Targeting thioredoxin reductase for anti-tuberculosis drug discovery: the inhibitor OSSL_632214

Abstract

Abstract

The persistent prevalence of multidrug-resistant tuberculosis (TB) and extensively drug-resistant TB has progressively undermined the efficacy of TB infection control measures. Thus, developing innovative drugs that act on novel Mycobacterium tuberculosis (Mtb) targets is a strategic priority for global TB control efforts. This research direction offers a robust, evidence-based solution to the urgent, growing challenge of anti-TB drug resistance. Thioredoxin reductase (TrxB2, Rv3913) is an essential enzyme for the in vitro growth and survival of Mtb. At present, no anti-TB lead compounds targeting TrxB2 have been advanced for clinical translation, primarily due to challenges such as off-target drug side effects and insufficient target specificity. We employed a virtual screening approach to screen compounds from small-molecule libraries. We conducted further experiments, including enzyme activity assays and studies using overexpression strains, to verify the therapeutic target and mechanism of action of the hit compound. Using virtual screening, we discovered compound OSSL_632214, which exhibited potent anti-Mtb activity (MIC 3.13–6.25 μg/mL) in vitro and showed no cross-resistance with existing anti-TB drugs. In this study, we report for the first time the potent in vitro anti-Mtb activity of OSSL_632214, which possibly acts by targeting TrxB2 and inducing lipid peroxidation in Mtb. These findings provide a theoretical basis for developing novel anti-TB drugs targeting TrxB2.

Key points

• Development of anti-TB drugs based on the novel target TrxB2

• OSSL_632214 exhibits anti-Mycobacterium tuberculosis activity

• OSSL_632214 exhibits potential for the development of TrxB2-targeted drug

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00253-026-13949-0.

Introduction

Recent epidemiological data highlight that tuberculosis (TB) remains the leading cause of death globally among single infectious diseases and still ranks as the top infectious killer (WHO 2025). The persistent prevalence of multidrug-resistant TB (MDR-TB) and extensively drug-resistant TB (XDR-TB) has progressively undermined the efficacy of infection control measures (Mancuso et al. 2023). Furthermore, clinical drug resistance has further exhausted the available options for clinical treatment, posing a significant challenge to global TB control programs (Tan et al. 2025). The antibacterial activity of conventional anti-TB drugs (e.g., isoniazid and rifampicin) primarily arises from their inhibition of essential processes, including cell wall assembly and gene expression (Kia et al. 2023; Arun et al. 2020). However, the emergence of drug-resistant strains has led to the ineffectiveness of these therapeutic agents. Thus, the development of innovative drugs against novel molecular targets is not only aligned with global strategic health priorities but also critical for overcoming drug resistance in Mycobacterium tuberculosis (Mtb) (Yang et al. 2024).

Endogenous oxidative stress is an inevitable challenge for aerobically adapted microbes. To replicate and persist within the host, Mtb must counteract both endogenous and exogenous oxidative stress (Fan et al. 2023). Mtb possesses multiple antioxidant systems, including catalase, alkyl hydroperoxidase, superoxide dismutase, mycothiol, ergothioneine, thiol peroxidase, and membrane-associated oxidoreductase complex (Lin et al. 2016). Targeting the oxidative stress defense system, an entirely novel therapeutic target with a mechanism of action distinct from that of current clinically available drugs, exhibits considerable potential for the efficient eradication of drug-resistant strains (Lu et al. 2023).

Thioredoxin reductase (TrxB2, Rv3913), a core and indispensable component of the Mtb oxidative stress defense system, is an essential enzyme that governs the in vitro growth and survival of Mtb (Lin et al. 2016). TrxB2 accepts reducing equivalents from NADPH to reduce its downstream thioredoxins, thereby maintaining the reduced state of intracellular thiols and coordinating with the mycothiol system to sustain intracellular redox homeostasis (Sugandhi et al. 2025). Deletion of TrxB2 leads to rapid bacterial death and lysis, regardless of whether Mtb is in the replicative or nonreplicative phase (e.g., under nutrient starvation) (Lin). Additionally, TrxB2 plays a critical role in both the acute and chronic phases of infection (Lu and Holmgren 2014). In 2015, the small-molecule inhibitor auranofin, originally an antirheumatic drug, was found to inhibit TrxB2 in Mtb (Harbut et al. 2015). This inhibition leads to the accumulation of reactive oxygen species (ROS), thereby exerting an anti-TB effect. Unfortunately, in the phase II clinical trials, auranofin not only failed to demonstrate anti-TB activity but also posed substantial safety risks (Wallis et al. 2021). In 2023, gliotoxin was identified as an inhibitor of TrxB2 (Yan et al. 2024). It exhibits in vitro anti-TB activity, but the druggability of gliotoxin still requires further research, which limits its clinical translation.

For the development of TrxB2 inhibitors, the NADPH-binding domain is the optimal target region. First, the CXXC motif, which serves as the core catalytic active site of Mtb TrxB2, is exclusively located within the NADPH-binding domain (Akif et al. 2005). This active site is responsible for accepting reducing equivalents from NADPH and transferring them to the downstream thioredoxin substrates TrxB and TrxC, a process that is indispensable for maintaining intracellular redox homeostasis in Mtb. Moreover, the NADPH-binding domain of Mtb TrxB2 presents significant species-specific structural differences from TrxR orthologs in other species (Sugandhi et al. 2025). Therefore, based on the advantages of targeting TrxB2 for anti-TB therapy, the development of drugs targeting TrxB2 is expected to break through the current challenge of drug resistance, identify novel drug targets, and open up a new dimension for antibacterial treatment.

In this study, a virtual screening approach was employed to screen compounds targeting TrxB2 from small-molecule databases. The compounds obtained from virtual screening were subsequently subjected to in vitro anti-Mtb activity validation. We further performed enzyme activity assays and overexpression strain-based experiments to confirm the therapeutic target and mechanism of action of the active compound. This research aims to obtain lead compounds with anti-TB activity by targeting TrxB2.

Materials and methods

Antimicrobial agents and mycobacterial strains

OSSL_632214 was purchased from ChemDiv. Mtb strain H37Rv (ATCC 27294) for this study was obtained from Beijing Chest Hospital. H37Rv was cultured in Middlebrook 7H9 broth (Difco, Becton, Dickinson and Company, Franklin Lakes, NJ, USA) with 0.2% glycerol (Sinopharm Chemical Reagent Co., Ltd., Beijing, China), 0.05% Tween 80 (Sigma-Aldrich, St. Louis, MO, USA), and 10% oleic acid-albumin-dextrose-catalase (OADC, Becton, Dickinson and Company). After 2 weeks of incubation at 37 °C with continuous shaking at 180 rpm and 5% CO₂, the bacterial culture was used for further experiments.

Docking analysis of TrxB2 structure of Mtb

The small-molecule library for virtual screening was sourced from the open-access ZINC 20 database. Two predefined filtering criteria, namely, in-stock commercial availability and drug-like properties, were applied during data retrieval. The MOL-format structural files of the top 400,000 filtered compounds were downloaded to establish the primary screening library. The molecular docking of OSSL_632214 with the NADPH-binding site of the TrxB2 NADPH domain was assessed using AutoDock Vina. The crystal structure of the NADPH domain (PDB: 2a87) of TrxB2 was downloaded for the receptor. The ligand binding site was defined using the reference ligand NADPH for the NADPH domain. Docking simulations for virtual screening with AutoDock Vina were conducted according to the standard protocol. After molecular docking simulations, all compounds were ranked by their calculated binding affinity. The top 200 ranked compounds were subjected to in silico ADMET (absorption, distribution, metabolism, excretion, and toxicity) prediction using the ADMETlab 3.0 platform (https://admetlab3.scbdd.com/server/screening). Two strict criteria were established to prioritize candidate compounds: (1) full physicochemical compliance with Lipinski’s rule of five and (2) a favorable oral safety profile with a risk of oral absorption (ROA) value < 0.5. Finally, five candidate small molecules were shortlisted for subsequent experimental validation based on confirmed commercial availability. The SMILES strings and IUPAC names of the compounds are fully listed in Table S1 of the Supplementary Materials.

Minimum inhibitory concentration (MIC) was determined by a microplate-based Alamar Blue assay

The MICs of OSSL_632214 against Mtb and clinical strains were tested using the microplate Alamar Blue assay (MABA). Mycobacterial susceptibility testing was performed using sterile 96-well microplates. Briefly, 2 μL of OSSL_632214 stock solution was added to 198 μL of Middlebrook 7H9 medium in the first well to achieve an initial concentration of 50 μg/mL. Twofold serial dilutions were then prepared by transferring 100 μL of the mixture into subsequent wells containing 100 μL of fresh 7H9 medium. Finally, 100 μL of Mycobacterium tuberculosis H37Rv or clinical isolate suspension was inoculated into each well, resulting in final compound concentrations ranging from 25 to 0.02 μg/mL and a final bacterial concentration of 1 × 10⁵ CFU/mL. Each plate included duplicate drug-free positive growth controls and duplicate medium-free negative controls, and plates were incubated at 37 °C in a 5% CO₂ humidified atmosphere for 7 days. After incubation, 32.5 μL of a freshly prepared Alamar Blue /20% Tween 80 mixture (20:12.5 ratio) was added to each well under light-shielded conditions, followed by a further 24-h incubation at 37 °C in 5% CO₂. Color changes were assessed visually, and fluorescence intensity was measured using a Tecan Infinite M200 multimode microplate reader. The growth inhibition rate was calculated as (1−fluorescence of treated wells/mean fluorescence of positive control wells) × 100%, and MIC was defined as the minimum compound concentration yielding 90% growth inhibition of Mtb. Specific experimental methods referred to previous studies (Zhong et al. 2024).

Minimal bactericidal concentration (MBC)

The MBC against Mtb H37Rv was determined using the colony-forming unit (CFU) enumeration method. OSSL_632214 was serially diluted in 24-well plates to generate a concentration gradient spanning 1 × MIC to 16 × MIC. Bacterial cultures were inoculated into each well, resulting in final compound concentrations ranging from 3.13 to 50 μg/mL and a final bacterial concentration of 1 × 10⁵ CFU/mL. After 7 days of compound treatment in 24-well plates, 100 μL of each culture was serially diluted tenfold in Middlebrook 7H9 medium (10^–1 to 10^–6 dilutions). 100 μL aliquots of appropriate dilutions were then plated in triplicate on 7H10 agar plates and incubated at 37 °C for 28 days. The MBC was defined as the lowest drug concentration resulting in a ≥ 99.9% reduction in CFU compared to the untreated control at day 0.

Cellular toxicity assays

The cytotoxicity was evaluated via MTT assay. In brief, Vero and J774A.1 cells were seeded in 96-well plates at a density of 4 × 10⁵ cells/well and incubated for 24 h or 72 h at 37 °C; the cells were then exposed to OSSL_632214 at different concentrations (0 to 50 μg/mL) for 24 h. Subsequently, 10 µL of MTT solution was added to the cells, and cell viability was analyzed after incubation for 4 h by dissolving the formazan precipitate in DMSO and recording the absorbance at 490 nm using a microplate reader (iMark, Bio-Rad, USA).

TrxB2 overexpression, purification, and activity assays

To construct a Mtb strain with overexpressed trxB2 gene (Rv3913), first, we amplified the full-length trxB2 coding sequence via PCR using Mtb genomic DNA as the template (NC_000962.3). The full-length coding sequence of trxB2 (Rv3913) spans genomic nucleotides 4401728–4402735 (total length 1008 bp). Primers were designed to amplify the complete coding sequence, with BamHI and HpaI restriction endonuclease sites added to the 5′ end of the forward and reverse primers, respectively, to enable directional cloning into the pMV261 expression vector (forward primer: 5′-CGCGGATCCATGACCGCCCCGCCTG-3′; reverse primer: 5′-GAAGTTAACTCATCGTTGTGCTCCTATCAATGCG-3′). Subsequently, we introduced the recombinant vector into competent Mtb cells (H37Rv) through electroporation. After electroporation, we plated the cells on Middlebrook 7H10 agar supplemented with kanamycin to select positive transformants. Finally, we verified trxB2 overexpression in the selected clones via quantitative real-time PCR confirming the successful construction of the trxB2-overexpressing Mtb strain. Primers were designed to amplify a 115 bp conserved fragment within the trxB2 coding sequence (genomic coordinates: 4401891–4402005). The validated qPCR primer sequences are as follows: forward primer: 5′-CCACCGACGTGGAGAACTAC-3′; reverse primer: 5′-ACGTCTTCCATACGCAGGTC-3′.

Mtb trxB2 was cloned into the pET28a expression vectors and purified by Ni-NTA chromatography. The SDS-PAGE image of the purified TrxB2 protein and Western blot analysis of the final purified protein are detailed in Figure S1 and Figure S2, respectively. TrxB2 activity assays were performed in a total volume of 50 μL in black clear bottom 96-well plates at 37 ℃. Standard reaction mixtures contained 25 nM TrxB2, 100 μM DTNB, 100 μM NADPH, and 50 μM TrxC, in 50 mM HEPES, pH 7.5, with 2 mM EDTA. Absorbance at 412 nm was read on a Spectramax M5 plate reader as an indication of enzyme activity. Specific experimental methods referred to previous studies (Harbut et al. 2015).

Lipid peroxidation assay

Equal amounts of bacteria treated with DMSO or OSSL_632214 were harvested, sonicated, and then centrifuged for 10 min at 6800 × g and 4℃. The supernatant was collected and placed on ice. Lipid peroxidation MDA assay kit (Beyotime, Shanghai, China) was used for lipid peroxidation measurement.

Statistical analysis

SPSS (version 23.0, Chicago, IL, USA) was used to analyze the experimental data. Normality of the data distribution was tested using the Shapiro–Wilk test. For data conforming to a normal distribution with equal variances, one-way analysis of variance (one-way ANOVA) was used to compare differences among three or more independent groups. When the overall ANOVA result was statistically significant (p < 0.05), the least significant difference test was performed for all pairwise between-group comparisons to identify specific significant differences. All data are expressed as mean ± SEM, and statistical significance was set at p < 0.05.

Results

Pharmaceutical screening based on the NADPH domain of target TrxB2

Targeting the NADPH-binding domain can effectively inhibit TrxB2 function through mechanisms such as suppressing electron transfer, blocking conformational transition, or interfering with substrate binding. To screen for compounds capable of targeting the TrxB2 protein, we collected 400,000 compounds from the ZINC database and used the NADPH-binding pocket as the target site, and the screening process is illustrated in Fig. 1A. Subsequently, the top 200 compounds with the highest binding affinity were selected. Further filtering based on ADMET (absorption, distribution, metabolism, excretion, and toxicity) predictions and the availability of commercial availability resulted in five compounds, which were subjected to in vitro activity validation against Mtb H37Rv (Fig. 1B).

Fig. 1.

Pharmaceutical screening based on the NADPH domain of target TrxB2. A Compound screening process based on the TrxB2 NADPH domain. B Five candidate compounds. C Docking results of OSSL_632214 with TrxB2. D In vitro verification of compound activity

The results demonstrated that compound OSSL_632214 exhibited superior anti-TB activity among the five screened compounds, with a MIC 3.13–6.25 μg/mL (Fig. 1D). Molecular docking analysis revealed that compound OSSL_632214 could bind to the drug-binding pocket of TrxB2 (Fig. 1C, Figure S3). The results showed that OSSL_632214 was stably docked into the functional active pocket of TrxB2, with a calculated binding affinity of −6.685 kcal/mol.

The central ligand molecule is stably bound to key amino acid residues through multiple favorable interactions. Specifically, the ligand forms multiple conventional hydrogen bonds with Asp165, Ser166, His250, and Arg295, which provide the core specificity and stability for the binding. Meanwhile, the aromatic π-system of the ligand forms a π-anion interaction with the negatively charged Asp266, a π-cation interaction with the positively charged Arg292, and a weak π-σ interaction with Ser166. These auxiliary interactions further enhance the binding affinity of the ligand to the protein. Overall, the binding between the ligand and the binding pocket is dominated by highly specific hydrogen bonds and multiple types of π-based interactions, establishing a stable and clearly selective binding mode.

Compound OSSL_632214 exhibits potent anti-Mtb activity

MIC of compound OSSL_632214 was determined to be 3.13–6.25 μg/mL using the MABA. Subsequently, the MIC of OSSL_632214 was evaluated against 17 clinically isolated drug-resistant Mtb strains, and the results showed that the MIC values ranged from 6.25 to 12.5 μg/mL. These findings indicated that compound OSSL_632214 does not exhibit cross-resistance with first-line anti-TB drugs such as rifampicin and isoniazid (Table 1).

Table 1.

MIC ranges of clinical strains and reference strain of Mtb for INH, RIF, CFZ, BDQ, and OSSL_632214 (μg/mL)

Strains	CFZ	RIF	OSSL_632214	INH	BDQ
H37Rv	0.25	0.02	3.13–6.25	0.04	0.02
24-9c	0.25	> 10	6.25	2.5	0.02
30819c	0.25	> 10	6.25	> 10	0.02
16995c	0.13	> 10	6.25	> 10	0.02
24-5c	0.25	> 10	6.25	1.25	0.02
22-2c	0.13	10	12.5	2.5	0.04
20c	0.13	10	12.5	1.25	0.02
14-6a	0.13	> 10	6.25	0.02	0.02
4–7b	1	0.08	6.25	2.5	0.32
8a	0.13	> 10	6.25	0.02	0.01
106c	0.25	> 10	6.25	2.5	0.02
317c	0.13	> 10	6.25	2.5	0.01
12–14c	0.13	> 10	12.5	5	0.02
110c	0.13	> 10	12.5	5	0.02
17-11c	0.13	2.5	6.25	1.25	0.02
152a	0.13	> 10	6.25	0.02	0.01
95c	0.13	> 10	6.25	2.5	0.01
10-2c	0.25	> 10	6.25	2.5	0.02

CFZ clofazimine, BDQ bedaquiline, RIF rifampicin, INH isoniazid

^aMonoresistant Mtb

^bPolyresistant Mtb

^cMultidrug-resistant Mtb

Furthermore, after treating the standard H37Rv strain with drugs for 7 days, the bacterial load in the group treated with 4 × MIC of compound OSSL_632214 decreased by 5.42 Log₁₀CFU/mL compared to the blank control group (Table 2). This result demonstrated that the minimum bactericidal concentration (MBC) of compound OSSL_632214 is 4 × MIC, confirming it as a bactericidal agent.

Table 2.

The minimum bactericidal concentration of compound OSSL_632214

	Log10 CFU/mL (day 7)					Log10 CFU/mL (day 0)
Strain	1 × MIC	2 × MIC	4 × MIC	8 × MIC	16 × MIC
H37Rv	8.04	5.65	0	0	0	5.42

Cellular toxicity assays revealed that the half-maximal cytotoxic concentration (CC₅₀) of OSSL_632214 against Vero cells was 26 μg/mL (24 h) and 7 μg/mL (72 h), while its CC₅₀ against J774A.1 cells ranged from 19 to 20 μg/mL (24 h) and 5 to 6 μg/mL (72 h). Collectively, the above results suggest that compound OSSL_632214 has the potential to serve as a lead anti-TB compound.

Compound OSSL_632214 exerts an inhibitory effect on the biological function of TrxB2

To further verify the inhibitory effect of compound OSSL_632214 on TrxB2, a trxB2-overexpressing Mtb strain was constructed. It was found that the MIC of OSSL_632214 against the trxB2-overexpressing strain was 4–8 times higher than that against the empty vector-containing strain. This observation suggests that OSSL_632214 may potentially inhibit the biological functions of TrxB2 (Table 3).

Table 3.

The MIC of OSSL_632214 against the trxB2-overexpressing strain

Strains	MIC (μg/mL)
	OSSL_632214	RIF	INH
H37Rv	3.13	0.02	0.04
PMV261	3.13	0.02	0.04
PMV261-trxB2	12.25 ~ 25.00	0.02	0.04

Additionally, enzyme activity assays were performed. The results showed that as the concentration of OSSL_632214 increased, the activity of the TrxB2 enzyme was further inhibited (Fig. 2). OSSL_632214 exerts an inhibitory effect on the enzymatic activity of TrxB2, with a half-maximal inhibitory concentration (IC₅₀) of 50.39 μg/mL. This result indicated that compound OSSL_632214 may exert anti-TB effects by targeting TrxB2.

Fig. 2.

Validation of TrxB2 as a potential target of compound OSSL_632214. A M. tuberculosis TrxB2 enzymatic activity after treatment with gradient concentrations of OSSL_632214. B Dose–response curve showing the relative percent activity of TrxB2 with increasing concentrations of OSSL_632214. *Compared with the DMSO group, p < 0.05. n = 3

Compound OSSL_632214 induces lipid peroxidation in Mtb via targeting TrxB2

Inhibition of TrxB2 function disrupts the essential biological processes for bacterial growth, leading to lipid peroxidation in Mtb (Sugandhi et al. 2025). Using the BODIPY 581/591 C11 fluorescent probe assay, it was observed that compared with the DMSO group, compound OSSL_632214 induced significant lipid peroxidation in the bacteria (Fig. 3A). Furthermore, lipid peroxidation levels were evaluated in H37Rv following treatment with the TrxB2-specific inhibitor OSSL_632214 at 6.25 μg/mL. The level of lipid peroxidation was markedly reduced in the TrxB2-overexpressing bacterial strain (Fig. 3B).

Fig. 3.

Compound OSSL_632214 induces lipid peroxidation in Mtb via TrxB2. A OSSL_632214 induced significant lipid peroxidation in the bacteria; B lipid peroxidation levels were evaluated in H37Rv and the TrxB2-overexpressing bacterial strain following treatment with the TrxB2-specific inhibitor OSSL_632214 at 6.25 μg/mL. *Compared with the DMSO + H37Rv group, p < 0.05. ^&Compared with the OSSL_632214 + H37Rv group, p < 0.05. n = 3

Design and molecular docking simulation of compound OSSL_632214 derivatives

Structural modification of compounds was performed based on the chemical structure of compound OSSL_632214. Through ADMET (absorption, distribution, metabolism, excretion, and toxicity) prediction and molecular docking analysis, seven derivatives were identified to exhibit improved properties compared with the parent compound OSSL_632214 (Fig. 4). Table S2 and Table 4 report the comprehensive results for drug-likeness and ADMET prediction of the designed compounds, respectively. Specifically, the designed compounds reduced toxicity. The binding energy values of compounds 2, 3, 5, 6, 7, 8, isoniazid and compound OSSL_632214 bound to TrxB2 protein were obtained from docking simulation (Table 5). This indicates that compound OSSL_632214 has the potential to be optimized as a lead compound.

Fig. 4.

The designed derivatives of compound OSSL_632214

Table 4.

Pharmacokinetic (ADME) and toxicity prediction results for the designed compounds

Propertiesa	1	2	3	4	5	6	7	8
Log S	−2.01	−4.10	−2.26	−1.66	−3.61	−3.52	−2.28	−2.56
Caco2	−5.10	−5.18	−5.25	−5.19	−5.13	−5.22	−5.19	−5.27
HIA	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.00
PPB	39.02	97.06	48.43	17.54	93.30	82.34	41.52	53.67
BBB	0.04	0.03	0.01	0.01	0.10	0.18	0.04	0.01
CL	5.03	2.49	4.37	3.98	3.94	5.11	5.65	5.43
T1/2	1.24	1.15	1.58	1.65	1.57	1.10	1.19	1.25
CYP1A2 inhibitor	1.00	1.00	1.00	0.83	1.00	1.00	1.00	1.00
CYP2C19 inhibitor	0.02	0.95	0.31	0.00	0.87	0.54	0.10	0.91
CYP2C9 inhibitor	0.01	0.89	0.17	0.00	0.69	0.10	0.12	0.79
CYP2D6 inhibitor	0.03	0.01	0.00	0.01	0.00	0.00	0.02	0.00
CYP3A4 inhibitor	0.04	0.28	0.03	0.00	0.07	0.43	0.03	0.12
CYP2B6 inhibitor	0.32	0.86	0.30	0.03	0.75	0.98	0.36	0.55
CYP2C8 inhibitor	0.12	0.85	0.01	0.00	0.74	0.11	0.13	0.10
hERG blocker	0.22	0.44	0.05	0.42	0.32	0.31	0.22	0.22
H-HT	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
DILI	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Ames	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Respiratory	0.74	0.19	0.30	0.75	0.30	0.41	0.61	0.09
Drug-induced nephrotoxicity	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.00
Genotoxicity	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Drug-induced neurotoxicity	0.00	0.01	0.01	0.01	0.01	0.01	0.01	0.00
ROA	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.00

^aLog S  logarithm of aqueous solubility (− 4.5 to 0.5 log mol/L), Caco-2  human colon adenocarcinoma cell line permeability (> − 5.15 log cm/s), HIA  human intestinal absorption (0–0.3: excellent), PPB  plasma protein binding (> 80%), BBB  blood–brain barrier penetration (0 ~ 0.3: excellent; 0.3–0.7: medium; 0.7–1.0: poor), CL  clearance plasma (> 15 mL/min/kg: high clearance; 5–15 mL/min/kg: moderate clearance; < 5 mL/min/kg: low clearance), T 1/2  half-life of drug (ultrashort half-life drugs: 1/2 < 1 h; short half-life drugs: T 1/2 between 1 and 4 h; intermediate short half-life drugs: T 1/2 between 4 and 8 h; long half-life drugs: T 1/2 > 8 h), hERG blocker  human ether-a-go-go related gene (0–1), H-HT  the human hepatotoxicity (0–1), DILI  drug-induced liver injury (0–1), Ames  the Ames test for mutagenicity (0–1), respiratory toxicants (0–1), drug-induced nephrotoxicity (0–1), genotoxicity (0–1), drug-induced neurotoxicity (0–1). Enzyme inhibitor  category 1: inhibitor; category 0: noninhibitor. ROA  rat oral acute toxicity

Table 5.

Molecular docking result of the designed compounds with TrxB2 proteins of MTB

Compound ID	Binding affinity (kcal mol−1)
1	−6.685
2	−8.58
3	−7.044
4	−6.567
5	−8
6	−8.278
7	−6.69
8	−7.058
Isoniazid	−5.131
NADPH	−9.772

Discussion

Through virtual screening, we identified compound OSSL_632214, which exhibits in vitro anti-TB activity with a MIC of 3.13–6.25 μg/mL and no cross-resistance to existing anti-TB drugs by targeting TrxB2. In this study, we report for the first time that compound OSSL_632214 exerts in vitro anti-TB effects, indicating its potential as a promising anti-TB lead compound.

TrxB2 is a core component of the thioredoxin system in Mtb (Lu and Holmgren 2014). It reduces thioredoxins (TrxB and TrxC) using electrons donated by NADPH, thereby maintaining the reduced state of intracellular thiols (Harbut et al. 2015). Targeting the NADPH-binding domain can effectively inhibit TrxB2 function through mechanisms such as suppressing electron transfer, blocking conformational transition, or interfering with substrate binding (Sugandhi et al. 2025). In this study, we screened for compounds targeting the NADPH domain of TrxB2 using a database. This screening led to the identification of compound OSSL_632214, which exhibited in vitro anti-TB activity with a minimum inhibitory concentration (MIC) of 3.13–6.25 μg/mL and no cross-resistance with existing anti-TB drugs. Compound OSSL_632214 was confirmed to be a bactericidal agent, with a minimum bactericidal concentration (MBC) of 4 × MIC. Furthermore, validation using enzymatic activity assays and trxB2-overexpressing strains demonstrated that compound OSSL_632214 exerts its anti-TB activity by inhibiting the function of TrxB2. Collectively, these results suggest TrxB2 is a potential target of OSSL_632214, although further experimental validation is still required to confirm this interaction.

Targeting TrxB2 leads to its loss of function, which results in the accumulation of intracellular thiol oxidative stress. This stress further impairs protein folding, enzyme activity, DNA integrity, and lipid peroxidation, ultimately causing bacterial death (Dechow et al. 2021). The stability of bacterial membrane lipids is crucial for Mtb survival. The occurrence of lipid peroxidation may further disrupt the integrity of the membrane structure, exacerbating bacterial death (Batt et al. 2020). The results have shown that OSSL_632214 induced significant lipid peroxidation, and overexpression of TrxB2 led to a decrease in lipid peroxidation. This indicates that OSSL_632214 induces lipid peroxidation by targeting TrxB2, causing damage to the bacterial cell membrane, and ultimately leading to bacterial cell death.

Compound OSSL_632214 is a complex heterocyclic compound (Korolev et al. 2017). In this study, cytotoxicity assays revealed that OSSL_632214 exhibits moderate cytotoxicity, which presents a challenge for its further development as an anti-TB agent. Nevertheless, rational structural modification of OSSL_632214 was performed to address this limitation. A total of seven novel compounds were designed and seven derivatives were identified to exhibit improved properties compared with the parent compound OSSL_632214. Specifically, these derivatives showed reduced toxicity. These active compounds exhibit excellent drug-likeness properties by adhering to Lipinski’s rule of five with zero rule violations, indicating the compounds’ suitability for administration through the oral route. The results of the in silico ADMET predictions support that the pharmacokinetic properties of the active compounds are well within the acceptable range for use in humans, which supports these compounds as promising candidates for further drug development and discovery of new antitubercular agents. Compared with the ADMET parameters of auranofin, the OSSL_632214 and seven derivatives exhibit superior performance in absorption, metabolism, distribution, and excretion, respectively (Dubey et al. 2025). Therefore, from the perspectives of druggability and safety of the compound, using compound OSSL_632214 as a lead compound holds potential for the development of drugs targeting TrxB2.

In conclusion, this study demonstrated that compound OSSL_632214 exhibits in vitro anti-TB activity. It was further found that compound OSSL_632214 may exert its anti-TB effects by targeting TrxB2, which induces lipid peroxidation in Mtb. These findings provide a theoretical basis for the development of anti-TB drugs based on the novel TrxB2 target.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

QYJ: Formal analysis, Data curation, Visualization, Investigation, Writing- Original draft. JHS: Methodology, Validation, Investigation. YXY: Investigation, Methodology. BW: Investigation, Methodology. YL: Conceptualization, Formal analysis, Writing - Review & Editing, Project administration, Funding acquisition. All authors reviewed, edited, and approved the final version of the article.

Funding

This study was supported by the National Natural Science Foundation of China (82173862) and the Beijing Municipal Administration of Hospitals’ Ascent Plan (DFL20221402).

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval

This article contains no studies with human participants or animals performed by authors.

Competing interests

The authors declare no competing interests.