N-Derivatives of (Z)-Methyl 3-(4-Oxo-2-thioxothiazolidin-5-ylidene)methyl)-1H-indole-2-carboxylates as Antimicrobial Agents—In Silico and In Vitro Evaluation

Herein, we report the experimental evaluation of the antimicrobial activity of seventeen new (Z)-methyl 3-(4-oxo-2-thioxothiazolidin-5-ylidene)methyl)-1H-indole-2-carboxylate derivatives. All tested compounds exhibited antibacterial activity against eight Gram-positive and Gram-negative bacteria. Their activity exceeded those of ampicillin as well as streptomycin by 10–50 fold. The most sensitive bacterium was En. Cloacae, while E. coli was the most resistant one, followed by M. flavus. The most active compound appeared to be compound 8 with MIC at 0.004–0.03 mg/mL and MBC at 0.008–0.06 mg/mL. The antifungal activity of tested compounds was good to excellent with MIC in the range of 0.004–0.06 mg/mL, with compound 15 being the most potent. T. viride was the most sensitive fungal, while A. fumigatus was the most resistant one. Docking studies revealed that the inhibition of E. coli MurB is probably responsible for their antibacterial activity, while 14a–lanosterol demethylase of CYP51Ca is involved in the mechanism of antifungal activity. Furthermore, drug-likeness and ADMET profile prediction were performed. Finally, the cytotoxicity studies were performed for the most active compounds using MTT assay against normal MRC5 cells.


Introduction
Infectious diseases symbolize a consequential global strain on public health security and impact socio-economic stability all over the world. They have monopolized for centuries the dominant factors of death and disability of millions of humans.
Since their discovery, antimicrobial agents have been a reliable weapon in fighting lifethreatening infections. Despite the availability of effective antibiotics for the most common infections, the emergence of multidrug-resistant bacteria pathogens and the spread of new infectious diseases threaten to weaken the efficiency of the drugs currently approved for infectious disease therapy [1,2].
On the other hand, infections evoked by several pathogenic fungi species result in 1.5 million deaths worldwide every year [3]. The therapeutic approach to these diseases is still challenging due to the continuous increase of resistance to antifungal drugs in sufferers, especially in immuno-suppressed patients (cancer, transplants, HIV) and those who are In addition, 2 hodamine derivatives demonstrate a broad spectrum of pharmacological activities, such as antimicrobial [6][7][8][9], anticancer [10][11][12], anti-HIV [13,14], anti-diabetic [15,16], anti-tubercular [17], and immunoproteasome inhibitory activity [18], suggesting that this scaffold occupies a vital position in drug discovery.
On the other hand, infections evoked by several pathogenic fungi species result in 1.5 million deaths worldwide every year [3]. The therapeutic approach to these diseases is still challenging due to the continuous increase of resistance to antifungal drugs in sufferers, especially in immuno-suppressed patients (cancer, transplants, HIV) and those who are frequently treated with antimitotic therapy. Therefore, there is an urgent necessity to put efforts into discovering novel agents against invasive microbial infections.
Since the molecular hybridization approach may result in compounds that inhibit more than one target, we set up our investigation to discover novel antibacterial and antifungal agents based on this rationale [44].
So far, many indole-based rhodanine derivatives ( Figure 3) have been proposed [6,41,[45][46][47][48][49] as antimicrobial agents, some of which possess high efficacy in treating multi-drug resistant pathogens [6,41,[47][48][49]. Therefore, it is noteworthy that the combination of indole and rhodanine scaffolds into new chemical entities could be a promising strategy for antimicrobial therapies. Inspired by this strategy, we have described in our previous paper [6,41] the design and synthesis of potential antimicrobial agents by incorporating indole and rhodanine moieties into new molecules. Herein, we report the synthesis, the evaluation of antimicrobial activity, molecular docking studies, and the prediction of pharmacokinetic and toxicity profiles of our compounds. In recent years, the concept of designing hybrid molecules that contain two or more pharmacophore groups bound together covalently into one molecular framework is gaining ground. There are some publications regarding the importance of co-operative hydrogen bonding in antibiotics and the molecular hybridization approach of sugar-fused indoles [42,43].
Since the molecular hybridization approach may result in compounds that inhibit more than one target, we set up our investigation to discover novel antibacterial and antifungal agents based on this rationale [44].
So far, many indole-based rhodanine derivatives ( Figure 3) have been proposed [6,41,[45][46][47][48][49] as antimicrobial agents, some of which possess high efficacy in treating multi-drug resistant pathogens [6,41,[47][48][49]. Therefore, it is noteworthy that the combination of indole and rhodanine scaffolds into new chemical entities could be a promising strategy for antimicrobial therapies. Inspired by this strategy, we have described in our previous paper [6,41] the design and synthesis of potential antimicrobial agents by incorporating indole and rhodanine moieties into new molecules. Herein, we report the synthesis, the evaluation of antimicrobial activity, molecular docking studies, and the prediction of pharmacokinetic and toxicity profiles of our compounds.

Prediction of Toxicity
The prediction of compound toxicities is a key part of the drug development pipeline. Toxicity assessments in silico are not only faster than the determination of toxic doses in animals but can also facilitate the lessening of the number of experiments in vivo. In our case, a computational toxicity study was carried out prior to the in vitro evaluation of compounds 1-17 with the aim to identify and remove potentially toxic structures. ProTox-II web server was used for the prediction of various toxicity endpoints, such as rat acute toxicity, hepatotoxicity, cytotoxicity, carcinogenicity, mutagenicity, immunotoxicity, and adverse outcome (Tox21) pathways (Tables S1 and

Prediction of Toxicity
The prediction of compound toxicities is a key part of the drug development pipeline. Toxicity assessments in silico are not only faster than the determination of toxic doses in animals but can also facilitate the lessening of the number of experiments in vivo. In our case, a computational toxicity study was carried out prior to the in vitro evaluation of compounds 1-17 with the aim to identify and remove potentially toxic structures. ProTox-II web server was used for the prediction of various toxicity endpoints, such as rat acute toxicity, hepatotoxicity, cytotoxicity, carcinogenicity, mutagenicity, immunotoxicity, and adverse outcome (Tox21) pathways (Tables S1 and S2). In addition, ADMET Predictor 10.4 (Simulation Plus) was used to investigate other critical toxicity issues such as phospholipidosis and hERG-related cardiotoxicity (Table S1).
It is known that toxic doses are often given as LD 50 values in mg/kg body weight, where LD 50 represents the median lethal dose upon exposure to a compound. Toxicity classes are defined according to the globally harmonized system of classification of the labeling of chemicals (GHS), that is, Class I: fatal if swallowed (LD50 ≤ 5); Class II: fa- tal if swallowed (5 < LD50 ≤ 50); Class III: toxic if swallowed (50 < LD50 ≤ 300); Class IV: harmful if swallowed (300 < LD50 ≤ 2000); Class V: may be harmful if swallowed (2000 < LD50 ≤ 5000); and Class VI: non-toxic (LD50 > 5000). In our preliminary investigation, it was estimated (54-66% accuracy) that all examined compounds are classified in category IV, indicating a promising profile towards oral rat acute toxicity. Furthermore, our compounds did not predict to demonstrate adverse outcomes (Table S2).

Chemistry
Compounds were synthesized according to the process described in detail in our previous paper [41] and presented in Scheme 1. The structures of all synthesized compounds are illustrated in Table 1 (Table S1) and confirmed by 1 H and 13 C NMR spectroscopy and elemental analysis, which are described in the experimental part. S2). In addition, ADMET Predictor 10.4 (Simulation Plus) was used to investigate other critical toxicity issues such as phospholipidosis and hERG-related cardiotoxicity (Table  S1).
It is known that toxic doses are often given as LD50 values in mg/kg body weight, where LD50 represents the median lethal dose upon exposure to a compound. Toxicity classes are defined according to the globally harmonized system of classification of the labeling of chemicals (GHS), that is, Class I: fatal if swallowed (LD50 ≤ 5); Class II: fatal if swallowed (5 < LD50 ≤ 50); Class III: toxic if swallowed (50 < LD50 ≤ 300); Class IV: harmful if swallowed (300 < LD50 ≤ 2000); Class V: may be harmful if swallowed (2000 < LD50 ≤ 5000); and Class VI: non-toxic (LD50 > 5000). In our preliminary investigation, it was estimated (54-66% accuracy) that all examined compounds are classified in category IV, indicating a promising profile towards oral rat acute toxicity. Furthermore, our compounds did not predict to demonstrate adverse outcomes (Table S2).

Chemistry
Compounds were synthesized according to the process described in detail in our previous paper [41] and presented in Scheme 1. The structures of all synthesized compounds are illustrated in Table 1 (Table S1) and confirmed by 1 H and 13 C NMR spectroscopy and elemental analysis, which are described in the experimental part.

Antimicrobial Activity
The synthesized compounds (Table 1) were evaluated for their antibacterial activity against a panel of eight species using a microdilution method with ampicillin and streptomycin as reference drugs. The antibacterial activity of tested compounds was good to very good with MIC/MBC values ranging from 0.004 mg/mL to 0.045 mg/mL and Scheme 1. Preparation of compounds 1-17 Reagents and conditions: CH 3 COONH 4 , CH 3 COOH, reflux 2 h.

S2
). In addition, ADMET Predictor 10.4 (Simulation Plus) was used to investigate other critical toxicity issues such as phospholipidosis and hERG-related cardiotoxicity (Table  S1).
It is known that toxic doses are often given as LD50 values in mg/kg body weight, where LD50 represents the median lethal dose upon exposure to a compound. Toxicity classes are defined according to the globally harmonized system of classification of the labeling of chemicals (GHS), that is, Class I: fatal if swallowed (LD50 ≤ 5); Class II: fatal if swallowed (5 < LD50 ≤ 50); Class III: toxic if swallowed (50 < LD50 ≤ 300); Class IV: harmful if swallowed (300 < LD50 ≤ 2000); Class V: may be harmful if swallowed (2000 < LD50 ≤ 5000); and Class VI: non-toxic (LD50 > 5000). In our preliminary investigation, it was estimated (54-66% accuracy) that all examined compounds are classified in category IV, indicating a promising profile towards oral rat acute toxicity. Furthermore, our compounds did not predict to demonstrate adverse outcomes (Table S2).

Chemistry
Compounds were synthesized according to the process described in detail in our previous paper [41] and presented in Scheme 1. The structures of all synthesized compounds are illustrated in Table 1 (Table S1) and confirmed by 1 H and 13 C NMR spectroscopy and elemental analysis, which are described in the experimental part.  The synthesized compounds (Table 1) were evaluated for their antibacterial activity against a panel of eight species using a microdilution method with ampicillin and streptomycin as reference drugs. The antibacterial activity of tested compounds was good to very good with MIC/MBC values ranging from 0.004 mg/mL to 0.045 mg/mL and 0.008 mg/mL to 1.2 mg/mL, respectively, presented in Table 2. The order of activity can be presented as: 8 > 11 > 2 > 1 > 12 > 3 > 17 > 7 > 5 > 13 > 14 > 16 > 9 > 4 = 6 > 15 > 10. The best activity was expressed by compound 8 with MIC at 0.004-0.03 mg/mL and MBC at 0.008-0.06 mg/mL, while compound 10 was the least active one. Compounds 1, 2, and 3 exhibited good activity against B. cereus with MIC at 0.015 mg/mL. The same good activity was observed for compounds 2, 3, 5, 6, and 7 against S. aureus; 8 and 12 against L. monocytogenes; and compounds 2-6 and 12 against En. Cloacae. Compound 12, together with compounds 1, 3, 7, and 11 (MIC 0.015 mg/mL), showed good activity against S. typhimurium, while derivatives 1 and 11 showed good activity against E. coli. On the other hand, compounds 8 and 12 showed excellent activity with MIC at 0.004 mg/mL against En. cloacae and E. coli, respectively. Very good activity (MIC 0.008 mg/mL) was observed for compounds 11 and 17 against B. cereus and S. aureus, respectively. At the same time, compound 11 exhibited good activity with MIC at 0.011 mg/mL against En. cloacae and P. aeruginosa. It should be mentioned that all compounds were more potent than both reference drugs against all bacteria tested. The most sensitive bacteria appeared to be En. cloacae, while E. coli was the most resistant one, followed by M. flavus.  The structure-activity relationship revealed that the presence of 3-methylbutanoic acid as a substituent on the nitrogen of the 2-thioxothiazolidin-4-one moiety of (Z)-2-(5-((5-fluoro-2-(methoxycarbonyl)-1H-indol-3-yl)methylene)-4-oxo-2-thioxothiazolidin-3-yl)-3-methylbutanoic acid, as well as methylformate in the indole ring (8), is beneficial for antibacterial activity. The replacement of 3-methylbutanoic acid by an acetic acid substituent on the nitrogen of 2-thioxothiazolidin-4-one ring, as well as the removal of the F-substituent at position 5 of the indole ring, decreased a little in activity (11), while the presence of benzoic acid at the same position led to a slightly less active compound 2. The removal of F-substituent at position 5 on the indole ring from compound 8 led to compound 12, which is fifth in the order of activity, while the presence of 4-hydroxybenzene substituent on the nitrogen of 2-thioxothiazolidin-4-one moiety appeared to be detrimental for antibacterial activity. Thus, the structure-activity relationship studies revealed that the activity of compounds depends not only on the substituent at the 2-thioxothiazolidin-4-one ring but as well as at the indole ring.

Antifungal Activity
The antifungal activity of compounds was tested against eight fungal species using the drugs ketoconazole and bifonazole as reference. The antifungal activity of tested compounds was good to excellent with MIC in the range of 0.004-0.06 mg/mL and MFC at 0.008-0.12 mg/mL, presented in Table 3. The order of activity of tested compounds can be presented as follows: 15 > 3 > 16 > 10 > 7 > 6 > 2 > 5 > 11 > 9 > 13 > 4 > 12 > 17 > 14 > 1 > 8. Thus, the best activity was achieved by compound 15 with MIC/MFC at 0.008-0.015/0.015-0.03 mg/mL, respectively, while compound 8 was the least active. The opposite was observed in the case of antibacterial activity, where compound 8 was the most active and 15 was one of the least active. Some compounds showed excellent activity against some fungal species being more potent than both reference drugs. Thus, compound 1 expressed activity with MIC 0.004 mg/mL against P. ochramensis, while 3, 5, 7, and 10 expressed activity against T. viride. At the same time, compounds 10 and 9 showed the same good activity against A. ochraceus. Most of the compounds demonstrated very good activity against A. ochraceus with MIC at 0.008 mg/mL (1, 2, 4-8, 11), while compounds 4, 6, 9, 11, 12, 15, and 16 were also very good against T. viride. Furthermore, compounds 6, 7, 10, and 15 were also very potent against P. ochramensis, whereas 3 7, 10, and 15 also demonstrated very good activity against A. niger and P. funiculosum, respectively. Finally, compound 15 appeared to be very active also against P. cyclpoium var verucosum after A. fumigatus, the most resistant fungi. T. viride was found to be the most sensitive one.
The structure-activity relationship studies revealed that the presence of the methyl group substituent on the nitrogen of 2-thioxothiazolidin-4-one (15) moiety of (Z)-methyl 5- fluoro-3-((3-methyl-4-oxo-2-thioxothiazolidin-5-ylidene)methyl)-1H-indole-2-carboxylate is favorable for antifungal activity. The replacement of the methyl group by 4-hydroxybenzene (3) and the removal of the F-substituent of the indole ring slightly decreased the activity. The introduction of morpholine as a substituent instead of methyl in compound 15 led to a less active compound (16), nevertheless being among the third most active compounds, while the removal of the fluorine atom from the indole ring resulted in compound 6 being sixth in order of activity.
Finally, the presence of 3-methylbutanoic acid as a substituent at the nitrogen of 2thioxothiazolidin-4-one moiety (8) was detrimental to antifungal activity, while showing the best antibacterial activity among all tested compounds. Thus, the antifungal activity of compounds, as in the case of antibacterial, depends not only on substituent at 2-thioxothiazolidin-4-one but also at the indole ring as well.

Docking Studies 2.4.1. Docking to Antibacterial Targets
In order to predict the possible mechanism of activity of synthesized compounds, docking studies in different targets were carried out. In this direction, for docking studies, we used enzymes responsible for the most common mechanisms of activity of antibacterial agents, such as E. coli DNA gyrase, thymidylate kinase, E. coli primase, and E. coli MurA and E. coli MurB enzymes.
A low Free Energy of Binding represents a strong binding of ligand to the enzyme. Taking this into account, docking studies revealed that the Free Energy of Binding of all tested compounds to E. coli DNA gyrase, thymidylate kinase, E. coli primase, and E. coli MurA enzymes was higher than that of E. coli MurB (−7.54-10.88 kcal/mol). Therefore, it may be suggested that the inhibition of E. coli MurB is probably the most suitable mechanism of action of the compounds where binding scores were consistent with biological activity (Table 4). The docking pose of the most active compound 8 in the E. coli MurB enzyme showed two favorable hydrogen bond interactions. The first one is between the oxygen atom of the carbonyl group of the compound and the hydrogen of the side chain of Ser229 (distance 3.11 Å), and the other one is between the S atom of thiazolidinone moiety and Ser50 residue (distance 3.64 Å). Moreover, hydrophobic interactions between residues Val52, Arg159, Ile110, and the compound were detected, contributing to the stability of the complex ligandenzyme ( Figure 4). It is noteworthy to highlight that the hydrogen bond with the residue Ser229 is crucial for the inhibitory action of this compound because this residue takes part in the proton transfer at the second stage of peptidoglycan synthesis [50]. Hydrogen bond interactions with the residue Ser229 were also observed for most of the compounds ( Table 4).
The docking pose of the most active compound 8 in the E. coli MurB enzyme showed two favorable hydrogen bond interactions. The first one is between the oxygen atom of the carbonyl group of the compound and the hydrogen of the side chain of Ser229 (distance 3.11 Å), and the other one is between the S atom of thiazolidinone moiety and Ser50 residue (distance 3.64 Å). Moreover, hydrophobic interactions between residues Val52, Arg159, Ile110, and the compound were detected, contributing to the stability of the complex ligand-enzyme ( Figure 4). It is noteworthy to highlight that the hydrogen bond with the residue Ser229 is crucial for the inhibitory action of this compound because this residue takes part in the proton transfer at the second stage of peptidoglycan synthesis [50]. Hydrogen bond interactions with the residue Ser229 were also observed for most of the compounds (Table 4). Moreover, detailed analysis of the docking pose of the most active compounds showed that they bind MurB in a similar way to FAD and they fit into the binding center of the enzyme in the same way, interacting with the same residues, such as Ser50, Arg213, Arg158, and Ser229 ( Figure 4A). This is probably the reason why these compounds showed good inhibitory activity comparable to that of ampicillin. Moreover, detailed analysis of the docking pose of the most active compounds showed that they bind MurB in a similar way to FAD and they fit into the binding center of the enzyme in the same way, interacting with the same residues, such as Ser50, Arg213, Arg158, and Ser229 ( Figure 4A). This is probably the reason why these compounds showed good inhibitory activity comparable to that of ampicillin.

Docking to Antifungal Targets
All the synthesized compounds and the reference drug ketoconazole were docked to lanosterol 14α-demethylase of C. albicans and DNA topoisomerase IV (Table 5) in order to explore the possible mechanism of antifungal activity of compounds.
Docking studies showed that all compounds can bind the CYP51 Ca enzyme in an analogous mode to that of ketoconazole. Compound 15 is placed inside the enzyme by the side of the heme group, interacting hydrophobically aromatically and binding to the Fe of the heme. Furthermore, compound 15 forms one H-bond between the oxygen atom of the CO 2 group and the side-chain hydrogen of Thr311. Hydrophobic interactions were detected between residues Ile304, Leu300, and Ile131 and the benzene ring of the compound 15 ( Figure 5B). Interaction with the heme group was also observed with the benzene ring of ketoconazole, which forms aromatic and hydrophobic interactions ( Figure 5). However, compound 15 formed a more stable complex of the ligand with the enzyme, indicating their interaction with the Fe of heme. This is probably the reason why compound 15 ( Figure 5A) has better antifungal activity than ketoconazole ( Figure 6). Table 5. Molecular docking free binding energies (kcal/mol) to antifungal targets.

Drug-Likeness and Bioavailability
The drug-likeness and bioavailability scores of all tested compounds are shown in Table 6. According to prediction, the bioavailability score of most of the compounds was about 0.55, except for compounds 2, 9, 12, 14, and 17 with a value of 0.11. According to the BOILED-Egg illustration ( Figure 7A), all compounds are predicted to demonstrate moderate to high gastrointestinal (GI) absorption. Especially, compounds 5, 6, 7, 15, and 16 are predicted to be passively absorbed by the gastrointestinal tract. In addition, none of the examined compounds are predicted to passively permeate the blood-brain barrier. All compounds showed no violation of Lipinski's rule of five. Half of the compounds demonstrated a TPSA value < 140 Å, thus indicating their good oral bioavailability. Furthermore, all compounds displayed moderate drug-likeness scores ranging from −0.95 to 0.38, probably due to the enhanced TPSA values and the absence of a basic moiety on rhodanine's nitrogen. The best in silico prediction was achieved for compound 5, bearing the basic propyl-morpholine group with a drug-likeness score of 0.24 (Figure 7, Table 6).

Drug-Likeness and Bioavailability
The drug-likeness and bioavailability scores of all tested compounds are shown in Table 6. According to prediction, the bioavailability score of most of the compounds was about 0.55, except for compounds 2, 9, 12, 14, and 17 with a value of 0.11. According to the BOILED-Egg illustration ( Figure 7A), all compounds are predicted to demonstrate moderate to high gastrointestinal (GI) absorption. Especially, compounds 5, 6, 7, 15, and 16 are predicted to be passively absorbed by the gastrointestinal tract. In addition, none of the examined compounds are predicted to passively permeate the blood-brain barrier. All compounds showed no violation of Lipinski's rule of five. Half of the compounds demonstrated a TPSA value < 140 Å, thus indicating their good oral bioavailability. Furthermore, all compounds displayed moderate drug-likeness scores ranging from −0.95 to 0.38, probably due to the enhanced TPSA values and the absence of a basic moiety on rhodanine's nitrogen. The best in silico prediction was achieved for compound 5, bearing the basic propyl-morpholine group with a drug-likeness score of 0.24 (Figure 7, Table 6).

ADMET Properties
All compounds have been assessed for their ADMET profile in ADMET Predictor version 10.4 provided by the Simulation Plus software package [1-53]. We have found

ADMET Properties
All compounds have been assessed for their ADMET profile in ADMET Predictor version 10.4 provided by the Simulation Plus software package [1-53]. We have found that  compounds 1, 8, 9, 11, 12, 14, and 17 demonstrate optimal properties regarding distribution and, especially, absorption (Table 7). Although these compounds have moderate permeability indices due to their carboxylate moiety, they showed low potential to penetrate the blood-brain barrier (decreased logBB values) as well as acceptable water and salt solubility at a normal blood pH value of 7.4. Moreover, we believe that these compounds could be well absorbed orally after food in the small intestine based on the favorable solubility values in fed state simulated intestinal fluid (FeSSIF). On the other hand, they exhibit low fraction unbound values (less than 6.0%) in comparison with other derivatives in this series (e.g., neutral or basic analogs). Nevertheless, the values of volume of distribution (Vd) and blood-to-plasma ratio (RBP) in humans were predicted to be satisfactory for all compounds in this series.  (µg/mL); d water solubility in mg/mL at pH 7.4; e solubility expressed in mg/mL in fast state simulated gastric fluid; f solubility expressed in mg/mL in fast state simulated intestinal fluid; g solubility expressed in mg/mL in fed state simulated intestinal fluid; h possibility to penetrate blood-brain barrier; i logarithm of the brain/blood partition coefficient; j human %fraction unbound; k human volume of distribution in steady state (L/kg); l blood-to-plasma ratio in humans.
Furthermore, it has been predicted that the clearance mechanism of all compounds is metabolism (possibility of 74-99%) rather than hepatic uptake or renal elimination (99% both). In view of this fact, a comprehensive in silico analysis of cytochrome P450-mediated metabolism of our compounds was executed ( Table 8). First of all, we have found that the different groups incorporated on the nitrogen atom of the rhodanine scaffold can alter compounds' metabolism. Among the most well-absorbed compounds, 8 and 12 have been assessed to show high clearance through CYP2C9 isoenzyme; thus, they exhibit a CYP risk greater than zero. In addition, compounds 1, 9, 11, 14, and 17 display zero possibilities of CYP risks. Therefore, they show optimal metabolism characteristics for further studies. We have also estimated that compounds 1, 2, 8-12, 14, and 17 are not CYP2E1 substrates (67-87%), whereas compounds 3-7, 13, 15, and 16 are predicted to be good CYP2E1 substrates. Moreover, the majority of compounds may inhibit CYP3A4 isoenzyme (33-80%) except for compounds 1, 2, 8, 11, and 12 (71-75%). Despite their differences in metabolism characteristics, all compounds share some common characteristics. Particularly, all compounds may inhibit CYP1A2 (68-97%) but not the isoenzymes of CYP2C9 (62-95%), CYP2C19 (94-99%), and CYP2D6 (95%). As illustrated in Table 8, it has been estimated that the major metabolizing enzymes for compounds 1, 9, 11, 14, and 17 are CYP2C9 and CYP2C19 (over 90% both). On the other hand, all compounds may be also metabolized through CYP2C8 (70-91%) rather than CYP2A6 (67-99%) and CYP2B6 (57-98%). In addition, these compounds may be subjected to glucuronidation by UDP-glucuronosyltransferases 1A3 and 1A9. Last but not least, we suggest that compounds 1, 9, 11, 14, and 17 may exhibit the most promising metabolism profiles according to the low intrinsic clearance values concerning cytochrome P450 metabolism (CYP-CLint) as well as molecule-level hepatic clearance in humans (HEP-CLint). We have also predicted the metabolic pathways that will take place for the most potent compounds with optimal metabolism profiles, e.g., compounds 1 ( Figure 8) and 11 ( Figure 9). For both compounds, two main routes of oxidation and demethylation have been found. In particular, demethylated carboxylic acid derivatives account for 10% and 6% of the metabolisms of compounds 1 and 11, respectively. It is obvious that CYP2C19 and, even more, CYP2C9 are converting the free esters to carboxylic acid derivatives with similar yields. During oxidation, the 6-hydroxy-indole derivatives are also formed in almost equal amounts of 27% and 22% of the metabolic pathway of compounds 1 and 11, respectively. In addition, it has been shown that CYP2C9 is the major isoenzyme that may be responsible for the sulfone formation of these compounds. Thus, M4 is formed in 27% of the metabolism of compound 1 with regard to M2, which occurs in 18% of compound 11. Furthermore, it was also assumed that the main metabolites of these compounds could be the 2,4-thiazolidinones, namely 1-M3 and 11-M1, with surprising values of 36% and 53%, respectively. On the other hand, CYP2C8 has also an impact on the metabolism of compound 1 since it may cleave the acidic tail from the nitrogen atom of the rhodanine group, leaving a second metabolite of 6-oxohexanoic acid to be formed.
However, this procedure has been established to occur only in compound 1 and not in the case of compound 11, which possesses the short acetic acid chain. Last but not least, it is worth mentioning that all metabolites of compounds 1 and 11 which have been estimated during at least three metabolic cycles have been found to be non-toxic.   We further studied in detail the possibility of our compounds being substrates or inhibitors against selected human transporters, which play a critical role in pharmacokinetic properties, especially distribution and excretion ( Table 9). The ADMET Predictor software Pharmaceuticals 2023, 16, x FOR PEER REVIEW 18 of 25 Figure 10. Cell viability levels (%) of MRC-5 cells exposed for 48h at various concentrations in culture to compounds 1, 2, 8, and 11. All values are presented as mean ± standard deviation (SD) of triplicate independent cellular experiments. With asterisk (*), the significant differences between each concentration and the control group are presented. The significance level was determined at p < 0.05.

Prediction of Toxicity
The prediction of toxicity was performed using Protox II webserver [55].
Chemical shifts of nuclei 1 H were measured relatively in the residual signals of deuteron solvent (δ = 2.50 ppm). Coupling constants (J) are reported in Hz. Τhe assignment was based on 2D NMR techniques. Melting points were determined using the Fisher-Johns Melting Point Apparatus (Fisher Scientific, Hampton, NH, USA) and are uncorrected. Elemental analysis was performed by the classical method of microanalysis.
All compounds were synthesized analogous to the process described in our previous paper [6].
A mixture of a corresponding amino acid (50 mmol), a cooled solution of KOH in water (20 mL) (150 mmol in case of dicarbonic acids), and CS2 (mmol) were stirred in a flat-bottomed flask until a solution was formed. A solution of monochloracetic acid (55 mmol) was added with stirring, pre-neutralized with sodium bicarbonate (55 mmol) in water (25 mL), and left at room temperature for 2 days.
Then, to the formed solution, a 6N HCl solution (20 mL) was added and heated to boiling and kept at a slow boil for 1 h. After cooling, the precipitate formed was filtered off, dried, and recrystallized, alternately, from diluted acetic acid, ethanol, and toluene.
In a round-bottom flask equipped with a reflux condenser, 2.5 mmol of methyl 3-formyl-5,6-disabstituted-1H-indole-2-carboxylate, 3.3 mmol of 3-substituted-2-thioxothiazolidin-4-one, 2.5 mmol of ammonium acetate, and 5 mL of acetic acid were placed. The reaction mixture was boiled for 2 h and cooled. Then, the precipitate was filtered off, washed with acetic acid and water, dried, and recrystallized.   Figure 10. Cell viability levels (%) of MRC-5 cells exposed for 48h at various concentrations in culture to compounds 1, 2, 8, and 11. All values are presented as mean ± standard deviation (SD) of triplicate independent cellular experiments. With asterisk (*), the significant differences between each concentration and the control group are presented. The significance level was determined at p < 0.05.

Prediction of Toxicity
The prediction of toxicity was performed using Protox II webserver [55].
Chemical shifts of nuclei 1 H were measured relatively in the residual signals of deuteron solvent (δ = 2.50 ppm). Coupling constants (J) are reported in Hz. The assignment was based on 2D NMR techniques. Melting points were determined using the Fisher-Johns Melting Point Apparatus (Fisher Scientific, Hampton, NH, USA) and are uncorrected. Elemental analysis was performed by the classical method of microanalysis.
All compounds were synthesized analogous to the process described in our previous paper [6].
A mixture of a corresponding amino acid (50 mmol), a cooled solution of KOH in water (20 mL) (150 mmol in case of dicarbonic acids), and CS 2 (mmol) were stirred in a flat-bottomed flask until a solution was formed. A solution of monochloracetic acid (55 mmol) was added with stirring, pre-neutralized with sodium bicarbonate (55 mmol) in water (25 mL), and left at room temperature for 2 days.
Then, to the formed solution, a 6N HCl solution (20 mL) was added and heated to boiling and kept at a slow boil for 1 h. After cooling, the precipitate formed was filtered off, dried, and recrystallized, alternately, from diluted acetic acid, ethanol, and toluene.

Docking Studies
AutoDock 4.2 ® software was used for the in silico studies, and a detailed procedure is reported in our previous paper [58].

Drug Likeness
Five filters were used to predict drug-likeness [59] by the Molsoft software and Swis-sADME program (http://swissadme.ch, accessed on 25 October 2022) via the ChemAxon's Marvin JS structure drawing tool.

Assessment of Cytotoxicity
The normal human fetal lung fibroblast MRC-5 cell line was maintained and used in our laboratory (Dr. I.s.Vizirianakis, Laboratory of Pharmacology, School of Pharmacy, Aristotle University of Thessaloniki, Greece) (passage 15 < 40), as previously published [60]. Cells were grown in culture under the following conditions: 37 • C, humidified atmosphere containing 5% v/v CO 2 , in DMEM (Dulbecco's modified Eagle Medium) enriched with 10% FBS and 1% PS. The compounds tested were dissolved in DMSO and stored at 4 • C. For the assessment of cytotoxicity, the cells were cultivated in a 96-well plate at an initial concentration of 5 × 10 4 cells/mL and allowed to attach for 20 h before the addition of the compounds at concentrations of 1 × 10 −5 M (10 µM), 1 × 10 −6 M(1 µM), and 1 × 10 −7 M (0.1 µM). It is worth noting that the concentration of DMSO in the culture was 0.02% v/v, in which no detectable effect on cell toxicity was observed. To assess the cytotoxicity of each compound, the cells were allowed to grow for 48 h under the effect of each substance. Subsequently, Cell Counting Kit-8 (CCK8, St. Louis, MO, USA, Sigma-Aldrich) reagent was added to each well and incubated for 4 h at 37 • C. After the incubation, the OD for each well was determined at 450 nm in a multifunction microplate reader. Wells containing only the CCK-8 reagent were used as blank control. Data are presented as mean ± standard deviation (SD) of triplicate incubations. Statistical t-test and one-way analysis of variance (ANOVA) were performed via the use of the SPSS program and the significance level was determined at p < 0.05.
The evaluation revealed that all compounds were more potent than both reference drugs, ampicillin, and streptomycin against all bacteria tested. En. cloacae appeared to be the most sensitive bacterial strain towards our derivatives, whereas E. coli was the most resistant one, followed by M. flavus.
Concerning antifungal action, the tested compounds exhibited very good to excellent activity against all the fungal species tested, being more active than ketoconazole and bifonazole. Most of the compounds appeared to be very potent against A. ochraceus and T. viride with the last one being the most sensitive to tested compounds. Filamentous A. fumigatus was the most resistant strain.
It can be observed that the growth of both Gram-negative and Gram-positive bacteria and fungi responded differently to the tested compounds, suggesting that different substituents may lead to different modes of action or that the metabolism of some bacteria/fungi was able to overcome the effect of the compounds or adapt to it.
Docking analysis to DNA Gyrase, Thymidylate kinase, and E. coli MurB indicated MurB inhibition as a putative antibacterial mechanism of compounds tested, while docking analysis to 14α-lanosterol demethylase (CYP51) and tetrahydrofolate reductase of Candida albicans pointed out a probable implication of CYP51 reductase in the antifungal activity of the compounds.
Even though compounds displayed moderate to good drug-likeness scores (−0.89 to +0.24), no violation of the Lipinski rule was observed. Furthermore, according to predicted results, six out of seventeen compounds can be orally absorbed since their TPSA are less than 140 Å. All compounds have been assessed for their ADMET profile in ADMET Predictor version 10.4, provided by the Simulation Plus software package. We have found that compounds 1, 8, 9, 11, 12, 14, and 17 demonstrate optimal properties regarding distribution and especially absorption.
The prediction of metabolic pathways that could take place for the potent compounds 1 and 11 with optimal metabolism profiles indicated the transformation of rhodanine to thiazolidinone and indole ring hydroxylation as the two main routes. The evaluation of the cytotoxicity of compounds on normal human fetal lung fibroblast MRC-5 cell lines revealed that compounds are not toxic.
Thus, these derivatives can be considered as lead compounds for the development of novel potent and safe antimicrobial agents.