2.1 DprE1 Preparation

Gene–gene and protein–protein interaction networks (Figure S1, Supporting Information) observations underscore the central importance of DprE1 in the metabolic network of M. tuberculosis, providing additional support for its potential as a promising drug target. The 3D arrangement of M. tuberculosis DprE1, covalently linked to PBTZ169, was acquired from the Protein Data Bank database, featuring a resolution of 1.9 Å (PDB accession no. 4NCR).^[27] Using Modeller software, the non-terminal missing residues were modeled to complete the DprE1 structure.^[29] To streamline the structure, all extraneous components, such as water molecules, heteroatoms, ions, and co-crystallized ligand, were eliminated. Furthermore, polar hydrogens were incorporated into the structure at a pH of 7, employing the H++ web server.^[30]

2.2 Covalent Ligands Preparation

In an effort to identify the most potent covalent inhibitors targeting DprE1, 63 nitro-containing compounds were collected from the StreptomeDB database in SDF format.^[28] Employing the Omega2 software, 3D representations of each StreptomeDB compound were crafted.^[31, 32] Afterward, the generated structures underwent minimization employing the MMFF94S force field within the SZYBKI software.^[33, 34] Atomic charges for the StreptomeDB compounds were determined via the Gasteiger–Marsili method.^[35] Tautomer enumeration of the StreptomeDB compounds was carried out using the tautomer application implemented by means of the QUACPAC software package.^[36] Finally, the FixpKa program, functioning within QUACPAC, was utilized to predict the most probable molecular ionization status.

2.3 Covalent Docking Protocol

As a part of this in-silico study, covalent docking calculations were conducted with the AutoDock4.2.6 software,^[37] leveraging the flexible side chain method.^[38] In this method, the ligand is overlapped with the protein residue, and subsequently used AutoDock to optimize its conformation.^[38] To comply with the AutoDock4.2.6 protocol, the pdb file of DprE1 was converted to a pdbqt file.^[39] Concerning the covalent docking parameters, the default values were maintained, apart from the number of genetic algorithm (GA) runs, which was set to 250, and the maximum number of energy evaluations (eval), which was adjusted to 25 000 000. To ensure accurate binding predictions, a grid box was defined to encompass the DprE1 binding site. The grid box with dimensions of 40 Å × 40 Å × 40 Å was carefully chosen, and centered at coordinates x = 17.176, y = −20.119, and z = 1.875. A grid spacing of 0.375 Å was employed.

2.4 Molecular Dynamics Simulations (MDS)

The most promising StreptomeDB compounds in complex with DprE1 underwent MDS to evaluate their long-term stability using AMBER20 software.^[40] Technical details regarding the employed MDSs can be found in Ref. [^41-45] Succinctly, the GAFF2 (General AMBER force field) was applied to describe the investigated StreptomeDB compounds.^[46] However, DprE1 was described by relying on the FF14SB AMBER force field.^[47] To facilitate atomic charge estimates, N-terminal acetyl, and C-terminal methyl amide groups were added to cap the irreversible covalent inhibitors with CYS387 residue. Subsequently, geometric optimization was performed using the Gaussian09 software.^[48] The B3LYP/6-31 G* level of theory was employed for optimization. The atomic charges of the StreptomeDB compounds were then computed employing the restrained electrostatic potential (RESP) fitting technique at the HF/6-31 G* level.^[49] An antechamber module, an integral component of the AMBER package, was applied to define the parameters and the atom types of covalent inhibitors with the CYS387 residue. Following that, the ligand-DprE1 complexes were dissolved in a truncated octahedron box filled with TIP3P water molecules, maintaining an average distance of 12 Å from the solute edge. To neutralize the solvated complexes, the appropriate quantity of Na⁺/Cl⁻ ions was added. Moreover, 0.15 M NaCl was incorporated to preserve physiological ionic conditions. The solvated complexes then underwent energy minimization for 5000 cycles. After minimization, the ligand-DprE1 complexes were heated incrementally to 310 K over 50 ps, and then an equilibration simulation of 10 ns was performed under NPT conditions. Subsequently, production MDS of the equilibrated complexes were run over 5, 25, 50, 100, and 300 ns, recording snapshots at intervals of 10 ps. Bonds involving hydrogens were restricted utilizing the SHAKE algorithm with a time step of 2 fs.^[50] Additionally, the long-range electrostatic forces were handled utilizing the particle mesh Ewald (PME) method.^[51] Lennard-Jones interactions were assessed with a 12 Å cutoff. Throughout the production steps, an NTP ensemble was sustained with a Langevin thermostat (310 K) and an anisotropic Berendsen barostat (1 atm).^[52, 53] The PMEMD.CUDA GPU, as implemented in the AMBER20 software, was utilized to carry out MDS. Ultimately, graphical representations of the investigated StreptomeDB compounds were generated with the help of the BIOVIA Materials Studio software.^[54]

2.5 MM-GBSA Binding Affinity

For determining the binding energies (ΔG_binding) of the top StreptomeDB compounds complexed with DprE1, a molecular mechanics-generalized Born surface area (MM-GBSA) approach was implemented.^[55] Due to its lower computational cost and time, the MM-GBSA approach was selected for binding energy calculations over other approaches, such as molecular mechanics-Poisson Boltzmann surface area (MM-PBSA).^[56] The energy contribution from polar solvation was evaluated using the generalized Born (GB) model (igb = 2) proposed by Onufriev et al..^[57] The corresponding binding energies (ΔG_binding) were estimated relying on the uncorrelated snapshots compiled from MDS, using the following equation

()

In which the G expression is

()

()

()

E_MM stands for the energy of molecular mechanics (MM) in the gas phase. G_solv denotes the solvation energy. E_int signifies the internal MM energy, involving bond (E_bond), angle (E_angle), and dihedral (E_torsion) energies. Electrostatic interactions and van der Waals forces are represented by E_ele and E_vdW, respectively. A single-trajectory technique was employed to obtain the spatial coordinates for the covalent inhibitors, DprE1, and the ligand-DprE1 complexes. For all examined StreptomeDB compounds, the consideration of entropy contribution (S) was avoided owing to its expensive computational cost.^[58, 59]

2.6 Physicochemical Features

The molecular characteristics have a crucial role to play in shaping the effectiveness of therapeutic medications, highlighting their profound impact on treatment outcomes.^[60] The drug-likeness features were evaluated utilizing an online SWISS-ADME package (http://www.swissadme.ch). For each investigated StreptomeDB compound, the following molecular properties were estimated: molecular weight (MW); H-bond donors (HBD); MlogP (n-octanol/water partition coefficient); H-bond acceptors (HBA); and topological polar surface area (TPSA).

2.7 ADMET

Evaluating the ADMET characteristics, including chemical absorption, distribution, metabolism, excretion, and toxicity, is pivotal in ascertaining the therapeutic efficacy of prospective drug candidates.^[61] To expedite the evaluation process and minimize the risk of drug failure, the pkCSM server (http://biosig.unimelb.edu.au/pkcsm/prediction) was employed for the prediction of ADMET features for the identified StreptomeDB compounds. The absorption (A) property was assessed by anticipating the Caco-2 permeability and skin permeability (log Kp). The distribution (D) characteristic was evaluated by investigating the values of the logarithm of the log BB (blood-brain barrier) partition coefficient and CNS (central nervous system) permeability. Using cytochromes P450 (CYPs) models, predictions regarding metabolism (M) were conducted. Estimation of the excretion (E) property was made using the total clearance of the inhibitor. Furthermore, the toxicity (T) was examined by considering parameters such as AMES toxicity, maximum tolerated dose (MTD), hERG inhibition, and skin sensitization.

3.1 Covalent Docking Assessment

In an effort to ensure the reliability of the employed covalent docking protocol using AutoDock4.2.6 software, a validation process was conducted. First, the co-crystalized PBTZ169 (macozinone) inhibitor was prepared and re-docked against the DprE1. The RMSD between the co-crystalized and the predicted binding modes was estimated and found to be 0.73 Å (Figure 2).^[62, 63]

Furthermore, the covalent docking computations demonstrated a favorable docking score of −7.8 kcal mol⁻¹ for PBTZ169 complexed with DprE1. The strong docking score likely results from the formation of an irreversible covalent bond between the nitro group and the sulfur atom of CYS387 (1.16 Å), accompanied by the establishment of two H-bonds (Figure 2). More exactly, an H-bond formed between the oxygen atom of the thiazine-4-one and the NH₃ group of LYS134 (2.52 Å). In addition, the NO₂ group established an H-bond with the NH₃ group of LYS418 (2.00 Å). Besides hydrogen bonding, an array of other interactions was detected, including C–H bonds with HIS132, GLY117, and SER228 residues (Figure 2). As well, interactions such as halogen, π-sigma, alkyl, and π-alkyl were found between PBTZ169 and GLY133, VAL365, LYS367, and HIS132 residues, respectively (Figure 2). These insightful findings highlight the exceptional proficiency of AutoDock4.2.6 in accurately forecasting the ligand-DprE1 binding mode, solidifying its application in the subsequent calculations.

3.2 Screening of StreptomeDB Database

At the outset of drug discovery, virtual screening emerges as a powerful technique for rapidly identifying potential bioactive inhibitors.^[64, 65] To this end, an extensive covalent docking computation was conducted to screen a set of 63 nitro-containing compounds sourced from the StreptomeDB database against DprE1. The covalent docking study was performed with specific parameters. Specifically, the GA value was adjusted to 250, and the eval value was fixed at 25 000 000. Presented in Table S1, Supporting Information are the computed covalent docking scores for the investigated StreptomeDB compounds. Relying on the covalent docking scores, just 30 StreptomeDB compounds demonstrated docking scores lower than the calculated score of PBTZ169 (−7.8 kcal mol⁻¹). Depicted in Figure S2, Supporting Information are the molecular interactions of the potent compounds complexed with DprE1, represented in both 2D and 3D formats. Valuable insights into the 2D chemical structures, estimated covalent docking scores, and binding features of the most promising StreptomeDB compounds with DprE1 can be found in Table 1. It is pertinent to mention that the hydroxythaxtomin A (SDB12202) and lajollamycin B (SDB9226) were chosen relying on the estimated MM-GBSA binding energy throughout a 300 ns MDS, as will be elaborated in the following sections.

Table 1. Computed covalent docking scores (in kcal mol⁻¹), 2D chemical structures, and binding features of PBTZ169 and the most promising StreptomeDB compounds inside the DprE1 binding pocket.

No.	StreptomeDB Code/Namea)	2D Chemical Structure	Covalent Docking Score [kcal mol−1]	Binding Featuresb)
	PBTZ169 (Macozinone)		− 7.8	LYS134 (2.52 Å), LYS418 (2.00 Å)
1	SDB12202 (Hydroxythaxtomin A)		− 14.2	HIS132 (2.19 Å), ASP318 (2.33 Å), GLN336 (2.16 Å), TYR415 (2.09, 2.20 Å)
2	SDB9226 (Lajollamycin B)		− 13.9	ARG58 (2.00 Å), HIS132 (1.87 Å), TYR327 (2.80 Å), ALA330 (2.11 Å), GLN336 (2.32 Å), LYS418 (1.83 Å)

• a) The potent StreptomeDB compounds ranked relying on the covalent docking score.
• b) Only conventional H-bonds (in Å) are presented.

Upon analyzing the data presented in Table 1 and Figure S2, Supporting Information, it was observed that a significant proportion of the investigated StreptomeDB compounds exhibited remarkably similar docking poses within the DprE1 binding pocket. These identified StreptomeDB compounds demonstrated the ability to form an irreversible covalent bond with CYS387 and many hydrogen bonds with key residues such as HIS132, GLN336, and LYS418. Furthermore, π-based and hydrophobic interactions were prominently indicated between the investigated StreptomeDB compounds and key amino acids within the DprE1 binding site.

To enhance visual comprehension, Figure 3 provides comprehensive 2D and 3D representations of hydroxythaxtomin A and lajollamycin B; the nitro (NO₂) group of the two inhibitors created an irreversible covalent bond with the sulfur atom of CYS387 (1.75 and 1.32 Å, respectively) (Figure 3).

As outlined in Table 1, hydroxythaxtomin A, a herbicide with favorable toxicological properties,^[66] exhibited the most favorable covalent docking score of −14.2 kcal mol⁻¹ toward DprE1. This promising docking score of hydroxythaxtomin A toward DprE1 can be ascribed to the creation of five intermolecular H-bonds with the essential residues of DprE1 (Figure 3). Precisely, the OH group of hydroxy-piperazine-dione exhibited an H-bond with the CO group of HIS132 (2.19 Å). Also, the NH group of the pyrrole established an H-bond with the backbone CO of ASP318 (2.33 Å). The oxygen of the nitro group formed an H-bond with the CO group of GLN336 (2.16 Å). Moreover, the two OH groups of pyrocatechol contributed to two H-bonds with the OH group of TYR415 (2.09, 2.20 Å). These interactions indicate a dual inhibition mechanism. Covalent inactivation occurs through irreversible binding to CYS387, while structural destabilization results from non-covalent interactions with HIS132, ASP318, GLN336, and TYR415. Modification of CYS387 likely disrupts redox chemistry essential for epimerization. HIS132 and TYR415 are crucial for substrate binding, and stable hydrogen bonds with them may obstruct substrate access or proton relay. Additionally, ASP318 and GLN336 may help stabilize an inactive enzyme form. Thus, hydroxythaxtomin A may function as a competitive or mechanism-based inhibitor of DprE1, affecting mycobacterial cell wall biosynthesis.

Lajollamycin B demonstrates potent anti-cancer properties and significant antimicrobial activity, making it a promising candidate for combatting cancer and microbial infections.^[67, 68] Lajollamycin B exhibited the second superior covalent docking score with DprE1, recording a value of –13.9 kcal mol⁻¹. In addition, as outlined in Table 1, lajollamycin B participated in the formation of six H-bonds with key amino acid residues located within the DprE1 binding pocket. In detail, the OH group (i.e., N-(2 E,4 E,6 R,7S,9 R)-6-hydroxy) exhibited an H-bond with the CO group of ARG58 (2.00 Å) (Figure 3). The OH group of hydroxypyrrolidinone demonstrated an H-bond with the CO group of HIS132 (1.87 Å). Besides, the OH group (i.e., (R,4Z,6 E,8 E,10Z)-3-hydroxy) contributed to an H-bond with the NH group of TYR327 (2.80 Å). An H-bond was anticipated between the CO group of 11-nitrododeca-4,6,8,10 tetraenamide and the NH group of ALA330 (2.11 Å) (Figure 3). The oxygen atom and the CO group of oxetanone exhibited two H-bonds with the NH₂ group of the GLN336 (2.32 Å) and the NH₃ group of LYS418 (1.83 Å), respectively. These interactions indicate a potential inhibitory mechanism where lajollamycin B may hinder DprE1’s catalytic function by targeting residues essential for enzymatic activity and stability. Covalent bonding with the nucleophilic CYS387 likely leads to irreversible inactivation. The interaction with HIS132 may obstruct substrate positioning or proton transfer, while GLN336 and LYS418 could limit the conformational flexibility necessary for enzyme activity. Additionally, residues like ARG58, TYR327, and ALA330 may help stabilize an inactive enzyme state. Collectively, these interactions suggest that lajollamycin B irreversibly alkylates the active site cysteine and promotes an inactive conformation, effectively inhibiting DprE1 and disrupting mycobacterial cell wall synthesis.

Overall, hydroxythaxtomin A and lajollamycin B exhibited lower docking scores with DprE1 than PBTZ169 (calc. −7.8 kcal mol⁻¹), suggesting that both investigated compounds have the potential to be DprE1 inhibitors.

3.3 MDS

MDS serves as an effective strategy to validate detailed findings related to protein–ligand interactions with a dynamic aspect.^[69, 70] Herein, MDS was performed on the 30 potent StreptomeDB compounds complexed with DprE1, which were selected based on estimated covalent docking scores less than the calculated score of PBTZ169 (–7.8 kcal mol⁻¹). To streamline computational resources and minimize time requirements, a short MDS of 5 ns was performed, succeeded by calculations of MM-GBSA binding energies (Table S2, Supporting Information). As implied in Table S2, Supporting Information, only 14 StreptomeDB compounds out of the 30 StreptomeDB compounds had ΔG_binding lower than that of PBTZ169 (calc. –37.9 kcal mol⁻¹) with DprE1. To ensure reliable outcomes, these 14 StreptomeDB compounds complexed with DprE1 underwent 25 ns MDS, followed by an estimation of the corresponding binding energies (Table S3, Supporting Information). Notably, Table S3, Supporting Information indicated that 8 compounds demonstrated ΔG_binding less than that of PBTZ169 (calc. –42.9 kcal mol⁻¹), which served as the threshold for identifying potent DprE1 inhibitors. These 8 potent compounds complexed with DprE1 underwent MDS for a longer time of 50 ns, and the corresponding binding energies were compiled and gathered in Table S4, Supporting Information. As enrolled in Table S4, Supporting Information, 5 out of 8 StreptomeDB compounds demonstrated binding energies less than −45.1 kcal mol⁻¹. The MDSs of these StreptomeDB compounds complexed with DprE1 were prolonged to 100 ns. Moreover, the corresponding binding affinities were computed (Table S5, Supporting Information). Intriguingly, from this set of 5 compounds, the binding affinities calculations indicated that lajollamycin B (SDB9226) and hydroxythaxtomin A (SDB12202) exhibited significant binding energies (ΔG_binding) of −50.2 and −48.3 kcal mol⁻¹, respectively, compared to PBTZ169 (calc. –49.8 kcal mol⁻¹) (Table S5, Supporting Information). To further enhance the precision of binding energies, the MDS for the 2 investigated StreptomeDB compounds was elongated to 300 ns. Additionally, the binding energies of these identified StreptomeDB compounds with DprE1 over 300 ns MDS were calculated and displayed in Figure 4. Compared to PBTZ169 (calc. −49.3 kcal mol⁻¹), the estimated MM-GBSA//300 ns MDS of hydroxythaxtomin A and lajollamycin B were −51.2 and −50.5 kcal mol⁻¹, respectively. As per data portrayed in Figure 4, the ΔG_binding of lajollamycin B remained relatively unchanged throughout 100 and 300 ns MDS against DprE1. As indicated by employing long MDS, hydroxythaxtomin A and lajollamycin B are prospective DprE1 inhibitors and thus hold potential as antituberculosis drug candidates.

To examine the binding of hydroxythaxtomin A, lajollamycin B, and PBTZ169, the calculated binding affinities were decomposed into their individual components (Table 2). According to the MM-GBSA binding energy decomposition, hydroxythaxtomin A, lajollamycin B, and PBTZ169 in complex with DprE1 exhibited electrostatic (ΔE_ele) interactions, with an average value of −86.9, −66.7, and −20.4 kcal mol⁻¹, respectively. Furthermore, van der Waals (ΔE_vdW) interactions were crucial in the binding of hydroxythaxtomin A, lajollamycin B, and PBTZ169, exhibiting average values of −47.2, −61.5, and −52.1 kcal mol⁻¹, respectively. Looking at Table 2, hydroxythaxtomin A, lajollamycin B, and PBTZ169 complexed with DprE1 contribute modestly to ΔE_int, with average values of 7.9, 3.2, and 5.3 kcal mol⁻¹, respectively (Table 2).

Toward deeper insight into their binding modes and intermolecular interaction, energy decomposition per-residue was carried out for hydroxythaxtomin A, lajollamycin B, and PBTZ169 with DprE1. A specific focus was placed on amino acids that had energy contributions below −0.5 kcal mol⁻¹ (Figure 5a). Based on the data presented in Figure 5a, LYS134, GLN334, VAL365, and CYS387 residues conferred notable contributions to the binding of hydroxythaxtomin A, lajollamycin B, and PBTZ169 with DprE1. Notably, the investigated complexes displayed remarkably consistent interaction patterns with significant amino acid residues. CYS387 residue played a major role in the overall ΔG_binding, contributing significantly with values of −1.3, −0.8, and −1.1 kcal mol⁻¹ for hydroxythaxtomin A-, lajollamycin B-, and PBTZ169-DprE1 complexes, respectively (Figure 5a).

Additionally, the 2D visualizations of the final snapshot of 300 ns MDS for hydroxythaxtomin A, lajollamycin B, and PBTZ169 with DprE1 are depicted in Figure 5b. Interestingly, those investigated StreptomeDB compounds in complex with DprE1 exhibited strikingly similar interaction patterns with nearby residues, suggesting similarity in their binding modes. Obviously, hydroxythaxtomin A retained the essential H-bonds with HIS132, ASP318, and GLN336 and formed new H-bonds with GLY117, ASN385, and LYS418. Moreover, lajollamycin B preserved the essential H-bonds with ARG58 and GLN336 and exhibited new H-bonds with GLY117, THR118, and PHE320. This highlights the importance of conducting MDS to get more reliable results.

3.4 Post-MD Analyses

The steadiness of hydroxythaxtomin A (SDB12202) and lajollamycin B (SDB9226) complexed with DprE1 was assessed through structural and energetic analyses during 300 ns MDS. Post-MD analyses included the binding energy per frame, CoM distance, RMSD, RMSF, Rg, and the number of H-bonds.

3.4.1 Binding Energy Per Frame

The steadiness of hydroxythaxtomin A and lajollamycin B with DprE1 was examined throughout 300 ns, focusing on the correlation between binding energy and simulation time (Figure 6a). According to data presented in Figure 6a, hydroxythaxtomin A and lajollamycin B demonstrated overall energetic stability over the course of 300 ns MDS with ΔG_binding values of –51.2 and –50.5, respectively, compared to –49.3 kcal mol⁻¹ for PBTZ169. These intriguing findings underscored the enduring stability of the investigated StreptomeDB compounds with DprE1 throughout the entire MDS.

3.4.4 Root-Mean-Square Fluctuation (RMSF)

The Cα atom is fundamental in defining the backbone direction of each amino acid within the DprE1 chain. Figure 7a shows the RMSF values (in nm) for the Cα atoms, providing insights into the residue flexibility within the DprE1 structure. The apo-, hydroxythaxtomin A-, lajollamycin B-, and PBTZ169-DprE1 complexes demonstrated average RMSF values of 0.14, 0.14, 0.12, and 0.14 nm (Figure 7a). The RMSF analysis revealed significant fluctuations in certain DprE1 residues within the 260–290 and 310–330 ranges, implying that the interactions between the identified StreptomeDB compounds and DprE1 imposed various constraints on the motions near these amino acids (Figure 7a). The observed variations in RMSF values may reflect differences in the internal dynamics and interaction intensities.

3.4.6 H-Bond Analysis

The number of H-bonds versus time was estimated to comprehend the stability of hydroxythaxtomin A, lajollamycin B, and PBTZ169 inside the DprE1 binding site over the course of 300 ns MDS (Figure 8). Notably, the average number of H-bonds observed for hydroxythaxtomin A, lajollamycin B, and PBTZ169 complexed with DprE1 were 5, 4, and 3, respectively. The H-bond analysis demonstrated the steadiness of examined StreptomeDB compounds within the DprE1 binding site during the entire MDS.

3.5 Physicochemical Features

The drug-likeness properties and bioactivity of hydroxythaxtomin A (SDB12202) and lajollamycin B (SDB9226) as DprE1 inhibitors were assessed using SWISS-ADME server,^[71] compared to those of PBTZ169 (Figure 9). Notably, the HBD of hydroxythaxtomin A, lajollamycin B, and PBTZ169 were less than 5, with values of 4, 4, and 0, respectively. HBA values of hydroxythaxtomin A, lajollamycin B, and PBTZ169 were either equal to or less than 10, with specific values of 7, 10, and 8, respectively. The MW of hydroxythaxtomin A, lajollamycin B, and PBTZ169 were determined as 454.4, 673.8, and 456.5 gm mol⁻¹, respectively. Importantly, the slight increase in the MW of lajollamycin B is not expected to substantially affect its transmission and diffusion, given that numerous FDA-approved drugs exceed the typical molecular weight limit of 500.^[72] Additionally, the TPSA values of the investigated StreptomeDB compounds were 162.92 and 191.45 Ų, respectively, signifying their advantageous properties concerning oral absorption and membrane permeability.^[73] Besides, hydroxythaxtomin A, lajollamycin B, and PBTZ169 exhibited MLogP values close to five, with specific values of 1.54, 4.95, and 3.49, respectively, suggesting their good lipophilicity.

3.6 ADMET

To elucidate the behavior of the identified StreptomeDB compounds within the human body, an in-depth investigation of their pharmacokinetic properties was conducted. ADMET properties of hydroxythaxtomin A (SDB12202), lajollamycin B (SDB9226), and PBTZ169 were predicted utilizing pkCSM online server (Table 3). In line with the absorption (A) data listed in Table 3, hydroxythaxtomin A, lajollamycin B, and PBTZ169 were found to have Caco-2 permeability values of 0.26, 0.28, and 0.61 cm s⁻¹. The Caco-2 permeability is determined high if its value >0.90 cm s⁻¹.^[74] Hydroxythaxtomin A, lajollamycin B, and PBTZ169 exhibited low skin permeability (log Kp) with values of −2.74, −2.72, and −2.72 cm h⁻¹, respectively, suggesting there no possible dermatological effect. To gauge drug distribution (D), the permeability of the blood-brain barrier (BBB) membrane and the CNS was assessed. Hydroxythaxtomin A, lajollamycin B, and PBTZ169 estimated log BB values of −1.09, −1.43, and −0.9, respectively (Table 3). For CNS permeability, hydroxythaxtomin A, lajollamycin B, and PBTZ169 revealed log PS scores of −2.95, −3.51, and −2.1, respectively (Table 3). These absorption findings suggested that the identified StreptomeDB compounds are unlikely to penetrate the BBB or cross the CNS. Focusing on metabolism (M) predictions, hydroxythaxtomin A and lajollamycin B were confirmed as non-inhibitors of the CYP3A4 enzyme. Conversely, PBTZ169 was found to inhibit CYP3A4, with additional metabolism testing showing that both the identified StreptomeDB compounds and PBTZ169 function as substrates for the CYP3A4 enzyme, as outlined in Table 3. Furthermore, the total clearance property provided insights into the excretion (E) rates, with values of 0.42, 1.36, and 0.02 log ml/min/kg of hydroxythaxtomin A, lajollamycin B, and PBTZ169, respectively. The assessment of the toxicity (T) effect highlighted lajollamycin B as non-toxic, while hydroxythaxtomin A and PBTZ169 were classified as slightly toxic inhibitors. This study reveals maximum tolerated dose (MTD) values of 0.473, −0.51, and 0.003 mg/kg/day for hydroxythaxtomin A, lajollamycin B, and PBTZ169, respectively, shedding light on their distinct toxicity profiles. All three compounds lack hERG I inhibitory activity, indicating low cardiotoxicity risk, but inhibit hERG II. Moreover, the skin sensitivity test confirmed the absence of adverse reactions, ensuring the safety of contact with these proposed StreptomeDB compounds. Regarding cell wall penetration, potential structural modifications could be utilized to enhance the compounds’ ability to traverse the mycobacterial cell wall. Figure S4, Supporting Information illustrates the strategies for structural modifications of hydroxythaxtomin A and lajollamycin B to facilitate penetration through the mycobacterial cell wall.

Overall, hydroxythaxtomin A and lajollamycin B showed encouraging results, with certain values much surpassing PBTZ169, as detailed in Table 3.