New marine-derived indolymethyl pyrazinoquinazoline alkaloids with promising antimicrobial profiles

Due to the emergence of multidrug-resistant pathogenic microorganisms, the search for novel antimicrobials is urgent. Inspired by marine alkaloids, a series of indolomethyl pyrazino [1,2-b]quinazoline-3,6-diones was prepared using a one-pot microwave-assisted multicomponent polycondensation of amino acids. The compounds were evaluated for their antimicrobial activity against a panel of nine bacterial strains and five fungal strains. Compounds 26 and 27 were the most effective against Staphylococcus aureus ATCC 29213 reference strain with MIC values of 4 μg mL−1, and a methicillin-resistant Staphylococcus aureus (MRSA) isolate with MIC values of 8 μg mL−1. It was possible to infer that enantiomer (−)-26 was responsible for the antibacterial activity (MIC 4 μg mL−1) while (+)-26 had no activity. Furthermore, compound (−)-26 was able to impair S. aureus biofilm production and no significant cytotoxicity towards differentiated and non-differentiated SH-SY5Y cells was observed. Compounds 26, 28, and 29 showed a weak antifungal activity against Trichophyton rubrum clinical isolate with MIC 128 μg mL−1 and presented a synergistic effect with fluconazole.


Introduction
Infectious diseases caused by microorganisms stand as a major threat to public health. 1 Since antibiotics were rst introduced as medicines, these drugs have been used to prevent or treat infections in several applications. 2,3 Nonetheless, antibacterial resistance has increased dramatically, becoming an emergency in healthcare during the last 40 years. 4-6 Among 50 emerging infectious agents that have been identied, 10% have developed resistance to multiple drugs including antibiotics such as vancomycin, 7,8 methicillin, 9 carbapenems, 10 and cephalosporins. [11][12][13] Despite enormous efforts, the number of therapeutically useful compounds aiming to circumvent resistance is continuously decreasing and no truly novel class of compounds has been introduced into therapy, causing the World to face the "post-antibiotic era". 14,15 In order to restrain the clinical consequences of the development and spread of antimicrobial resistance both the preservation of current antimicrobials through their appropriate use and the discovery and development of new agents are mandatory. 16 Promising antimicrobial agents for overcoming multidrug resistance are emerging from a variety of sources and methodologies, [17][18][19][20][21][22][23] with the very recent example of articial intelligence aiding this discovery. 24 Several reports emphasized the discovery of new sophisticated antimicrobials from marine sources as a promising strategy to overcome the ever-increasing drug-resistant infectious diseases. [25][26][27][28][29] Particularly, fungal alkaloids containing an indolomethyl pyrazino [1,2-b]quinazoline-3,6-dione scaffold were isolated from marine organisms and presented very interesting antimicrobial activities. 30 For instance, glyantrypine (1, Fig. 1), isolated from Cladosporium sp. PJX-41, exhibited moderate inhibitory activity against a bacterium Vibrio harveyi (MIC ¼ 32 mg mL À1 ) 31 and neoscalin A (2), obtained fromNeosartorya siamensis KUFC 6349, exhibited a potent antibacterial activity against Staphylococcus aureus and Enterococcus faecalis (MIC ¼ 8 mg mL À1 ). 32 Regarding antifungal activity, fumiquinazoline F (3) for instance, obtained from a culture of Aspergillus fumigatus LN-4 showed good activity (MIC ¼ 12.5-50 mg mL À1 ) against several plant pathogenic fungi, 33 and cottoquinazoline D (4) obtained from Aspergillus versicolor LCJ-5-4, showed moderate antifungal activity against Candida albicans (MIC ¼ 22.6 mM). 34 Inspired by the relevant antibacterial activity of these quinazolinedione natural products 30 and the fact that these pure alkaloids are obtained from the fugal culture in low yields, 35 we have embarked on a synthesis of pyrazino [1,2-b]quinazoline-3,6-dione derivatives with a simpler indolomethyl moiety decreasing the number of stereogenic centers. Our rst approach (5-12, Fig. 1) involved the synthesis of enantiomeric pairs of two members of this quinazolinone family (structural modications at C-1 and C-4 stereochemistry), inspired by the marine-derived alkaloid scalin B (7). 36 The second approach (13-21, Fig. 1) consisted in the synthesis of other derivatives of these natural alkaloids, but this time with modication of the C-1 side chain and the stereochemistry, by using different amino acids. 37 Inuenced by a large number of halogenated marine natural products with interesting antimicrobial activities isolated over the last few years, 29,38 in the present work (22-32, Fig. 1) we present a third approach through the introduction of halogen atoms in the aromatic ring of anthranilic acid (Ant) which led to the discovery of promising antimicrobial agents within this series. The potential mechanism of action was studied using in silico docking on three potential antibacterial targets of antimicrobial alkaloids (S. aureus DNA gyrase B (GyrB), and S. aureus FtsZ), followed by molecular dynamics.

Chemistry
The eleven new indolomethyl pyrazino [1,2-b]quinazoline-3,6dione derivatives 22-32 were synthesized by a previously described approach using a microwave-assisted multicomponent polycondensation of amino acids with modications (Table 1). 36,37,39 The coupling of halogenated commercial anthranilic acids (33) to N-protected L-a-amino acids (34), and further dehydrative cyclization using triphenyl phosphite [(PhO) 3 P], generated a kind of benzoxazin-4-ones intermediates 35 which, followed by the addition of D-tryptophan methyl ester (36) under microwave irradiation, furnished the desirable nal products 22-32 (2-14% yield) with partial epimerization (Table  1). Starting from halogenated anthranilic acid 33, the reaction conditions had to be optimized to obtain the intermediates 35: the amount of (PhO) 3 P was increased from 1.2 eq. to 2.4 eq. and the reaction time was increased from 16 to 24 h. With these conditions we were able to increase the scope of this one-pot reaction to halogenated compounds. Interestingly, only anti isomers (1S,4R) were obtained by this methodology whereas the different side chains at C-1 were achieved by selecting diverse La amino acids-valine, leucine, and isoleucine. The purity of the compounds was determined, by a reversed-phase liquid chromatography (RP-LC, C18, MeOH : H 2 O; 50 : 50), to be on average higher than 95%, whereas that for 25 and 32 was 90%.
Structure elucidation of compounds 22-32 was accomplished by 1D and 2D NMR (using CDCl 3 or DMSO-d 6 as solvents) spectral analysis and conrmed by HRMS. Particularly, the 1 H NMR spectra of the monohalogenated derivatives, 22-24, and 28-31 whose halogen atoms (Cl, Br, or I) are at position 9, exhibited a signal corresponding to the resonance of H-8 in the form of a doublet at d H ca. 8.5 with a small coupling constant ( 4 J ¼ 2.4 Hz), evidencing a long range coupling with H-10; while the signal of H-10 appeared as double doublets at d ca. 7.9 ( 3 J ¼ 8.7 Hz and 4 J ¼ 2.4 Hz). In addition, the signal corresponding to the resonance of H-11 appeared as doublet at d ca. 7.5 ( 3 J ¼ 8.7 Hz), indicating a short-range coupling to H-10. Similarly, the 1 H NMR spectra of the dihalogenated derivatives, i.e. 25-27 and 32, with two halogen atoms (Cl or I) at positions 9 and 11, also appeared as a doublet at d ca. 7.76 ( 4 J ¼ 2.4 Hz) corroborating a long-range coupling between H-8 and H-10 (See in Material and methods).
It was observed also that the signals of the NH of the indole moiety (H-7 0 ) and the amide group (H-2) appeared at higher frequencies when DMSO-d 6 was used as a solvent. For example, for 25, 26, and 32, the signals H-2 and H-7 0 appeared at d H 10.20 and 7.11 ppm, respectively, and this was due to the establishment of H-bonding (see the Experimental data).

Microbiology
Antibacterial susceptibility testing. An initial screening of the antibacterial activity of 5-32 against different reference strains of Gram-positive, Gram-negative bacteria, and clinically relevant multidrug-resistant (MDR) strains was performed by the disk diffusion method. 40,41 This primary assessment was followed by the determination of minimal inhibitory concentrations (MIC) of reference strains. For active compounds, this determination was also made for MDR strains. In the range of concentrations tested, none of the compounds was active against Gram-negative bacteria, and none of 5-21 ( Fig. 1), 28, 29, and 32 was active against any of the tested strains (see Tables S3 and S4 of ESI † for details on remaining strains). The results of the antibacterial activity on Gram-positive strains regarding all other compounds are presented in Table 2. None of the halogenated derivatives exhibited antibacterial activity,  Synergistic effects with vancomycin and oxacillin were evaluated for MDR strains, but no effect was observed. These antibiotics are relevant in the treatment of infections caused by Enterococcus spp. and Staphylococcus aureus, respectively. The compounds showed activity only for Gram-positive strains and, overall, this activity was greater for reference strains than for clinically relevant strains, whether MDR or not. Regarding Gram-positive strains, the range was not equal for all compounds, with a greater number of compounds being active against S. aureus than E. faecalis. As for E. faecalis there appeared to exist an inverse relationship between the compound activity and a resistance against clinically important antibiotics; however, there was not a clear tendency for S. aureus. Noteworthy, the rst series of compounds showed no relevant effect in the growth of non-malignant cells.
The ability to prevent biolm formation was evaluated for compounds with antibacterial activity. Concentrations ranging from 2 Â MIC to 1 4 MIC were tested against S. aureus ATCC 29213, S. aureus 66/1 and E. faecalis ATCC 29212. The highest concentration tested was 64 mg mL À1 , in order to keep a nal intest concentration of DMSO below 1%. The results were interpreted using a comparative classication that classies adherence capabilities of tested strains into four categories: (i) nonadherent, (ii) weakly adherent, (iii) moderately adherent and, (iv) strongly adherent. 42 Optical density cut-off values (ODc) for each microtiter plate were dened as three standard deviations above the mean OD of the negative control. The classication criteria are summarized in Table 3. This classication which uses the negative control as a starting point, instead of using the positive control as reference, reduces the risk of inconsistencies due to external factors that inuence biolm production. 43 The compounds tested did not inhibit biolm formation of E. faecalis ATCC 29212 (data not shown). S. aureus ATCC 29213 was classied as strong biolm producer, and (À)-26 (2 Â MIC, 8 mg mL À1 ) was able to impair this ability (Table 4), while 22-27, 30 and 31 were not able to prevent the formation of a strong bio-lm (data not shown). The biolm forming ability of S. aureus 66/1, which was also classied as a strong biolm producer, was impaired by 26 (MIC, 8 mg mL À1 and 2 Â MIC, 16 mg mL À1 ), (À)-26 (2 Â MIC, 8 mg mL À1 ) and 27 (2 Â MIC, 16 mg mL À1 ) ( Table 5).
Time-kill kinetics. Even though the quinazolinones with antimicrobial activity described herein have bacteriostatic activity, given a potent activity of 26 against S. aureus strains, the time-kill kinetics of this compound was evaluated in order to ascertain its effect over a 24 h incubation. In the presence of 64 mg mL À1 of 26 (the highest concentration within the 1% DMSO threshold), it was possible to observe a rise in bacterial growth aer 12 h of incubation whereas, for the control, the growth was evident from the start of the assay. In addition, at the nal time point there was a 10 2 CFU mL À1 difference between the two conditions, as observed in Fig. 2.
Scanning electron microscopy (SEM). In order to evaluate a possible effect of 26 on S. aureus cell surface and morphology as well as cellular arrangement, SEM analysis was performed for S. aureus ATCC 29213 and S. aureus 66/1 (Fig. 3). The surfaces of  Classication Criteria Non-adherent OD less than or equal to ODc Weakly adherent OD more than ODc and less than two-fold ODc Moderately adherent OD more than two-fold ODc and less than four-fold ODc Strongly adherent OD more than four-fold ODc a OD, optical density; ODc, optical density cut-off value.
cells treated with 26 at MIC and 1 2 MIC (Fig. 3B, C, E and F) were relatively smooth and regular, similar to those of the untreated control ( Fig. 3A and D). No shrinkage compatible with permeabilization of the membrane and leakage of cytoplasmic material was observed. However, the irregular clusters predominantly observed in the untreated control, an arrangement known to occur in S. aureus, 44 were not present in treated samples, being replaced by short chains, pairs and single cells, with a more uniform size. This was also observed by Gram staining (data not shown).
Antifungal activity. The antifungal activity of the test compounds was evaluated against Candida albicans, Aspergillus fumigatus and Trichophyton rubrum by determining MICs. 45,46 None of the compounds tested showed activity against C. albicans and A. fumigatus strains. Nevertheless, 26, 28, and 29 exhibited a weak inhibitory effect on a dermatophyte strain (T. rubrum FF5) with a MIC value of 128 mg mL À1 and a minimal fungicidal concentration (MFC) higher than 128 mg mL À1 , suggesting that this compound has a fungistatic activity (for details, see Table S5, ESI †). Given these results, 26, 28, and 29 were tested against two additional dermatophyte strains, i.e. Microsporum canis FF1 and Epidermophyton occosum FF9; however, no activity was observed. Compounds 26, 28, and 29 were also evaluated for synergistic effects for T. rubrum. A synergistic effect was observed for 28 and 29 with uconazole. Fractional inhibitory concentrations index (FICI), determined by checkerboard method, was < to 0.5 (0.06 for 28 and 0.13 for 29).
In order to evaluate in vitro enantioselectivity activities, including antibacterial and antifungal activities, the most promising derivatives 22, 23, and 26 were obtained in milligram scale by a semi-preparative enantioselective liquid chromatography, employing a tris-3,5-dimethylphenylcarbamate amylose column with multiple injection in a 200 mL loop (see ESI † for details).
The pure enantiomers of 22, 23, and 26 were evaluated for antibacterial and antifungal activities. The enantiomer (À)-26 showed a MIC value of 4 mg mL À1 for reference strain S. aureus ATCC 29213, sensitive clinical isolate S. aureus 40/61/24, and methicillin-resistant strain S. aureus 66/1, while the enantiomer (+)-26 showed no effect (Table 2). Noteworthy, these derivatives showed higher potency than the natural product neoscalin A (2), (tested by the same group with the same conditions). 32,47,48 None of the pure enantiomers was active against the fungi tested.
Structure-activity relationship (SAR) study. Regarding antibacterial activity, SAR (Fig. 4) suggested that the presence of a halogen atom at positions C-9 and/or C-11 plays a crucial role for this activity since all the non-halogenated derivatives were inactive against all the tested strains (for details see Tables S3 and S4, ESI †). In fact, compounds containing chlorine at one or both positions exhibited better antibacterial activity compared to those having bromine and iodine. Higher antibacterial activity was obtained when the halogen atom is present at C-9 and C-11 positions (i.e., 25, 26, 27 and 32) and/or the presence of longer side chains at C-1. Compounds 26, 28, and 29 were also evaluated for their synergistic effects on T. rubrum. A synergistic effect was observed for 28 and 29 with uconazole (data not shown). The enantiopure (À)-26 showed signicant antibacterial effect against a resistant strain of S. aureus whereas its antipode [(+)-26] did not. This result emphasizes that the conguration (1S,4R) is crucial for the antibacterial activity of quinazolinone scaffold.
To understand the structural basis of the potential ligand binding to GyrB, and Fts, each prepared ligand was individually docked with the enzymes at their active site using Auto-dockVina. Test molecules present free binding energies on GyrB ranked from À9.3 to À9.6 kcal.mol À1 (Table 6), and all of the test compounds are predicted to establish more stable complexes with GyrB (lower docking score than test compound 08B; À7.8 kcal.mol À1 ). All the test compounds are predicted to have lower affinity to FtsZ than test compound 9PC (docking scores higher than À10.1 kcal.mol À1 , respectively). Therefore, as GyrB has already been reported as a target for other quinazoline derivatives 49,50 it is hypothesized as a possible target for the tested antimicrobial pyrazinoquinazolines. However, further target-specic testing will be performed in the near future to counterproof the in silico ndings. s Spatial conformation of test compounds and key interactions of 26 in the active site of GyrB are presented in Fig. 5.
The pyrazinoquinazoline moiety binds to the ATP binding site of GyrB, participating in hydrogen interactions with Ser-55, Glu-58, Gly-85, Ile-86 and Thr-173, and halogen interaction with Asp-57whereas the indole group is involved in hydrogen interactions with Asn-54 and Asp-81 (Fig. 5B). Alternatively, another docking pose suggests hydrogen interactions with Glu-58, Arg-84, and Gly-85, as well as halogen interactions with Asn-54 and Ser-55 (Fig. 5C). Some of these residues have already been described as being involved in the binding of known inhibitors to GyrB 51-54 In both docking poses ( Fig. 5B and C), the halogen atoms are predicted as being involved in the interaction with GyrB. In fact, currently, the halogenation of compounds has become an important strategy in drug design. 55 Halogens are found to increase membrane permeability and the t 1/2 by lowering metabolic degradation, and enhance binding affinity to targets. 56,57 Halogens, especially chlorine and bromine, are present in a signicant number of drugs, contributing favourably to ligand-protein interaction. 58 Halogen bonding is a non-covalent interaction similar to the hydrogen bond (halogens act as hydrogen bond acceptors), and characterized by its directionality. 59 There are several physico-chemical properties of chloride that may justify its more favourable contribution to activity when compared to bromide and iodide. In fact, chloride is more electronegative and it is a smaller atom, forming halogen bonds with specic lengths and angles, which contribute to a more appropriate t to the target stereoelectronic prole. Furthermore, the presence of halogens has already been described as being responsible for the increase in the affinity of known inhibitors to the GyrB ATP binding site. 60,61 However, molecular docking using a rigid target may be insufficient for the determination of the structure and the stability of the ligand:target complex. To gain a detailed insight in the energetic and geometric behaviour of the 26:GyrB complex in aqueous environment, a 2 ns MD simulation was performed based on two most stable complex structures obtained from the docking study considering the effects of the target exibility and the explicit water solvation.    )). Polar interactions are depicted as yellow broken lines (residues evolved are labelled); GyrB is represented as transparent surface, with carbons, oxygens, and nitrogens coloured grey, red, and blue, respectively. Hydrogen atoms are omitted for simplification. Potential energy plot of the complexes during MD simulation is represented below each MD image (U ¼ potential energy; t ¼ time). interactions are established to Ile-51 and to Ser-129 via a water molecule; and hydrogen bonds are formed between 26 and Asn-54, Glu-58, and Arg-84. 62 In conclusion, the halogenation of pyrazinoquinazolines was a successful approach in drug design of new antimicrobial compounds, potentially increasing the binding affinity and binding selectivity to GyrB.
Cytotoxicity of compound 26. The cytotoxicity of 26 (0-25 mM) was evaluated in differentiated and non-differentiated SH-SY5Y cells, by the neutral red (NR) uptake, resazurin (REZ) reduction and sulforhodamine B (SRB) binding assays, 24 h aer exposure. As observed in Fig. 7, no signicant cytotoxicity towards both differentiated and non-differentiated SH-SY5Y cells was observed, as evaluated by the results obtained for the three adopted in vitro cytotoxicity assays.

Conclusion
Among a new series of indolomethyl pyrazino [1,2-b] quinazoline-3,6-diones, 26 and 27 exhibited a potent antibacterial activity against S. aureus strains, with MIC values of 4 mg mL À1 for a reference strain and 8 mg mL À1 for a methicillinresistant strain (S. aureus 66/1). Comparing with the marine natural product neoscalin A (2), a two-fold reduction in the MIC values was observed. Regarding the structural complexity and synthetic pathways, the excellent and inspiring results obtained in the present study show that simpler molecules than neoscalin A (2) are quite promising to nd new agents to overcome MDR strains. In silico docking and molecular dynamics revealed that halogenation was a successful design strategy to discover new indolymethyl pyrazinoquinazoline with antimicrobial activity. Although the investigated biological activities of scalins are still in preliminary stages, there is no doubt that these are privileged structures with potential as drug candidates. Nevertheless, mainly due to the limited amounts isolated, most of the studies correspond to phenotypic screening assays and little is known about their molecular targets and pharmacokinetic properties. One limitation of the one-pot synthesis is related to the low yields obtained, which can be a drawback on the scale up of these compounds. The main challenge in synthesis is the intricate stereochemistry found in the more complex structures. Although the threecomponent one-pot assembly of this scaffold is a very efficient procedure to obtain some of these secondary metabolites, the high temperatures involved resulted in partial epimerization. Differentiation between stereoisomers was noted in their antimicrobial action but also occurs in drug disposition and is of particular signicance for those processes that depend upon a direct interaction between the drug and a chiral biological macromolecule, e.g., active transport processes, binding to plasma and tissue proteins, and drug metabolism, deserving this issue special attention in future work.

Microbiology
Microbial strains and growth conditions. In the present study, two Gram-positive -Staphylococcus aureus ATCC 29213 and Enterococcus faecalis ATCC 29212-and two Gram-negative -Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853reference bacterial strains were used. When it was possible to determine a minimal inhibitory concentration (MIC) value for these strains, clinically relevant strains were also used. These included methicillin-resistant S. aureus (MRSA) 66/1, isolated from public buses, 64 and a vancomycin-resistant Enterococcus (VRE) strain isolated from river water, 65 E. faecalis B3/101. Frozen stocks of all strains were grown on Mueller-Hinton agar (MH -BioKar Diagnostics, Allone, France) at 37 C for 24 h. All bacterial strains were sub-cultured on MH agar and incubated overnight at 37 C before each assay, in order to obtain fresh cultures. For the antifungal activity screening, a yeast reference strain Candida albicans ATCC 10231, a lamentous fungi reference strain Aspergillus fumigatus ATCC 46645, and a dermatophyte clinical strain Trichophyton rubrum FF5 were used. For compounds showing some activity in the dermatophyte T. rubrum, the activity was enlarged to other species of dermatophytes (Microsporum canis FF1 and Epidermophyton occosum FF9). Frozen stocks of all fungal strains were sub-cultured in Sabouraud Dextrose Agar (SDA -Bio-Mèrieux, Marcy L'Etoile, France) before each test, to ensure optimal growth conditions and purity. Stock solutions of each compound (10 mg mL À1 ) were prepared in dimethylsulfoxide (DMSO -Alfa Aesar, Kandel, Germany). In the experiments, the nal in-test concentration of DMSO was kept below 1%, as recommended by the CLSI. 66 Antimicrobial susceptibility testing Antibacterial activity. An initial screening of the antibacterial activity of the compounds was performed by the Kirby-Bauer disk diffusion method, as recommended by the Clinical and Laboratory Standards Institute (CLSI). 67 Briey, sterile 6 mm blank paper disks (Oxoid, Basingstoke, England) impregnated with 15 mg of each compound were placed on inoculated MH agar plates. A blank disk with DMSO was used as a negative control. MH inoculated plates were incubated for 18-20 h at 37 C. At the end of incubation, the inhibition halos were measured. The minimal inhibitory concentration (MIC) was used to determine the antibacterial activity of each compound, in accordance with the recommendations of the CLSI. 68 Twofold serial dilutions of the compounds were prepared in Mueller-Hinton Broth 2 (MHB2 -Sigma-Aldrich, St. Louis, MO, USA) within the concentration range of 0.062-64 mg mL À1 . Cefotaxime (CTX) ranging between 0.031-16 mg mL À1 was used as a control. Sterility and growth controls were included in each assay. Purity check and colony counts of the inoculum suspensions were also performed in order to ensure that the nal inoculum density closely approximates the intended number (5 Â 10 5 CFU mL À1 ). The MIC was determined as the lowest concentration at which no visible growth was observed. The minimal bactericidal concentration (MBC) was assessed by spreading 10 mL of culture collected from wells showing no visible growth on MH agar plates. The MBC was determined as the lowest concentration at which no colonies grew aer 16-18 h incubation at 37 C. These assays were performed in duplicate.
Antifungal activity. The MIC of each compound was determined by the broth microdilution method, according to CLSI guidelines (reference documents M27-A3 (ref. 45) for yeasts and M38-A2 (ref. 46) for lamentous fungi). Briey, cell or spore suspensions were prepared in RPMI-1640 broth medium supplemented with MOPS (Sigma-Aldrich, St. Louis, MO, USA) from fresh cultures of the different strains of fungi. In the case of lamentous fungi, the inoculum was adjusted to 0.4-5 Â 10 4 CFU mL À1 for A. fumigatus ATCC 46645 and to 1-3 Â 10 3 CFU mL À1 for the dermatophytes. The inoculum of C. albicans was adjusted to 0.5-2.5 Â 10 3 CFU mL À1 . Two-fold serial dilutions of the compounds were prepared in RPMI-1640 broth medium supplemented with MOPS within the concentration range of 1-128 mg mL À1 , with maximum DMSO concentration not exceeding 2.5% (v/v). Sterility and growth controls were also included in each assay. The plates were incubated for 48 h at 35 C (C. albicans and A. fumigatus) or during 5 days at 25 C (T. rubrum, M. canis and E. occosum). MICs were recorded as the lowest concentrations resulting in 100% growth inhibition in comparison to the compound-free controls. Voriconazole MIC for Candida krusei ATCC 6258 was used as quality control. 45,46,69,70 The assay was validated when results obtained were within the recommended limits. The minimal fungicidal concentration (MFC) was determined by spreading 20 mL of culture collected from wells showing no visible growth on SDA plates. The MFC was determined as the lowest concentration showing 100% growth inhibition aer 48 h at 35 C (for C. albicans and A. fumigatus) or 5 days incubation at 25 C (T. rubrum, M. canis and E. occosum). All the experiments were repeated independently at least two times.

Biolm formation inhibition assay
For compounds with antibacterial activity, their effect on bio-lm formation was evaluated using the crystal violet method. Briey, bacterial suspensions of 1 Â 10 6 CFU mL À1 were prepared in Tryptone Soy broth (TSB-Biokar Diagnostics, Allone, Beauvais, France) supplemented with 1% (p/v) glucose (D(+)-glucose anhydrous for molecular biology, PanReac Appli-Chem, Barcelona, Spain); and four concentrations of compound were tested: 2 Â MIC, MIC, 1 2 MIC and 1 4 MIC, keeping nal intest concentration of DMSO below 1%. A control with inoculum and culture media, a control with appropriate concentration of DMSO, as well as a negative control (TSB alone) were included. Sterile 96-well at-bottomed untreated polystyrene microtiter plates were used. Aer a 24 h incubation at 37 C, the biolms were heat-xed for 1 h at 60 C and stained with 0.5% (v/v) crystal violet (Química Clínica Aplicada, Amposta, Spain) for 5 min. The stain was resolubilized with 33% (v/v) acetic acid (acetic acid 100%, AppliChem, Darmstadt, Germany) and the biolm biomass was quantied by measuring the absorbance of each sample at 570 nm in a microplate reader (Thermo Scien-tic Multiskan® EX, Thermo Fisher Scientic, Waltham, MA, USA). 40,43 Three independent experiments were performed, in triplicate.

Antimicrobial synergy testing
Antibiotic synergy. In order to evaluate the combined effect of the compounds and clinically relevant antimicrobial drugs, a screening was conducted using the disk diffusion method, as previously described. 40 A set of antibiotic disks (Oxoid, Basingstoke, England) to which the isolates were resistant was selected: cefotaxime (CTX, 30 mg) for extended spectrum betalactamase-producer E. coli SA/2, oxacillin (OX, 1 mg) for S. aureus 66/1, and vancomycin (VA, 30 mg) for E. faecalis B3/101. Antibiotic disks alone (controls) and antibiotic disks impregnated with 15 mg of each compound were placed on MH agar plates seeded with the respective bacteria. Sterile 6 mm blank papers impregnated with 15 mg of each compound alone were also tested. A blank disk with DMSO was used as a negative control. MH inoculated plates were incubated for 18-20 h at 37 C. Potential synergism was recorded when the halo of an antibiotic disk impregnated with a compound was greater than the halo of the antibiotic or compound-impregnated blank disk alone.
Antifungal synergy. In order to evaluate the combined effect of the compounds and clinically relevant antifungal drugs, checkerboard assay was conducted, as previously described. 29 Fluconazole was used in a range between (0.062-4 mg mL À1 ) and compounds were tested in a range between their MIC and progressive two-fold dilutions.

Time-kill kinetics assay
Time-kill kinetics of 26 was performed for S. aureus ATCC 29213 and MRSA S. aureus 66/1, according to CLSI guidelines. 72 Briey, a solution of 64 mg mL À1 of 26 was prepared in MHB2, as well as a control with appropriate concentration of DMSO; bacterial suspensions of 5 Â 10 5 CFU mL À1 in MHB2 were added and incubated at 37 C in an orbital shaker. Aliquots of 100 mL were taken at time intervals of 0, 2, 4, 8, 10, 12, and 24 h, and serially diluted (10-fold serial dilutions) in MHB2. Appropriate dilutions were inoculated into MH agar plates and incubated at 37 C for 24 h. Aer incubation, colony counts were performed, and log 10 CFU mL À1 was plotted against time.

Scanning electron microscopy (SEM)
The effect of 26 on S. aureus ATCC 29213 and MRSA S. aureus 66/ 1 cellular surface was observed using scanning electron microscope. 73

Enantioselective liquid chromatography
Quantitative analysis of enantioselective liquid chromatography. Compounds 22-32 were prepared using HPLC grades nhexane : EtOH (50 : 50) at a nal concentration 50 mg mL À1 , and the injection volume was 10 mL. The HPLC system comprised a JASCO model 880-PU intelligent HPLC pump (JASCO corporation, Tokyo, Japan), equipped with a 7125 injector (Rheodyne LCC, Rohnert Park, CA, USA) tted with a 20 mL LC loop, a JASCO model 880-30 solvent mixer involving a 875-UV intelligent UV/VIS detector, a system equipped with a chiral column (Lux® 5 mm amylose-1, 250 Â 4.6 mm). The data acquisitions were performed using ChromNAC chromatography Data system (version 1.19.1) from JASCO Corporation (Tokyo, Japan). The mobile phase consisted of the mixture of n-hexane : EtOH (90 : 10, v/v), at a ow rate of 0.5 mL min À1 . The mobile phase was prepared in a volume/volume ratio and degassed in an ultrasonic bath for at least 15 min before use. Chromatographic analyses were carried out in isocratic mode at 22 AE 2 C, in duplicate. The UV detection was performed at a wavelength of 254 nm. The volume void time was considered to be equal to the peak of solvent front and was taken from each particular run. The enantiomeric ratios (e.r) were determined by the mean percentage of peak area of eluted peaks.
Semi-preparative enantioselective resolution. Compounds 22, 23, and 26 were prepared in the mixture of HPLC grade solvent n-hexane : EtOH (50 : 50) at a concentration of 10 mg mL À1 , and the injection volume was 100-200 mL. The HPLC system is similar to that described in quantitative analysis, equipped with an in-house column amylose tris-3,5dimethylphenylcarbamate coated with Nucleosil (500 A, 7 mm, 20%, w/w) packed into a stainless steel (200 mm Â 7 mm I.D. size) column, prepared in the UFSCar laboratory. 63 Semipreparative chromatographic separations were rst achieved through multiple injection with 200 mL at a ow rate of 2 mL min À1 . Chromatographic analyses were carried out in isocratic mode at 22 AE 2 C. The UV detection was performed at a wavelength of 254 nm. The collected fractions were analyzed using the analytical column to determine their enantiomeric ratio/excess with the conditions described above.
Cell-based studies SH-SY5Y cell culture and differentiation. SH-SY5Y cells (ATCC, United States of America) were routinely cultured in 25 cm 2 asks using DMEM with 4.5 g L À1 glucose supplemented with 10% heat-inactivated FBS, 100 U mL À1 of penicillin and This journal is © The Royal Society of Chemistry 2020 RSC Adv., 2020, 10, 31187-31204 | 31199 100 mg mL À1 of streptomycin. Cells were maintained in a 5% CO 2 to 95% air atmosphere, at 37 C, and the medium was changed every 2-3 days. When 80-90% conuence was reached, the cultures were passaged by trypsinization (0.25% trypsin/ 1 mM EDTA). In all experiments, the cells were seeded in 96 well plates at a density of 25 000 cells per cm 2 and used 6 days aer seeding. To obtain cells with a dopaminergic neuronal phenotype, SH-SY5Y cells were differentiated as previously described. 74 Briey, SH-SY5Y cells were seeded in complete DMEM medium containing 10 mM RA, and cultured for 3 days at 37 C. Aer 3 days, 80 nM of TPA was added to the cultures, and cells were cultured for another 3 days, at 37 C. The nondifferentiated SH-SY5Y cells were maintained in complete DMEM medium for 6 days, mimicking the differentiation protocol but in the absence of both RA and TPA. The cells used in all experiments were taken between the 21 st and 25 th passages.
Evaluation of cytotoxicity of 26. Cytotoxity of 26 (0-25 mM) was evaluated in both differentiated and non-differentiated SH-SY5Y cells, by the neutral red (NR) uptake, resazurin (REZ) reduction and sulforhodamine B (SRB) assays, 24 h aer exposure. For that purpose, SH-SY5Y cells were seeded in 96-well plates at a density of 25 000 cells per cm 2 , submitted or not to a RA and TPA differentiation protocol, and exposed, 6 days aer seeding to 26 (0-25 mM), in fresh cell culture medium. Triton™ X-100 (0.1%) was used as positive control.
Neutral red uptake assay. The neutral read (NR) uptake assay, which is based on the capacity of viable cells to incorporate and bind the weak cationic dye NR into the lysosomes, provides a quantitative estimation of the number of viable cells in a culture. 75,76 The NR dye is rst incorporated into the cells according to their lysosomal functionality, and then extracted from the viable cells using an acidied ethanol solution, being the absorbance of the solubilized dye quantied using a spectrophotometer. 77 Twenty-four hours aer exposure to 26, the cell culture medium was removed, followed by the addition of fresh cell culture medium containing 50 mg mL À1 NR, and incubation at 37 C, in a humidied 5% CO 2 to 95% air atmosphere, for 60 minutes. Aer the 60 min incubation with NR, the cell culture medium was removed, the dye absorbed only by viable cells extracted with lysis buffer [absolute ethyl alcohol/ distilled water (1 : 1) with 5% acetic acid], and the absorbance measured at 540 nm in a multiwell plate reader (PowerWaveX BioTek Instruments, Vermont, USA). The percentage of NR uptake relatively to that of the control cells (0 mM) was used as the cytotoxicity measure. Four independent experiments were performed, in triplicate.
Resazurin reduction assay. The resazurin (REZ) reduction assay represents a simple, rapid, and sensitive method for the evaluation of cell's viability. It is based on the ability of metabolically active cells to reduce the nonuorescent dye REZ to the strongly-uorescent dye resorun, being the uorescence proportional to the number of viable cells in culture. 78 Aer incubation with compound 26 for 24 h, the cell culture medium was removed, followed by the addition of fresh cell culture medium containing 10 mg mL À1 REZ, and incubation at 37 C, in a humidied 5% CO 2 to 95% air atmosphere, for 60 minutes.
The uorescence was then read in a multiwell plate reader (PowerWaveX BioTek Instruments, Vermont, USA), using an excitation and emission wavelengths of 560 nm and 590 nm, respectively. The percentage of REZ reduction relatively to that of the control cells (0 mM) was used as the cytotoxicity measure. Four independent experiments were performed, in triplicate.
Sulforhodamine B assay. The sulforhodamine B (SRB) assay has been widely used to measure drug-induced cytotoxicity and cell proliferation, and it is based on the ability of the protein dye SRB to bind stoichiometrically to protein basic amino acid residues under mild acidic conditions. The bound SRB can then be extracted using basic conditions, and the amount of bound SRB can be used as a proxy for cell mass. 79 Aer incubation with 26 for 24 h, the cell culture medium was removed, the cells washed with HBSS (+/+) and xed overnight at À20 C with a methanolic solution of 1% acetic acid (v/v). The xing medium was then removed, replaced by 0.05% SRB solution (prepared in 1% acetic acid), followed by incubation at 37 C for 60 minutes. Aer incubation, the SRB solution was removed and the cells washed with 1% acetic acid (v/v) to remove unbound dye. The plates were let to air-dry at room temperature and the bound SRB extracted with a Tris base solution (10 mM, pH 10.5). The absorbance was measured at 540 nm in a multiwell plate reader (PowerWaveX BioTek Instruments, Vermont, USA). The percentage of SRB binding relatively to that of the control cells (0 mM) was used as the cytotoxicity measure. Three independent experiments were performed, in triplicate.
Statistical analysis. In the cytotoxicity assays, GraphPad Prism 6 for Windows (GraphPad Soware, San Diego, CA, USA) was used to perform all statistical calculations. Three tests were performed to evaluate the normality of the data distribution: KS, D'Agostino & Pearson omnibus and Shapiro-Wilk normality tests. For the cytotoxicity results, one-way ANOVA was used to perform the statistical comparisons, followed by the Dunnett's multiple comparisons test. Details of the performed statistical analysis are described in the gure legend. Differences were considered to be signicant for p values lower than 0.05.
Molecular dynamics simulation (MD). A 5Å spherical droplet containing 100 water molecules was placed surrounding two conformations of 26 obtained from the docking study on GyrB; the complexes were energy minimized using the MMFF94x force eld 83 until RMS gradient < 0.1 kcalÅ À1 . mol À1 . The 26:GyrB complexes were then subjected to MD simulation using MOEdynamic implemented in MOE 2014.09 (Chemical Computing Groups, Montreal, Canada). 84 MD simulation 85,86 was done by choosing MMFF94x force eld and NVT (N, total atom; V, volume; T, temperature) ensemble and Nosé-Poincaré-Andersen (NPA) algorithm, 87 with 0.002 ps time step and sampling every 0.5 ps. The system was heated from 0 K to 300 K in 100 ps (heat stage), followed by a 2000 ps of a production stage at 300 K; the system was then cooled back to 0 K in 100 ps (cooling stage). 88 Simulation observation was done by examining the 26:GyrB complex interaction between ligand atoms and target atoms at the end of the simulation. The system behaviour was monitored by the analysis of the potential energy over the course of the simulation.

Conflicts of interest
There are no conicts to declare.