Anti-Shigella activity and killing kinetics of compounds from MMV Pathogen and COVID Boxes: Identifying leads against Shigellosis

Abstract

Background

Shigellosis is a leading diarrheal disease, posing critical public health challenges in low- and middle-income countries. Rising antibiotic resistance exacerbates treatment limitations.

Methods

We screened 560 compounds from the MMV Pathogen and COVID Boxes for in vitro activity against Shigella flexneri, S. sonnei, and S. dysenteriae via resazurin-based microdilution. Cytotoxicity (RAW 264.7 cells) and killing kinetics against S. flexneri were assessed.

Results

From the primary screen, 13 compounds (9 from the Pathogen Box, 4 from the COVID Box) showed potent anti-Shigella activity against S. flexneri (MIC ≤10 μM). Among these, seven non-antibiotic Pathogen Box compounds exhibited low-micromolar MICs (0.31–5 μM) and low cytotoxicity (CC₅₀ > 100 μM). The known antibiotic levofloxacin (MMV687798) showed broad-spectrum activity (MIC 0.15–2.5 μM). MMV1804312 (salinomycin) was the most active COVID Box hit (MIC 1.25 μM). Time-kill assays revealed rapid, concentration-dependent bactericidal activity (≥3-log reduction within 4–6 h) for Pathogen Box compounds MMV675997 and MMV021660, while MMV000858 displayed bacteriostatic action.

Conclusion

This study identifies several novel, non-antibiotic leads from the MMV Pathogen Box with potent anti-Shigella activity and favorable cytotoxicity profiles, offering promising starting points for further development against drug-resistant shigellosis.

Graphical abstract

Highlights

• •Thirteen anti-Shigella compounds identified from MMV libraries.
• •Seven novel non-antibiotic leads show potent anti-Shigella activity.
• •Selected leads exhibit low MICs (≤5 μM) and low cytotoxicity (CC₅₀ > 100 μM).
• •Six Pathogen Box compounds (including levofloxacin) show ≥3-log kill within 4–6 h.

1. Introduction

Shigellosis is a severe diarrheal illness caused by Shigella species. Shigella is a group of Gram-negative, facultative anaerobic bacteria (Schnupf and Sansonetti Philippe, 2019). The disease can be caused by four main Shigella species which are Shigella flexneri, Shigella sonnei, Shigella boydii and Shigella dysenteriae with S. flexneri being the most prevalent, followed by S. sonnei and S. dysenteriae. Shigellosis represents a significant global public health challenge, especially in low- and middle-income countries where access to clean water, sanitation, and healthcare is limited (Bennish and Ahmed, 2020). Shigella is highly invasive and capable of penetrating the intestinal epithelium, which triggers inflammation and tissue damage. The pathogenesis of Shigella is driven by a sophisticated array of virulence factors that enable the bacterium to colonize the host, evade immune defences, and induce disease. After ingestion, Shigella traverses the gastrointestinal tract, overcoming the stomach's acidic environment to reach the colon (Pakbin et al., 2023). There, it employs a type III secretion apparatus to inject effector proteins into host cells, facilitating its invasion and intracellular survival. Once inside the epithelial cells, Shigella replicates and spreads to neighbouring cells, causing cell death, mucosal ulceration, and a robust inflammatory response (Pakbin et al., 2023). This inflammation is amplified by the release of pro-inflammatory cytokines, leading to the hallmark symptoms of shigellosis.

Clinically, the disease manifests as a spectrum of symptoms, ranging from mild, self-limiting diarrhoea to severe dysentery, which can be complicated by life-threatening conditions such as dehydration, bacteraemia, and hemolytic uremic syndrome (Raso et al., 2023).

Globally, Shigella infections are responsible for an estimated 80–165 million cases of disease and approximately 1 million deaths each year, with the majority of cases and fatalities occurring among children (Raso et al., 2023; Rogawski et al., 2020).

The current standard treatment for shigellosis relies on antibiotics such as ciprofloxacin as a first-line therapy, azithromycin, ceftriaxone and others as second-line options. However, the rapid rise of antimicrobial resistance (AMR) mainly in low- and middle-income countries has severely diminished the efficacy of these therapies, leaving limited options for managing severe cases (Raso et al., 2023; Lefèvre et al., 2023). In Africa, Notably, in the 2024 updated WHO list of priority bacterial pathogens, fluoroquinolone-resistant Shigella species have been elevated from medium to high priority. This reclassification underscores the critical need to explore alternative therapeutic and vaccine strategies to combat this growing public health threat (W.H. Organization, 2024).

While natural product libraries have historically been a rich source of antimicrobials, they often require extensive dereplication and optimization. In contrast, the MMV Pathogen and COVID Boxes consist of pre-curated, drug-like compounds with established activity against diverse human pathogens. These libraries are enriched with molecules that have favorable pharmacokinetic profiles and proven bioactivity, thereby increasing the likelihood of identifying hits with translational potential. Repurposing such libraries offers a time- and cost-effective strategy to discover new anti-Shigella agents, particularly in the context of rising antimicrobial resistance.

By exploiting these well-characterized collections, we aim to identify molecules offering new therapeutic options to address the dual challenges of shigellosis and antimicrobial resistance. Specifically, our objectives are to (i) evaluate the anti-Shigella activity of the compounds within these libraries, (ii) determine the killing kinetics of promising candidates.

2. Materials and methods

The chemicals, antibiotics, and reagents used in the experiments were purchased from Sigma-Aldrich (Sigma, St. Louis, MO, USA).

2.1. Preparation of test and reference compounds

The Pathogen Box and COVID Box were generously provided by the Medicines for Malaria Venture (MMV, Geneva, Switzerland). These boxes are collection of structurally diverse compounds that can provide new starting points for drug discovery. The Pathogen Box is a collection of 400 drug-like compounds specifically selected for their potential activity against some of the world's most critical neglected tropical diseases, including trypanosomiasis, leishmaniasis, cryptosporidiosis, toxoplasmosis, filariasis, schistosomiasis, dengue virus, and trichuriasis, in addition to malaria and tuberculosis. The MMV COVID Box, on the other hand, contains 160 diverse compounds, including marketed drugs and those under research or clinical development, representing 19 distinct drug classes with known or predicted activity against SARS-CoV-2. The compounds within both libraries represent a mix of synthetic small molecules and natural product-derived scaffolds, selected based on their drug-likeness, synthetic accessibility, and demonstrated activity against a range of infectious diseases. Both libraries were supplied in sealed 96-well plates, with each well containing 10 μL of a 10 mM solution prepared in 100 % dimethyl sulfoxide (DMSO). For practical use, the original master plates were replicated into working plates, each containing 10 μL of a 1 mM solution, in incomplete Dulbecco's Modified Eagle Medium (DMEM) (for cytotoxicity) or Mueller Hinton Broth (for antibacterial assay) and stored at −20 °C until further use. The original library plates contained compounds at 10 mM in DMSO.

2.2. Bacterial species and cell culture

Shigella flexneri 2457T (Catalog N^o NR-518, serotype 2a) and Shigella sonnei Strain WRAIR I Virulent (Catalog N^o NR519), were obtained from BEI Resources (https://www.beiresources.org/). A clinical isolate of S. dysenteriae was provided by the Centre Pasteur of Cameroon under a material transfer agreement. All bacterial species were maintained in slant culture tubes stored at 4 °C and sub-cultured 24 h before each experiment on Petri dishes containing Mueller Hinton agar medium, followed by incubation at 37 °C.

For cytotoxicity assays, the RAW 264.7 murine macrophage cell line (RCB0535) was used between passages 5 and 15. These cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10 % fetal bovine serum (FBS) and 1% penicillin-streptomycin, then incubated at 37 °C in a 5 % CO₂ atmosphere for 72 h.

All experiments were conducted under Biosafety Level 2 containment.

2.3. Assessment of the Anti-Shigella susceptibility

2.3.1. Preparation of the bacterial suspension

Colonies from an overnight bacterial culture were collected and suspended in physiological saline (0.9 % NaCl). The turbidity of the suspension was adjusted to match the 0.5 McFarland standard, which corresponds to a bacterial concentration of approximately 1.5 × 10⁸ cells/mL. This suspension was then diluted to a final concentration of 10⁶ cells/mL for use in subsequent tests.

2.3.2. Single-concentration preliminary screening

The broth microdilution assay was conducted following the Clinical and Laboratory Standards Institute (CLSI) guidelines (M07-A9 protocol, 2012) (Cockerill, 2012) with slight modifications, using a resazurin dye-based assay. All compounds were screened at a final test concentration of 10 μM. The assay was performed in two independent biological replicates, each conducted on separate days with freshly prepared bacterial cultures and compound dilutions. Each biological replicate included two technical duplicates, resulting in a total of n = 4 data points per compound. Briefly, in a sterile 96-well microplate, 98 μL of Mueller Hinton Broth was added to each well, followed by 2 μL of compounds from the intermediate test compounds plate concentrated at 1 mM. Subsequently, 100 μL of a standardized bacterial suspension (10⁶ cells/mL) was added to each well (except sterility controls), yielding a final compound concentration of 10 μM and a final DMSO concentration of 0.1% (v/v) in each well determined in preliminary experiments to have no significant effect on bacterial growth. Ciprofloxacin (1.44 μM; Sigma-Aldrich) prepared in acidified water served as the positive control. The negative control contained bacterial inoculum in medium with 0.1% DMSO, and the sterility control consisted of medium alone. The plates were sealed and incubated at 37 °C for 24 h. After incubation, 20 μL of freshly prepared resazurin solution (0.15 mg/mL in DPBS; Sigma-Aldrich) was added to each well. Plates were incubated for an additional 30 min in the dark. Crucially, the 24 h endpoint was confirmed to ensure the resazurin color change (blue to pink/colorless) is stable and clearly differentiated between growing and non-growing wells. Resazurin is reduced from blue to pink by metabolically active bacteria; wells that remained blue indicated no bacterial growth and were recorded as active. Compounds showing complete growth inhibition (100% inhibition, no colour change) at 10 μM were selected for minimum inhibitory concentration (MIC) determination via dose–response assay.

2.3.3. Determination of minimum inhibitory concentrations (MICs), and minimum bactericidal concentrations (MBCs) of active compounds

The assay was performed twice in two independent experiments using sterile 96-well microplates and the resazurin-based method described earlier. The compounds were tested at concentrations ranged from 10 μM to 0.3125 μM in duplicate, while ciprofloxacin concentrations were ranged from 1.44 μM to 0.045 μM. The final bacterial concentration in each well was 5 × 10⁵ cells/mL. Control wells were treated as described previously. The plates were incubated and processed under the same conditions as previously outlined. The lowest concentration of the compound in which no colour change (from blue to pink) was observed, indicating no visible bacterial growth, was recorded as the MIC value of the corresponding compound.

The minimum bactericidal concentrations (MBCs) of the most active compounds were determined following the CLSI (formerly NCCLS) M26-A microdilution procedure (Standards and Barry, 1999). A protocol similar to MIC determination was employed, with the exception that the primary inoculum was prepared from bacterial cultures in the logarithmic growth phase after MIC determination. Following 24 h of incubation at 37 °C, 10 μL of bacterial suspension from each well showing no visible growth (concentration ≥ MIC) was inoculated onto MH2 agar plates (bioMérieux) to determine the MBCs. The plates were then incubated at 37 °C for 48 h, and colony-forming unit (CFU) counts were assessed. These counts were compared against CLSI M26-A reference tables. The MBC was defined as the lowest concentration of the compound that achieved a ≥99.9 % reduction in the initial bacterial load.

2.4. Assessment of cytotoxicity assay

The cytotoxicity assay of active compounds was conducted using the RAW 264.7 mouse macrophage adherent cell line, following a previously described method (Yuan et al., 2018) with minor modifications. Cells were seeded at a density of 4 × 10⁴ cells/well in sterile 96-well tissue culture plates and allowed to adhere in 100 μL DMEM supplemented with 10% fetal bovine serum (FBS) for 24 h at 37 °C and 5% CO₂. After adherence, the cells were treated with 10 μL of varying concentrations (0–200 μM) of each selected compound and incubated for an additional 48 h under the same conditions. Podophyllotoxin at 10 μM and 1% DMSO were also included as positive and negative controls, respectively. Following incubation, resazurin dye (0.15 mg/mL in DPBS) was added to each well, and the plates were incubated for 4 h at 37 °C and 5% CO₂. Fluorescence was measured using an excitation wavelength of 550 nm and an emission wavelength of 590 nm. The assay was performed twice in duplicate and the results were expressed as the percentage of cell viability relative to untreated control cells.

Only promising compounds with MIC ≤5 μM and CC₅₀ > 100 μM were selected for the time-kill kinetic assay.

2.5. Assessment of killing kinetic assay of potent compound

To assess the bactericidal activity of the most promising compounds, time-kill kinetics were performed, adapting the method of Klepser et al. (1998). Briefly, selected compounds were serially diluted two-fold in sterile 96-well microplates to a final volume of 100 μL. Each well, except those for sterility control, was then inoculated with 100 μL of bacterial suspension at a density of 1 × 10⁶ cells/mL, resulting in compound concentrations ranging from 2 MIC to ½ MIC. The plates were incubated at 37 °C, and samples were taken at 0, 2, 4, 6, 8, 10, 12, and 24 h. Ciprofloxacin (at 1.44 μM) served as a positive control, and all tests were conducted in duplicate. At each time point, 2 μL aliquots were removed from each well, diluted 1000-fold in 0.9% NaCl, and plated on Muller Hinton Agar. Following a 24-h incubation at 37 °C, bacterial colonies were counted. The lower limit of detection for bacterial enumeration was 50 CFU/mL; values below this threshold were recorded as 50 CFU/mL for kill curve generation. Kill curves were plotted with the logarithmic mean of colony-forming units per millilitre (log10 CFU/mL) on the y-axis and time (hours) on the x-axis.

2.6. Definition of bactericidal activity

Bactericidal activity was defined using a unified, two-part criterion: (1) an MBC/MIC ratio ≤4, based on established CLSI interpretive guidelines for bactericidal action; and/or (2) a ≥ 3-log₁₀ reduction in viable bacterial counts in time-kill assays over a 24-h incubation period. Compounds meeting either or both of these criteria were classified as bactericidal. Compounds with an MBC/MIC ratio >4 and that did not achieve a ≥ 3-log reduction in time-kill assays were classified as bacteriostatic. This dual-parameter approach ensures consistent and reproducible classification of antibacterial pharmacodynamics throughout the study.

2.7. Statistical analysis

The CC₅₀ values were calculated from dose-response curves using sigmoid dose-response curves in GraphPad Prism 5.0 software. The results are expressed as the mean ± standard error of the mean (SEM).

3. Results

3.1. Minimum inhibitory concentrations (MICs) and minimum bactericidal concentrations (MBCs) of the MMV's chemical libraries

All the 560 compounds from the Pathogen and COVID boxes were screened for their anti-Shigella activity following a standardized drug discovery cascade (Fig. 1).

Fig. 1.

Screening workflow for identification of anti-Shigella compounds from the MMV Pathogen and COVID Boxes. A total of 560 compounds were screened at 10 μM against S. flexneri NR518 using a resazurin-based microdilution assay. Primary hits (n = 48) were defined as compounds showing no colour change (no bacterial growth) after 24 h. These hits were tested in dose-response to determine MIC values, yielding 13 confirmed actives (MIC ≤10 μM). Cytotoxicity screening identified 8 non-cytotoxic compounds (CC₅₀ > 100 μM), of which 6 were selected for time-kill kinetic studies based on potent anti-Shigella activity (MIC ≤5 μM). All time-kill candidates originated from the Pathogen Box.

Of these 560 compounds screened, 48 showed primary hits showed ≥99% growth (no colour changes after addition of resazurin) of S. flexneri growth at 10 μM in the primary single-concentration assay. From these, 13 compounds (Supplementary Table S1) (9 from the Pathogen Box, 4 from the COVID Box) were confirmed to have MIC values ≤ 10 μM against S. flexneri. Detailed antibacterial activity and cytotoxicity parameters for these confirmed hits are presented in Table 1.

Table 1.

Anti-Shigella activity parameters: MIC, MBC and CC₅₀.

Compounds	Trivial Name	Shigella flexneri NR518			Shigella sonnei NR519			S. dysenteriae (Isolate)			RAW (246.7 cells)
		MIC (μM)	MBC (μM)	MBC/MIC	MIC (μM)	MBC(μM)	MBC/MIC	MIC (μM)	MBC(μM)	MBC/MIC	CC50 (μM)
PATHOGEN BOX
MMV675997	–	0.62 ± 0.00	1.25 ± 0.00	2	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV000858	–	0.62 ± 0.00	5 ± 0.00	8	10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV021660	–	0.31 ± 0.00	5 ± 0.00	16	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV688771	–	5 ± 0.00	>5 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV004168	–	0.31 ± 0.00	2.5 ± 0.00	8	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV687699	–	0.62 ± 0.00	>0.62 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	>100 ± 0.00
MMV687798	Levofloxacin (−)-ofloxacin	0.15 ± 0.00	1.25 ± 0.00	8.00	0.62 ± 0.00	2.5 ± 0.00	4	2.5 ± 0.00	10 ± 0.00	4	>100 ± 0.00
MMV021057	Azoxystrobin	0.31 ± 0.00	0.31 ± 0.00	1	–	–	ND	–	–	ND	0.46 ± 0.10
MMV228911	–	10 ± 0.00	>10 ± 0.00	–	–	–	ND	–	–	ND	18.73 ± 3.23
COVID BOX
MMV003461	Niclosamide	1.25 ± 0.00	5 ± 0.00	4	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	67.32 ± 2.78
MMV002798	Papaverine	10 ± 0.00	>10 ± 0.00	ND	10 ± 0.00	>10 ± 0.00	ND	10 ± 0.00	>10 ± 0.00	ND	98.51 ± 1.34
MMV004066	Doxorubicin	5 ± 0.00	10 ± 0.00	2	5 ± 0.00	10 ± 0.00	2	>10 ± 0.00	>10 ± 0.00	ND	>100
MMV1804312	Salinomycin	1.25 ± 0.00	1.25 ± 0.00	1	>10 ± 0.00	>10 ± 0.00	ND	>10 ± 0.00	>10 ± 0.00	ND	56.19 ± 2.02
Positive controls
Ciprofloxacin		0.15	0.62	4	0.15	0.31	1.99	0.07	0.15	2	–
Podophyllotoxin		–	–	–	–	–	–	–	–	–	0.7 ± 0.003

CC₅₀: median cytotoxic concentration; MIC: minimal inhibitory concentration; MBC: minimal bactericidal concentration, MIC and MBC values represent the mean ± SEM from two independent biological replicates (n = 4). ND = Not determined (MBC > tested concentration); >value: No activity observed at the highest concentration tested (e.g., >10 μM). ‘–’ = not tested. MIC: Minimal Inhibitory Concentration; MBC: Minimal Bactericidal Concentration; CC₅₀: Median cytotoxic concentration. Levofloxacin (MMV687798) is included as a known antibiotic control. Bactericidal activity was defined as MBC/MIC ≤4 and/or ≥3-log kill in time-kill assays.

Dose-response studies of the nine identified active compounds (2.25% hit rate) from the Pathogen Box revealed MIC values ranging from 0.15 to 10 μM against S. flexneri (Table 1). Among these, seven were non-antibiotic leads with potent activity against S. flexneri (MIC 0.31–5 μM) and low cytotoxicity (CC₅₀ > 100 μM) (Table 1). The known antibiotic Levofloxacin (−)-ofloxacin (MMV687798) served as a reference controland, was the most active, with a MIC of 0.15 μM and MBC of 1.25 μM. Of the novel leads, MMV021660, MMV004168, and MMV021057 (Azoxystrobin) also showed good activity against S. flexneri with a MIC of 0.31 μM, and MBCs of 5.25, and 0.31 μM, respectively (Table 1).

Among the 4 identified active COVID box compounds (MMV003461, MMV002798, MMV004066 and MMV1804312), The known antibiotic, MMV1804312 (Salinomycin) was the most promising active against S. flexneri, with a MIC value of 1.25 μM and MBC value of 1.25 μM (Table 1).

To see the potent broad-spectrum anti-Shigella activity, active compounds were screened against S. sonnei NR519 and clinical isolate of S. dysenteriae. Susceptibility testing showed that the known antibiotic compound, MMV687798 (Levofloxacin (−)-ofloxacin), a Pathogen Box compound was active against the two Shigella species with MICs ranging from 0.62 to 2.5 μM (Table 1) with bactericidal effect, showing the broad-spectrum activity of the compound. MMV000858 inhibited the growth of S. sonnei NR519 with an MIC value of 10 μM with a bacteriostatic effect (MBC>10 μM). With MIC and MBC values of 5 μM, and 10 μM, respectively, MMV004066 presented the same activity profile on S. flexneri and S. sonnei.

3.2. Cytotoxicity

Cytotoxicity assessment in RAW 264.7 macrophages revealed that eight of the 13 active compounds were non-cytotoxic (CC₅₀ > 100 μM), including all seven non-antibiotic Pathogen Box leads and MMV004066 from the COVID Box (Table 1, of which 6 were selected for time-kill analysis based on potent anti-Shigella activity (MIC ≤5 μM).

3.3. Killing kinetic effect active compounds

Based on the anti-Shigella parameters activity and cytotoxicity profile, six identified promising compounds (MIC ≤5 μM and CC₅₀ > 100 μM), all from Pathogen Box were prioritized for time-kill study against S. flexneri to evaluate the speed and extent of the bactericidal effect of these compounds.

The time-kill studies revealed a range of antibacterial activities among the tested Pathogen Box compounds, from bacteriostatic inhibition to concentration-dependent bactericidal killing (Fig. 2).

Fig. 2.

Time kinetic profile of promising identified compounds. The bactericidal activity of promising compounds was evaluated by time-kill kinetic studies against S. flexneri NR518. MMV687798, displayed a concentration-dependent bactericidal effect after 4 h of incubation, similar to the positive control ciprofloxacin, at all tested concentrations. The time-kill kinetics profile of other compounds showed a significant decrease in viable bacteria counts when compared to the control. For MMV00858 a bacteriostatic action at all tested concentrations was observed. MMV675997 and MMV021660 are found to be bactericidal at 2 MIC and 4 MIC. MMV004168 presented bactericidal profiles at 1 MIC, 2 MIC and 4 MIC. MMV688771 and MMV687699 were bactericidal only at 2 MIC and 4 MIC, respectively.

As anticipated, the known antibiotic Levofloxacin (−)-ofloxacin (MMV687798), a known antibiotic, exhibited rapid, concentration-dependent bactericidal activity after 4 h, closely mirroring the effect of the ciprofloxacin control. It is noteworthy that the log reduction achieved by Levofloxacin (−)-ofloxacin plateaued at ≤3 log10, indicating a potential limitation in its killing efficacy under the experimental conditions used. In contrast, the novel non antibiotic compounds displayed distinct activity profiles. While MMV000858 exerted a consistently bacteriostatic effect, the other leads—MMV675997, MMV021660, MMV004168, MMV688771, and MMV687699—showed concentration-dependent bactericidal activity, with the extent and speed of killing varying by compound and concentration.

In particular, compounds MMV675997 and MMV021660 required higher concentrations (2x MIC and 4x MIC) to achieve a ≥3-log reduction, suggesting a concentration-dependent mechanism where a critical threshold must be reached to induce cell death. In contrast, MMV004168was more potent, showing bactericidal activity even at the lowest tested concentration (1x MIC), underscoring its potential as a highly effective antibacterial agent. Meanwhile, MMV687699 only demonstrated bactericidal activity at higher concentrations (2x MIC and 4x MIC), indicating that lower concentrations were insufficient to cause significant bacterial death.

3.4. Classification of bactericidal activity using a unified criterion

Bactericidal activity was defined using a unified criterion: either an MBC/MIC ratio ≤4 and/or a ≥ 3-log₁₀ reduction in viable bacterial counts in time-kill assays over 24 h. Based on this, eight of the 13 active compounds demonstrated bactericidal activity (Table 1, Fig. 2). From the Pathogen Box, MMV675997, MMV004168, MMV021660, MMV687699, MMV688771, and MMV021057 met the criterion, together with the known antibiotic levofloxacin (MMV687798). Among the COVID Box hits, MMV1804312 (salinomycin), MMV003461 (niclosamide), and MMV004066 (doxorubicin) also exhibited bactericidal profiles (MBC/MIC ≤4). Time-kill assays confirmed that MMV675997, MMV021660, MMV004168, MMV687699, and MMV688771 produced a ≥ 3-log reduction in CFU/mL within 4–6 h at elevated concentrations (2–4 × MIC) (Fig. 2). In contrast, MMV000858 and MMV002798 (papaverine) were classified as bacteriostatic, as they did not meet the bactericidal threshold under the tested conditions.

4. Discussion

Shigellosis remains a major global health concern, particularly in low- and middle-income countries, where it disproportionately affects children under five years of age (Baker, 2018). The alarming rise in antimicrobial-resistant Shigella strains underscores the urgent need for novel therapeutic options (Raso et al., 2023; Cohen, 2022). In this study, we leveraged two curated libraries from the Medicines for Malaria Venture—the Pathogen Box and COVID Box—to identify new anti-Shigella compounds. Screening of 560 compounds yielded 13 active hits including MMV675997, MMV000858, MMV021660, MMV688771, MMV004168, MMV687699, MMV687798, MMV021057, and MMV228911 (from the Pathogen Box), and MMV003461, MMV002798, MMV004066, and MMV1804312 (from the COVID Box) with MIC values ranging from 0.156 to 10 μM against S. flexneri. Of these, seven non-antibiotic leads from the Pathogen Box exhibited potent activity (MIC ≤5 μM) and low cytotoxicity (CC₅₀ > 100 μM), highlighting these libraries as valuable resources for antibacterial repurposing (Adam et al., 2024; Duffy et al., 2017).

The known fluoroquinolone antibiotic Levofloxacin (−)-ofloxacin (MMV687798) was identified as a potent hit in our screen, serving as an important internal control. Its activity against all three Shigella species (MIC 0.15–2.5 μM) closely mirrored that of ciprofloxacin and validated the reliability of our resazurin-based microdilution and time-kill assays. Its rapid, concentration-dependent bactericidal kinetics (≥3-log reduction within 4 h) represent a significant advantage, as fast-acting agents can reduce the risk of resistance development (Abouelhassan et al., 2019). This effect is attributable to inhibition of bacterial topoisomerase IV and DNA gyrase (Collins and Osheroff, 2024), and may be supplemented by disruption of membrane integrity or interference with RNA and protein synthesis (Furet and Pechère, 1990; Millanao et al., 2021). While levofloxacin is clinically established, its inclusion in the Pathogen Box underscores the utility of such libraries for both discovery and validation purposes.

The seven non-antibiotic Pathogen Box leads represent diverse chemotypes originally developed against other pathogens: anti-kinetoplastid (MMV675997, MMV004168), antimalarial (MMV000858, MMV021057), and anti-schistosomiasis (MMV688771) agents (Duffy et al., 2017; Veale and Hoppe, 2018). Among these, MMV675997 and MMV004168 were particularly promising, exhibiting low MICs (0.31–0.62 μM), bactericidal activity, and no significant cytotoxicity (CC₅₀ > 100 μM). MMV021660 also showed potent inhibition (MIC 0.31 μM), though its bactericidal activity required higher concentrations. In contrast, MMV000858 was bacteriostatic, which may offer advantages in contexts where sustained suppression is preferred.

The COVID Box yielded chemically distinct hits categorized by primary indications: antiparasitic (MMV003461: niclosamide), nervous system agent (MMV002798: papaverine), antispasmodic (MMV004066: doxorubicin), and antimicrobial (MMV1804312: salinomycin). MMV1804312 (salinomycin) exhibited potent bactericidal activity (MIC = MBC = 1.25 μM). This polyether ionophore, known for anticancer and antifungal effects (Cerqueira et al., 2024; Sharma et al., 2023; Antoszczak, 2019), may act through disruption of membrane potential and induction of oxidative stress. MMV003461 (niclosamide) and MMV004066 (doxorubicin) also showed bactericidal profiles, consistent with reported activity against Klebsiella pneumoniae (Cerqueira et al., 2024).

Using a unified criterion for bactericidal activity (MBC/MIC ≤4 and/or ≥3-log reduction in time-kill assays), we classified eight of the 13 active compounds as bactericidal. These included Pathogen Box compounds MMV675997, MMV004168, MMV021660, MMV687699, MMV688771, and MMV021057, alongside the control levofloxacin (MMV687798). From the COVID Box, MMV1804312, MMV003461, and MMV004066 also met this criterion. Time-kill kinetics confirmed rapid, concentration-dependent killing (≥3-log reduction within 4–6 h) for several Pathogen Box leads. In contrast, MMV000858 and MMV002798 (papaverine) were bacteriostatic.

Our findings align with and extend previous MMV library screens. Duffy et al. (2017) reported broad-spectrum antibacterial activity in the Pathogen Box but did not include Shigella; our work adds this pathogen to the susceptible list. Similarly, Veale & Hoppe (Veale and Hoppe, 2018) identified anti-trypanosomal leads from the same library. Several of our hits have documented activity against other pathogens: MMV021057, MMV688771, and MMV687699 inhibit methicillin-resistant Staphylococcus aureus (MRSA) with MICs of 0.75–25 μM (Adam et al., 2024), while MMV675997, MMV004168, and MMV228911 are active against polymyxin-resistant E. coli (Adam et al., 2024). MMV021057 and MMV021660 are known antimalarials, with IC₅₀ values of 0.04 μM and 0.160 μM, respectively (Duffy et al., 2017), and MMV021660 also shows anti-Giardia ((IC₅₀ ≤ 2 μM)) and anti-tubercular activity (Veale and Hoppe, 2018; Tiash et al., 2020). However, MMV021057's cytotoxicity (CC₅₀ 0.464 μM) limits its direct therapeutic potential.

To our knowledge, this is the first report of anti-Shigella activity for MMV000858, MMV675997, and MMV004168, compounds not previously associated with antibacterial action against this pathogen. This underscores the Pathogen Box as a valuable source of novel, safe chemotypes for infectious disease drug discovery(Duffy et al., 2017; Veale and Hoppe, 2018; Tiash et al., 2020; Bhandari et al., 2018).

5. Study limitations

While this study identifies promising anti-Shigella candidates, several limitations should be acknowledged. First, the primary screen used a single reference strain (S. flexneri); although follow-up testing included S. sonnei and S. dysenteriae, future screens should incorporate a broader panel of clinical and multidrug-resistant isolates, to better assess spectrum and translational relevance. Second, although the resazurin-based readout, is widely used for viability assessment, it relies on metabolic reduction and could theoretically yield false positives or negatives if compounds interfere with redox signalling or bacterial metabolism directly; however, key findings were confirmed by MBC and time-kill assays. Finally, mechanistic studies—such as efflux pump inhibition and outer-membrane penetration assays—were beyond the scope of this initial screening but will be essential for advancing the most promising leads.

6. Conclusion

Screening of Pathogen Box compounds identified seven promising anti-shigellosis candidates MMV687798, MMV000858, MMV675997, MMV004168, MMV688771, MMV687699, and MMV021660, while the COVID Box yielded fewer hits, MMV1804312 emerged as a promising candidate warranting further study. The favorable cytotoxicity profile and high efficacy of these compounds make them ideal candidates for lead-optimization, for future structure-activity relationship (SAR) and in vivo studies. Future work will focus on structure-activity relationship (SAR) studies to enhance potency and selectivity, as well as in vivo efficacy evaluations in relevant infection models. Additionally, these leads and their analogs will be tested against a broader panel of clinical and multidrug-resistant Shigella isolates to assess their spectrum of activity, frequency of resistance (FoR), and potential cross-resistance patterns. Mechanistic investigations, including effects on biofilm formation, efflux pump, membrane integrity, and specific molecular targets, will be essential to elucidate their modes of action and guide rational drug design.

CRediT author statement

Lauve Rachel Tchokouaha Yamthe: Conception, Supervision, Methodology, Investigation. Patrick Valere Tsouh Fokou: Conception, Supervision, Methodology, Investigation, Writing – original draft. Yanick Kevin Melogmo Dongmo: Methodology, Investigation, Formal analysis, Writing – original draft. Francine Kengne Mediesse: Methodology, Investigation. Sonia Raissa Matchuenkam Gayap: Methodology, Investigation. Darline Dize: Methodology, Investigation. Fabrice Fekam Boyom: Resources. All authors: Writing – review & editing, Validation. All authors read and approved the submitted version.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Patrick Valere Tsouh Fokou reports equipment, drugs, or supplies was provided by Medicines for Malaria Venture. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.