Synthesis and Biological Evaluation of New N-Acyl-α-amino Ketones and 1,3-Oxazoles Derivatives

In order to develop novel bioactive substances with potent activities, some new valine-derived compounds incorporating a 4-(phenylsulfonyl)phenyl fragment, namely, acyclic precursors from N-acyl-α-amino acids and N-acyl-α-amino ketones classes, and heterocycles from the large family of 1,3-oxazole-based compounds, were synthesized. The structures of the new compounds were established using elemental analysis and spectral (UV-Vis, FT-IR, MS, NMR) data, and their purity was checked by reversed-phase HPLC. The newly synthesized compounds were evaluated for their antimicrobial and antibiofilm activities, for toxicity on D. magna, and by in silico studies regarding their potential mechanism of action and toxicity. The 2-aza-3-isopropyl-1-[4-(phenylsulfonyl)phenyl]-1,4-butanedione 4b bearing a p-tolyl group in 4-position exhibited the best antibacterial activity against the planktonic growth of both Gram-positive and Gram-negative strains, while the N-acyl-α-amino acid 2 and 1,3-oxazol-5(4H)-one 3 inhibited the Enterococcus faecium biofilms. Despite not all newly synthesized compounds showing significant biological activity, the general scaffold allows several future optimizations for obtaining better novel antimicrobial agents by the introduction of various substituents on the phenyl moiety at position 5 of the 1,3-oxazole nucleus.


Introduction
Many heterocyclic compounds are very important in medicinal chemistry since they exhibit remarkable and various pharmacological activities, being present as active substances in the composition of numerous potent drugs.
Derivatives of the above classes were linked with a fragment derived from diphenyl sulfone with the purpose of obtaining new compounds with potent biological properties. The choice of this pharmacophore center is justified, on the one hand, by the fact that numerous diaryl sulfones (e.g., dapsone, acedapsone, glucosulfone, sulfoxone, thiazosulfone) were found to possess antimicrobial, antimalarial, antioxidant, anticancer properties [27][28][29][30][31][32][33] and, on the other hand, by data from the literature indicating that it was incorporated into various heterocyclic systems with biological value [34,35].

Spectral Characterization UV-Vis Spectral Data
The UV-Vis spectra of new compounds 2-5 showed the E band at λ max = 202.6 nm and the B band in the range of 241.0-252.0 nm. Furthermore, 2,5-diaryl-4-isopropyl-1,3oxazoles 5a,b spectra presented a third absorption band at a longer wavelength: 331.3 (5a) or 336.6 (5b) nm, due to the appearance of the 1,3-oxazole chromophore, which determined the extension of the π electrons conjugation.

IR Spectral Data
For acyclic intermediates 2 and 4a,b, a characteristic absorption band due to N-H stretching vibration, ν(N-H), was registered in the range of 3302-3424 cm −1 . In the IR spectrum of N-acyl-α-amino acid 2, the peak due to carbonyl valence vibration, ν(O=C- C), was recorded at 1720 cm −1 , and amidic carbonyl absorption, ν(O=C-N), at 1674 cm −1 . These two carbonyl absorption bands are overlapped in the case of N-acyl-α-amino ketones 4a,b, as suggested by the single very strong peak present in their spectra at 1657 (4b) or 1662 (4a) cm −1 . Representative bands for the N-acyl-L-valine 2 associated by hydrogen bonds are also: a strong, broad absorption peak between 2500 and 3300 cm −1 , and two medium broad bands at 2686 and 2609 cm −1 due to O-H stretching vibration, ν(O-H).

NMR Spectral Data
The NMR spectral data also proved the structures of the new compounds (Supplementary Materials).

H-NMR Spectral Data
The numbering of atoms used for assigning the NMR signals of the compounds 2-5 is presented in Figure 3. In 1 H-NMR spectra of new compounds 2 and 4a,b, for the deshielded proton of NH group, respectively H-3, the signal was registered as a doublet at δ H = 8.38 ppm (2) and 8.96 (4b) or 8.99 (4a) ppm, due to the vicinal coupling to H-4. In the 1 H-NMR spectrum of the N-acyl-α-amino acid 2, a doublet of doublets signal at 4.26 ppm was assigned to the H-4 proton, coupled to the proton of the NH group and the H-18 methine proton. In the case of N-acyl-α-amino ketones 4a,b, the H-4 proton signal appeared as a triplet at 5.35 (4b) or 5.38 (4a) ppm, due to the coupling to H-3 and H-18. In the 1 H-NMR spectra of acyclic intermediates 2 and 4a,b, for the isopropyl group, an octet or multiplet signal was recorded in the 2.12-2.26 ppm range due to the proton of the methine group (H-18) and two strongly shielded doublet signals at δ H = 0.89 (2, 4b) or 0.90 (4a) ppm and in the interval: 0.90-0.92 ppm, respectively, corresponding to the H-19 and H-20 nonequivalent protons from the two methyl groups.
A proof for the intramolecular cyclocondensations of precursors 2 and 4a,b is represented by the absence in 1 H-NMR spectra of heterocyclic compounds 3 and 5a,b of the signal assigned to the NH proton.
In the case of 4-isopropyl-1,3-oxazol-5(4H)-one 3, the H-4 signal was recorded at 4.32 ppm as a doublet due to coupling to the adjacent methine proton and the signal of H-18 at 2.39 ppm as a heptet of doublets, this proton being coupled to H-19 and H-20 protons (with 3 J = 6.9 Hz) and H-4 (with 3 J = 4.7 Hz). In the 1 H-NMR spectra of 1,3-oxazoles 5a,b, for the isopropyl group, a heptet signal appeared at δ H = 3.26 (5b) or 3.28 (5a) ppm due to the H-18 proton (shifted downfield compared to the corresponding proton signal registered for N-acyl-α-amino ketones 4a,b), and a strongly shielded doublet at δ H = 1.35 (5b) or 1.36 (5a) ppm, corresponding to protons of the two CH 3 groups. 13

C-NMR Spectral Data
In the 13 C-NMR spectrum of N-acyl-α-amino acid 2, the signal due to the C-4 atom was observed at a chemical shift value of 58.43 ppm. The isopropyl group is highlighted by the presence of the carbon atom of the methine group for which a signal was recorded at 29.48 ppm and by the presence of nonequivalent carbon atoms of methyl groups, which showed two signals at δ C = 18.61 and 19.24 ppm, respectively. The C-4 signal was shifted downfield with 12.59 ppm after cyclodehydration of compound 2 to saturated azlactone 3. Further, in the case of 1,3-oxazol-5(4H)-one 3, the C-2 atom resonated at 160.44 ppm, being more shielded with 5.43 ppm than the corresponding carbon of intermediate 2, and the C-5 at 177.05 ppm, being shifted downfield with 4.24 ppm than the corresponding atom of 2.

Mass Spectral Data
The mass spectra recorded by the L-ESI-MS/MS technique had an additional contribution to the elucidation of the structures of compounds 2, 3, 4a, and 5a.
In this study was used the ability of ionization at atmospheric pressure of electrospray ionization (ESI) type for amino acid derivatives analysis. Further, a special advantage of the ESI source is that negative ions can also be analyzed.
In our experiments, compounds 2 and 4a, dissolved in methanol/water with 0.1% ammonium carbonate (9:1, v/v), were ionized positively and negatively, respectively. For compounds 3 and 5a, methanol/water (containing 0.1% ammonium formate and 1% formic acid) 9:1 (v/v) was used as the solvent mixture. After evaporation of the solvents, the ions obtained were introduced into the mass spectrometer.
ESI is considered a mild source of ionization. Consequently, the mass spectra of these compounds are very simple and consist mainly of the protonated molecular ion [M + H] + for positive ionization or the deprotonated molecular ion [M−H]for negative ionization. Non-covalent dipole-dipole interactions can be highlighted by using ESI so that in the compounds spectra appeared combinations like [2M + H] + and [2M−H] − , respectively (e.g., for 4a).
Valuable structural information can be obtained from the fragmentation by the collision of pseudo-molecular ions with an inert gas (argon), as fragmentation in positive-ion mode can be different from fragmentation in negative-ion mode. In the positive ionization mode, the first fragmentation occurred with water (for 2), carbon monoxide (for 3), or methane loss (for 5a), and in the negative ionization mode with carbon dioxide (for 2) or isopropyl radical loss (for 4a). Protonated and/or deprotonated molecular ions and the main fragments of these compounds are reported in the Materials and Methods section.
In the mass spectrum of 1,3-oxazol-5(4H)-one 3 which was also achieved by GC-EI-MS analysis, the molecular ion, [M] + , being unstable, did not show a signal. 1,3-Oxazol-5(4H)one 3 was first fragmented at the side chain from 4-position with the removal of a propene molecule and the formation of cation-radical with m/z = 301, which is the base peak (BP). Other main fragments of 3 are indicated in Materials and Methods.

Qualitative Analysis of the Antimicrobial Activity
The results of the qualitative analysis of the antimicrobial activity of the newly synthesized compounds showed that the majority of the compounds did not produce growth inhibition zones, except for compounds 2 and 3, which inhibited the growth of the Grampositive strain Enterococcus faecium E5, producing a growth inhibition zone of 20 and 17 mm, respectively.

Investigation of the Influence of the Tested Compounds on the Antibiotic Susceptibility Spectrum of Enterococcus faecium E5
E. faecium is increasingly identified in nosocomial infections and it has rapidly adapted to newer anti-gram-positive agents (e.g., linezolid, quinupristin/dalfopristin, daptomycin, tigecycline) [47]. Therefore, the development of new drugs to combat these recalcitrant microorganisms is needed. For compounds 2 and 3 only, which proved to be active in the qualitative disk diffusion assay against E. faecium E5, their influence on the antibiotic susceptibility spectrum of this strain was evaluated. Antibiotic susceptibility tests were performed and interpreted according to the CLSI. The diameters of growth inhibition zones (mm) are shown in Table 1. Kirby-Bauer disk diffusion tests showed that E. faecium E5 was resistant to penicillin and susceptible to ampicillin, linezolid, and vancomycin. Regarding the effects of compounds 2 and 3, no changes in the E. faecium E5 strain susceptibility to the tested antibiotics were determined after cultivation in the presence of subinhibitory concentrations of the two compounds, suggesting both a different mechanism of action and a low selective pressure for resistance occurrence.

Quantitative Evaluation of Antimicrobial and Antibiofilm Activities
The urgent need for novel antimicrobial agents persists, as the emergence of multidrug pathogens is both unpredictable and inevitable. Therefore, the α-amino ketones and 1,3-oxazoles derivatives received significant importance due to their wide spectrum of biological applications across synthetic and medicinal chemistry [48,49]. The newly synthesized derivatives were screened for their antimicrobial activity using the two-fold serial microdilution method, following the CLSI guidelines. Additionally, the potential of the compounds to prevent initial cell attachment was investigated through the biofilm inhibition assay. The minimal inhibitory concentration (MIC) and the minimal biofilm eradication concentration (MBEC) values (in µg/mL) obtained for the tested compounds 2-5 are presented in Table 2. The quantitative testing of the antimicrobial activity showed that the majority of the tested compounds exhibited antimicrobial effects, with MIC values equal to or higher than 500 µg/mL. Among the analyzed compounds, 4b was found to have a good antimicrobial activity (MIC value of 62.5 µg/mL) against two Grampositive, i.e., Bacillus subtilis ATCC 6683, Staphylococcus aureus ATCC 6538, and one Gramnegative, i.e., Escherichia coli ATCC 8739 reference strains. Concerning the influence on the development of microbial biofilms on the inert substrate, the compounds 4a, 4b, 5a, and 5b did not interfere with the development of microbial biofilms on the inert substrate, at the tested range of concentrations. The compounds 2 and 3 exhibited antibiofilm effects in the case of the Gram-positive E. faecium E5 strain, with an MBEC value of 15.6 µg/mL. Comparing the MIC and MBEC values obtained for the newly synthesized derivatives with those obtained for ciprofloxacin and fluconazole, respectively, the tested compounds were shown to exhibit a weak antimicrobial activity.

Daphnia Magna Toxicity Assay
The Daphnia magna bioassay results are summarized in Table 3. Both at 24 and 48 h, compounds 2, 4b, 5b, and positive controls induced lethality values lower than 35%, and therefore, the LC 50 values couldn't be determined. The LC 50 was determined at 24 h only for compound 4a, its value being greater than the maximum tested concentration as shown by the 95% CI range. After 48 h of exposure, compounds 3, 4a, and 5a showed high to moderate cytotoxicity values, compound 3 being the most active of all. The correlation between concentrations and L% was higher than 0.7 for compounds 4a and 5a. The predicted values of 3, 4a, and 5a are significantly lower than those obtained experimentally, and for all other compounds, the values are ranging from 0.18 to 68.2 µg/mL, despite no significant toxicity was recorded experimentally.

PASS Prediction
The software prediction of activity spectra for substances (PASS) is an application that can predict a large number of biological activities for a given molecule using its structure as input data. The software returns the probability of the compound to be active (Pa) or inactive (Pi) for each target [50]. The corresponding Pa values are presented in Table 4 for biological activities related to the antibacterial effects. The Pa values are generally higher for compounds 4a and 4b than for the corresponding 1,3-oxazoles 5a and 5b, but this is an indication of the possibility that the new compounds produce an effect, and not for their potency. There are small differences between the not substituted derivatives (a) and the corresponding 4-methyl analogs (b).
The resulted Pa values indicate an acceptable potential for the new compounds to have a general antiinfective effect, and in the case of compound 3, to have a similar effect with the glycopeptides antibiotics. The PASS prediction functions on the basis of the structural similarity between the tested structures and those included in the prediction set. The higher the value, the higher is the possibility to find an active drug. Nevertheless, a small probability could be an indicator for an original active structure.

Structural Similarity Analysis
The similarity search on the ChEMBL database for compounds 2, 3, 4a,b, and 5a,b returned 62 analog compounds, with the highest degree of structural similarity (80.0%) being observed for the pair formed by compound CHEMBL2071499 and 4a. The results highlight the originality of the newly synthesized compounds.

Discussion
The 1,3-oxazole moiety gained attention in recent times due to its increasing importance in the field of medicinal chemistry. Being a doubly unsaturated 5-membered ring with one oxygen atom at position 1 and one nitrogen atom at position 3, the 1,3-oxazoles were identified as antibacterial, anticancer, and anti-inflammatory agents. The development of new antibacterial active substances is an ongoing process to improve the affinity for different bacterial strains. Some 1,3-oxazoles, e.g., 4-(1-benzofuran-2-yl)-1,3-oxazole-2-amine derivatives [51], showed appreciable antimicrobial activity as compared to the standard drugs, and a number of multi-substituted oxazoles containing a heterocyclic moiety exhibited pronounced antibacterial activity against S. aureus, E. coli, B. subtilis, and K. pneumonia [52].
In this paper, we report the synthesis of novel derivatives based on 1,3-oxazole and diphenyl sulfone scaffolds, starting from a natural α-amino acid, namely, from L-valine. The chemical structures of the novel compounds 2-5 were confirmed based on spectral studies. Furthermore, the newly synthesized compounds were evaluated for their antimicrobial and antibiofilm activities, for toxicity on D. magna, and by in silico studies regarding their potential mechanism of action and toxicity. Among the new tested compounds, the best antibacterial activity was revealed by compound 4b (MIC value: 62.5 µg/mL) against B. subtilis, S. aureus, and E. coli reference strains. Additionally, compounds 2 and 3 were very active against the biofilm formed by E. faecalis. The Daphnia magna model is frequently used for determining the toxicity of drugs, biocompounds, plant extracts, as well as in ecotoxicological evaluations, and can be a useful tool for the anticancer and antibacterial activities prediction [53]. In the present study, the newly synthesized compounds were tested at selected concentrations, mainly chosen based on their solubility. The most toxic compound was compound 3, followed by 4a and 5a. All other compounds showed low toxicity or no toxicity at all. The results indicate that these compounds have biological activity and are good candidates for further investigations. The high difference between the predicted and the experimentally obtained values of LC 50 can be attributed to the low solubility in water of these compounds.
It can be noticed that the N-acyl-α-amino acid derivative 2 is promising due to its antibacterial and antibiofilm activities and low toxicity. Comparing the structure of this compound with that of the most probable pharmacological target: CHEMBL4544788, the biological action is probably due to the presence in the structure of the valine residue and also of the fragment derived from diphenyl sulfone. The antibacterial activity and antibiofilm effect against the Gram-positive E. faecium E5 strain are maintained by conversion of compound 2 to the corresponding 1,3-oxazol-5(4H)-one 3, but the toxicity of the reaction product increases. Further, it appears that the transformation of N-acyl-α-amino acid 2 to N-acyl-α-amino ketone 4b (via 1,3-oxazol-5(4H)-one 3) does not increase the toxicity, but determines an improvement in the antimicrobial profile, with the disappearance of the antibiofilm activity. Compounds 4a, 5a, and 5b were proved inactive, in the tested range of concentrations. Subsequent studies will reveal the structure-activity relationships (SAR). The different activity profiles for N-acyl-α-amino ketones 4 and 1,3-oxazoles derivatives 5 are probably due to the fact that acyclic intermediates 4 are much more reactive than the corresponding cyclodehydration products 5. The higher antimicrobial potential of N-acylα-amino ketones 4 may be due to the presence in their structures of the two C=O double bonds and of the NH group which are both centers of reactivity and allow the formation of hydrogen bonds, being known that the intermolecular hydrogen-bonding capability is important to biological activity [54]. By cyclizing precursors 4, to the corresponding heterocycles 5, all these groups are no longer present in the structures of the molecules.
Further studies will be made for optimizations in order to obtain new derivatives with potent antimicrobial and antibiofilm activities. Thus, we identified three critical positions, which influence biological activity. A first possibility is to add various substituents on the diphenyl sulfone fragment, like nitro or fluoro, e.g., by using other aromatic compounds in the Friedel-Crafts sulfonylation. Desai et al. showed that the nitro group leads to a better efficacy against S. pyogenes, whereas the presence of the fluorine atom could increase the affinity for the active site of DNA gyrase [55]. The predictive studies indicated a high degree of similarity of 4a and 2 with CHEMBL4520114 and CHEMBL4544788, respectively outlining a potential future synthesis approach. As a second choice of future improvement of these derivatives, other amino acids will be taken into account-cysteine, α-alanine, phenylalanine, and serine. As previously showed by Mhlongo et al., the naturally occurring oxazole-containing peptides, such as muscoride A and microcin B17, possessed strong antibacterial activity against E. coli strains by inhibition of DNA gyrase [13]. A third possibility is to use another aromatic compound in the Friedel-Crafts acylation reaction with 1,3-oxazol-5(4H)-one 3. This step could also improve the antibacterial efficacy. Thus, in the future, we will consider the introduction of various substituents (e.g., a fluorine atom or a trifluoromethyl group) on the phenyl moiety at position 5 of the 1,3-oxazole nucleus for obtaining strong antimicrobial and antibiofilm analogs.

4.1.General Information
The melting points, m.p., were measured on a Boëtius apparatus (VEB Wägetechnik Rapido, PHMK 81/3026, Radebeul, Germany) and are uncorrected. The UV-Vis spectra were registered on a Specord 40 spectrophotometer (Analytik Jena AG, Jena, Germany) in a 1 cm pathlength quartz cuvette, for solutions in methanol (≈0.025 mM). The FT-IR spectra were acquired in KBr pellets on a Vertex 70 spectrometer (Bruker Optik GmbH, Ettlingen, Germany). Selected IR absorption bands are described as very strong, vs; strong, s; medium, m; weak; w. The NMR spectra were recorded on a Gemini 300 BB spectrometer (Varian, Inc., Palo Alto, CA, USA) in DMSO-d 6 or CDCl 3 , at room temperature, operating at 300 MHz for 1 H and 75 MHz for 13 C. Combined 2D spectra (COSY, HETCOR) were also registered. Chemical shifts, δ, are in parts per million (ppm) relative to tetramethylsilane (TMS) used as internal standard and coupling constants, J, are expressed in hertz (Hz). The 1 H-NMR signals multiplicity was abbreviated as follows: singlet, s; doublet, d; doublet of doublets, dd; triplet, t; triplet of triplets, tt; heptet, hp; heptet of doublets, hpd; octet, oct; multiplet, m; and a broad signal was abbreviated br. The 1 H-NMR data are reported in the following order: chemical shift (multiplicity, coupling constants, number of protons, proton assignment), and the 13 C-NMR data are quoted as follows: chemical shift (carbon attribution). Mass spectra of one representative from each class (2, 3, 4a, 5a) were recorded on a Varian 1200L MS/MS triple quadrupole mass spectrometer (Varian, Inc., Walnut Creek, CA, USA) with an electrospray interface, in positive and/or negative ionization mode. A solution in methanol/water with 0.1% ammonium carbonate (or 0.1% ammonium formate and 1% formic acid) 9/1 (v/v) of around 1 ppm of these compounds was directly infused with a Prostar 240 SDM at 70 µL/min flow rate, using a Rheodyne manual injector. Pseudo-molecular ions (protonated molecules or deprotonated molecular ions) were selected with the first quadrupole. Fragments were obtained by collision with argon at different energies up to 50 eV. GC-EI-MS analysis was performed on a GC 8000 gas chromatograph, equipped with an electron impact quadrupole, and coupled to an MD 800 mass spectrometer detector (Fisons Instruments SpA, Rodano, Milano, Italy), using a fused-silica capillary column coated with poly(5% diphenyl/95% dimethylsiloxane) (SLB-5ms, 30 m × 0.32 mm, d f 0.25 µm), a helium carrier gas flow rate of 2 mL/min, and dichloromethane as solvent. RP-HPLC chromatograms were acquired on a System Gold 126 liquid chromatograph (Beckman Coulter, Inc., Fullerton, CA, USA), with a System Gold 166 UV-Vis detector, a Rheodyne injection system, and a non-polar chromatography column (LiChrosorb RP-18, 25 cm × 4.6 mm, 5 µm particle size). The flow rate of the mobile phase (a mixture of methanol-water in various proportions) was 1 mL/min. Compounds' purity (%) and retention time, t R , in minutes (min) were indicated. The elemental analysis was performed on an ECS 4010 elemental analyzer (Costech Analytical Technologies Inc., Valencia, CA, USA).

Chemistry
All the chemicals and reagents were purchased from commercial suppliers and used without purification. Dichloromethane was dried over anhydrous calcium chloride.

Antimicrobial Activity Assessment
The antimicrobial activity of the compounds was investigated using the agar discdiffusion method, broth microdilution, and microtiter plate assay.

Qualitative Assessment of the Antimicrobial Activity
The qualitative assessment of the antimicrobial activity of the tested compounds was performed using the agar disc-diffusion method. The Mueller Hinton (MH) agar plates are inoculated with a standardized microbial inoculum prepared in phosphate-buffered saline and adjusted to 0.5 McFarland scale. Then, five microliters of the solution of the tested compound (of 5000 µg/mL concentration in DMSO) were pipetted on the inoculated agar surface. After incubation of the plates for 24 h at 37 • C, the diameters of inhibition growth zones were measured.

Investigation of the Influence of the Tested Compounds on the Antibiotic Susceptibility Spectrum of the Studied Strain
Subinhibitory concentrations (250 µg/mL) of tested compounds 2 and 3 were achieved in sterile liquid culture medium; the obtained tubes were inoculated with 0.5 McFarland suspensions obtained from the 24 h microbial culture of E. faecium E5. A control of DMSO (bacterial strain grown in the presence of DMSO) and control of microbial growth (culture medium inoculated with microbial suspension) were also prepared. The inoculated tubes were incubated at 37 • C for 24 h. The microbial cultures obtained in the presence of subinhibitory concentrations of tested compounds 2 and 3, DMSO and control cultures, respectively, were used to determine the influence of the tested compounds on the antibiotic susceptibility spectrum of the studied strain. Thus, the liquid microbial cultures were sedimented by centrifugation at 10,000× g for 5 min, and the obtained cellular sediment was washed 3 times in sterile saline by centrifugation at 10,000× g for 5 min. The cell pellet was resuspended in sterile saline until a turbidity corresponding to the 0.5 McFarland standard was obtained and the classical Kirby-Bauer method was then performed to assess the susceptibility to the following antibiotics (bioMérieux, France): ampicillin, penicillin, linezolid, vancomycin for the E. faecium E5 strain. The results were recorded after 24 h of incubation at 37 • C. The diameters of the areas of inhibition of bacterial growth were interpreted according to the recommendations of the current edition of the Clinical and Laboratory Standards Institute (CLSI).

Quantitative Assessment of the Antimicrobial Activity
Quantitative assessment of the antimicrobial activity of the tested compounds was performed using broth microdilution, in 96-well microtiter plates. The tested concentrations of the solutions of the different compounds in DMSO achieved through double serial dilutions, in columns 1-10, were between 500-0.97 µg/mL. Then, the wells were inoculated with 10 µL microbial suspension prepared in the same medium after dilution (1:100) of standardized microbial inoculum adjusted to 0.5 McFarland scale. Column 11 contained 10 µL of standardized inoculum and 90 µL of Mueller Hinton Broth, and column 12 contained 100 µL of Mueller Hinton Broth (as a control to monitor sterility). Ciprofloxacin (Sigma-Aldrich, St. Louis, MO, USA) and fluconazole (Sigma-Aldrich) served as positive controls. The microtiter plates were incubated without agitation for 24 h at 37 • C. In order to confirm the MIC value, the assays were performed in triplicate. The MIC was determined as the lowest concentration of tested compound that inhibited the growth of the microorganism as detected spectrophotometrically at 620 nm with an Apollo LB 911 ELISA Reader (Berthold Technologies GmbH & Co. KG, Waltham, MA, USA) [56].

Evaluation of the Antibiofilm Activity
Evaluation of the antibiofilm activity of tested compounds was carried out using the microtiter biofilm inhibition assay. Briefly, in a 96-well polystyrene microtiter plate containing 90 µL of Mueller Hinton Broth, 90 µL of the tested compound solution was added in column 1. Serial two-fold dilutions were performed in columns 1-10. A volume of 10 µL microbial suspension (final OD600 = 0.01), prepared from an overnight culture grown in TSB into MH broth, was added. Column 11 contained 10 µL of standardized inoculum and 90 µL of Mueller Hinton Broth, and column 12 contained 100 µL of Mueller Hinton Broth (as a control to monitor sterility). Ciprofloxacin (Sigma-Aldrich) and fluconazole (Sigma-Aldrich) served as positive controls. Microplates were incubated for 24 h at 37 • C under static conditions to allow for microbial growth and biofilm maturation. The wells of the microplate were emptied and washed twice with phosphate-buffered saline. The biofilms formed on the walls of wells of the microplate were fixed with methanol for 5 min and stained with 1% crystal violet solution for 15 min, then rinsed three times with distilled water to remove the unbound dye. The fixed dye was resuspended in 33% acetic acid and the A492 was recorded with an Apollo LB 911 ELISA Reader. The amount of biofilm inhibition was calculated relative to the amount of biofilm that was grown in the absence of the tested compound and the media sterility control. The minimal biofilm eradication concentration (MBEC) was determined to be the lowest concentration of the tested compounds at which the decrease in absorbance value, measured at 492 nm, was observed in comparison to the positive control. Results from at least three separate biological replicates were averaged [57].

Daphnia Magna Toxicity Assay
D. magna Straus was maintained parthenogenetically ('Carol Davila' University-Department of Pharmaceutical Botany and Cell Biology) at 25 • C with a photoperiod of 16 h/8 h light/dark cycle in a Sanyo MLR-351 H climatic chamber (Sanyo, San Diego, CA, USA). With 24 h prior to the determination, young daphnids were selected according to their size and maintained for 24 h in an artificial medium. Ten daphnids/replicate were used, and the determination was performed in tissue culture plates with 12 wells (Greiner Bio-One) [58,59]. Each compound was tested in six concentrations, ranging from 2.2 to 44 µg/mL. L-Valine-control 1 (2.5-50 µg/mL) and 4-(phenylsulfonyl)benzoic acidcontrol 2 (1.1-22 µg/mL) were used as positive controls, and a 1% DMSO solution as a negative control. The concentration ranges were selected based on the solubility and a pre-screening assay. All determinations were performed in duplicate. The lethality was evaluated after 24 and 48 h of exposure. LC 50 and 95% confidence intervals (95% CI) were calculated using the least square fit method. All calculations were performed using GraphPad Prism v 5.1 software (GraphPad Software, Inc., La Jolla, CA, USA). Freely available online GUSAR software (Institute of Biomedical Chemistry, Moscow, Russia) was used to predict the LC 50 values for 48 h exposure of the new compounds [60].

PASS Prediction
A virtual screening was performed using the software PASS (Prediction of Activity Spectra for Substances), an application designed to evaluate the pharmacological potential of newly synthesized compounds. The structures were inputted in PASS as SMILES and the results were analyzed if the Pa values were above the corresponding Pi values.

Structural Similarity Analysis
A similarity search was performed on the ChEMBL database for the newly synthesized compounds using a 50% threshold [58]. The resulting structures were extracted together with their assayed activities on bacteria [61]. The entries were filtered using DataWarrior v5.2.1 software [62] to remove duplicate structures.

Data Availability Statement:
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.