ATCC12228 of some synthesized anthra [1,2-d] [1,2,3]triazine-4,7,12(3H)-triones at the concentration of 0.125-0.5 µg/mL, while randomly selected antibiotics have MIC in the range of 0.125-8.0 µg/mL. The high antistaphylococcal potential of the most active compounds, namely 32-34 and 36, has been also confirmed against clinical isolates of S. aureus, including the MRSA strains. The obtained results create the prospect of their further in-depth studies as effective antibacterial agents.

General Information
Melting points were measured in open to air glass capillaries using a B-540 melting point apparatus (Büchi, New Castle, DE, USA) and are uncorrected. Elemental analysis was performed on a 2400 CHN-analyzer (PerkinElmer, Wilmington, DE, USA) and the results were found to be in good agreement with the calculated values. 1 H-NMR spectra (400 MHz) and 13 C-NMR spectra (100 MHz) were recorded in DMSO-d 6 (unless otherwise indicated) on a Mercury-400 spectrometer (Varian, Palo. Alto, CA, USA) with TMS as an internal standard. Mass spectra were recorded on a 1100 Series G1956B LC/MSD SL LCMS system (Agilent, Santa Clara, CA, USA) using electrospray ionization at atmospheric pressure (70 eV). X-ray structural analysis of a single crystal of compound 22 was performed at 173 K on a Smart Apex II diffractometer (Bruker, Madison, WI, USA) operating in the ω scans mode. The intensity data were collected within the range of 2.19 ≤ θ ≤ 26.72 • using Mo-K α radiation (λ = 0.71078 Å). The intensities of 11,753 reflections were collected (2580 unique reflections, R merg = 0.0526). The structure was solved by direct methods and refined by the full-matrix least-squares technique in the anisotropic approximation for non-hydrogen atoms using the Bruker SHELXTL program package [38]. All chemicals were of reagent grade and used without further purification. The solvents were purified according to the standard procedures [39]. 1-Amino-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid (1) and 1-amino-4-bromo-9,10-dioxo-9,10-dihydroanthracene-2-carboxylic acid (2) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Investigation of Antimicrobial Potential of Synthesized Agents
The two-fold broth microdilution method, according to the CLSI standard methodology [42], was applied for the determination of minimum inhibitory concentrations (MICs) of synthesized agents and antibiotics against bacteria (both reference strains and clinical isolates). Bacterial strains were plated on Luria-Bertani Agar medium and incubated overnight at 37 • C. Two to three bacterial colonies were transferred from the agar medium into PBS buffer (pH = 7.4). The bacterial suspension was adjusted to the optical density of OD 600 = 0.1 and diluted in MHB2 medium at a ratio of 1:100 v/v to the final cell concentration of approximately 0.5-1.0 × 10 6 CFU/mL. The synthesized anthra [1,2-d] [1,2,3]triazine-4,7,12(3H)-triones were dissolved in DMSO-final concentration of all substances tested in the stock solutions was 1280 µg/mL. The stock solutions with the same concentration (1280 µg/mL) of all antibiotics were used for investigation. Subsequently, a series of two-fold dilution of each agent in range 64.0-0.0625 µg/mL were prepared in the wells of columns 1-10 of 96-well plates using Mueller-Hinton Broth 2. An aliquot of 100 µL of inoculum was dispensed to the wells of columns 1-11. Column 11 contained 200 µL of inoculum without any agent, and column 12 contained 200 µL of the MHB2 broth only (served as a control of medium sterility). The plates were incubated 24 h under static conditions at 37 • C. The lowest concentration of agent with no visible bacterial growth was taken as a MIC value.
The determination of MIC values for yeast strains was also performed according to the appropriate NCCLS reference microdilution method [43]. The solutions of substances in DMSO were prepared in the same way as for the investigation of antibacterial activity (presented above). Serial two-fold dilutions of the tested substances were prepared in RPMI 1640 medium buffered to pH 7.0 with MOPS buffer (3-N-morpholinopropanesulfonic acid) in 96-well microtiter plates in a final volume of 100 µL. The final concentrations of the agents were in the range from 64 to 0.0625. Suspensions of the microorganisms were prepared by taking one loop of pure culture (cultivated 24 h at 37 • C on YPD agar) into sterile water and adjusting the optical density to 0.1 at 660 nm wavelength before further 50-fold dilution in an RPMI 1640 medium resulting in 2 × 10 4 CFU/mL. One hundred microliters of such suspension were inoculated to each well of the microtiter plate, leaving a drug-free column (number 12) as sterility controls. Plates were incubated for 24 h at 37 • C. MIC values were read visually as the first concentration where no growth was observed. The MIC assay for each tested strain of bacteria/yeast and agent was performed in triplicate. In the case of both assays (investigation of antibacterial and antifungal activity) the final concentration of the solvent (DMSO) in the medium did not exceed 2.5% (v/v), and did not influence the growth of bacteria/yeast.

Time-Kill Assay
Taking into account the results of previously performed analysis (high activity against staphylococci and lack of activity against Gram-negative bacteria and yeasts) the kinetic time-kill assay was performed for one reference strain of staphylococci (namely S. aureus ATCC 29213) and four, most promising, compounds (namely 32, 33, 34 and 36). The suspension of approx. cell density 0.5-1.0 × 10 6 CFU/mL of S. aureus ATCC 29,213 was prepared in MHB2 broth supplemented with a particular agent to the final concentrations equal to MIC, 2 × MIC or 4 × MIC and incubated at 37 • C with shaking (150 rpm). Bacterial suspension without agent addition was used as the untreated control. At predetermined time intervals (0, 1, 3, 6 and 24 h) samples were taken, serially diluted in PBS buffer (from 10 −1 to 10 −7 ) and spotted (10 µL) onto a Baird-Parker agar plate. After 24 h of incubation, the plates were enumerated. For each agent the assay was repeated in triplicate.

Biofilm Formation and Determination of Antibiofilm Activity
The assay for biofilm cultivation was performed according to the procedure described previously [44] with slight modifications. The suspensions of approx. cell density 1-5 × 10 8 CFU/mL of S. aureus ATCC 29,213 were diluted 1:100 (v/v) in TSB medium supplemented with 2.5% glucose. Cell suspensions (200 µL) were placed into the wells of columns 1-7 of vertically set plates. The wells of column 8 were fulfilled with 200 µL of sterile medium-these wells were used as a negative control. The plates were incubated for 24 h at 37 • C without shaking in order to allow bacteria to attach. In the next step, the formed biofilm was treated with four selected, most promising agents (32, 33, 34 and 36). The liquid medium was removed from the wells of the plates and the formed biofilm was gently washed with 200 µL of sterile PBS. Subsequently 200 µL of 2-fold serial dilutions of the agents in MHB2 medium, ranging from 64 to 2 µg/mL, were added to the wells of columns 1-6 (column 7-without agent but with formed biofilm served as positive control and wells of column 8-without biofilm, were used as negative control-control of sterility of the medium) and incubated for 24 h at 37 • C. The MTT assay was performed as described previously [44,45]. After 24-h treatment, the liquid content of the wells of the plates was removed and the treated biofilm was washed with 200 µL of sterile PBS buffer. Subsequently, 150 µL of PBS and 50 µL of MTT solution (0.3% in PBS) were added to the wells and mixed. Following 2 h incubation at 37 • C in the dark, the MTT solution was replaced with 200 µL of DMSO for dissolving the formed formazan crystals. The optical density of the obtained solutions was measured at 540 nm using a Victor 3 microtiter reader (Perkin Elmer, Waltham, MA, USA). The Minimal Biofilm Eradiation Concentration (MBEC 50 ) values were taken as the lowest concentration of the agent that caused eradication of at least 50% of living cells in comparison to the cells growing in the untreated control-measured as comparison of ability of living cells to the biotransformation of MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide) to insoluble in water violet formazan crystals [45].