Anti-virulence and bactericidal activities of Stattic against Shigella sonnei

ABSTRACT Shigella sonnei is an important human pathogen that provokes severe colitis and diarrhea, causing high mortality and morbidity worldwide. Because of increasing drug resistance, it is particularly important to develop new antibacterial agents against S. sonnei. In this study, we report that Stattic, a potent signal transducer and activator of transcription 3 inhibitor, was found to be effective against S. sonnei. Our results showed that Stattic has dual antibacterial mechanisms. At low concentrations, Stattic binds to GalU of S. sonnei, inhibits its enzyme activity, and then reduces biofilm formation and extracellular polysaccharide production, which ultimately attenuates S. sonnei virulence in human cell lines. At high concentrations, Stattic could kill bacteria by disrupting the integrity of the bacterial cell membrane, causing space collapse and the outflow of contents. Furthermore, Stattic showed excellent synergistic effects with antibiotics against S. sonnei. In addition, Stattic also showed anti-virulence activity against Escherichia coli and Klebsiella pneumoniae. Together, our findings suggest the promising potential and advantages of Stattic as a new therapeutic agent against bacterial pathogens. IMPORTANCE Shigella sonnei is a major human enteric pathogen that causes bacillary dysentery. The increasing spread of drug-resistant S. sonnei strains has caused an emergent need for the development of new antimicrobial agents against this pathogenic bacterium. In this study, we demonstrate that Stattic employs two antibacterial mechanisms against S. sonnei. It exerted both anti-virulence activity and bactericidal activity against S. sonnei, suggesting that it shows advantages over traditional antibiotics. Moreover, Stattic showed excellent synergistic effects with kanamycin, ampicillin, chloramphenicol, and gentamicin against S. sonnei. Our findings suggest that Stattic has promising potential for development as a new antibiotic or as an adjuvant to antibiotics for infections caused by S. sonnei.

serious antibiotic resistance (12).Therefore, there is an urgent need to develop new antibiotics or find another treatment against S. sonnei (13).
The mechanisms of bacterial resistance to antibiotics are varied, among which the formation of biofilms is one of the most effective methods (14,15).Bacterial biofilms are characterized by adhesion to the solid surface and the production of an extracellular polymer matrix composed of extracellular polysaccharides (EPS), proteins, DNA, and lipids (16)(17)(18)(19).EPS not only provides bacteria with adhesion but also protects bacteria from the host immune response and antibacterial therapy (20).In addition, EPS can create a pathogenic environment (such as acidic pH and hypoxia) and promote the release of key virulence factors (21).The galU gene encodes a highly conserved protein (known as GalU) that catalyzes the synthesis of UDP-α-D-glucose, which is involved in the lipopolysaccharide (LPS) core region biosynthetic process in Enterobacteriaceae (22)(23)(24).Besides, UDP-α-D-glucose is the important intermediate in several different metabolic pathways and biosynthetic reactions involved in the biosynthesis of capsular polysaccharides (CPS) and EPS in bacteria (25)(26)(27).Previous studies revealed that GalU is critical for bacterial attachment, biofilm formation, drug resistance, and immune evasion (28)(29)(30)(31).Mutation of galU caused reduced virulence in many gram-negative pathogens, suggesting the vital role of GalU in the pathogenesis of various bacterial pathogens (32)(33)(34).
Due to the importance of GalU in the physiological processes and pathogenesis of bacteria, it has attracted attention to be used as a target for the development of antibacterial drugs (24,35,36).It was found that pyrimidinyl benzamide could inhibit GalU to attenuate Listeria virulence (35), and some compounds target GalU and reduce Streptococcus virulence (36).In this study, we discovered that Stattic, which is a small molecule that selectively inhibits the activation of signal transducer and activator of transcription 3 (STAT3) by inhibiting phosphorylation and dimerization processes (37,38), displayed strong antimicrobial activity against S. sonnei by inhibiting virulence through targeting GalU or killing bacterial cells directly.Our findings suggest that Stattic might have an advantage over conventional antibiotics for the treatment of S. sonnei infections.

Screening of leading compounds that inhibit the virulence of S. sonnei
Biofilm development offers a protected mode of growth that not only permits cells to survive in adverse environments but also allows microorganisms from microbial clusters to colonize new niches, which are a major source of bacterial infection (39).Biofilm formation is also a key pathogenic phenotype of S. sonnei (40).We first evaluated the inhibitory effects of approximately 1,000 compounds on S. sonnei biofilm formation at a final concentration of 20 µM to screen for leading compounds against S. sonnei, and 42 candidate compounds were selected because they could effectively decrease more than 15% biofilm production (Fig. S1).Then, we used the HeLa cells infection model to detect whether these compounds had an effect on the virulence of S. sonnei, which was evaluated by the content of lactate dehydrogenase (LDH) in the supernatant of the culture medium when cells were co-incubated with S. sonnei.The results showed that among these selected 42 compounds screened with biofilm formation assays, only nine compounds could attenuate S. sonnei virulence by more than 15% (Fig. 1), among which Bilirubin, Stattic, and Ethyl gallate had significant inhibitory activity, lowering S. sonnei virulence by more than 30% (Fig. 1).

Stattic can bind to GalU
As GalU plays a key role in the biofilm formation and pathogenesis of bacterial patho gens, we continued to investigate whether there is a relationship between these compounds and GalU in S. sonnei.The gene galU is located between hns and hnr, and the homologous protein of GalU in S. sonnei shares 100% identity with that in E. coli, based on the National Center for Biotechnology Information Basic Local Alignment Search Tool (BLAST) program, which contains a nucleotide transferase domain (Fig. 2a  and b).Molecular docking is a common tool for predicting and understanding the binding of small molecules to proteins, and this method was performed by the AutoDock program (41,42).The docking results showed that Brazilin, Ungeremine, Stattic, and Bilirubin had a strong affinity for GalU, and their binding energies were less than −6 kcal/mol (Table S1).We then used microscale thermophoresis (MST), which is a novel biomolecule interaction analysis technique that measures ligand-target molecule affinity by observing changes in the target molecule's conformational size, charge, and solvation state after the ligand binds to the target molecule (43), to confirm the interaction between these compounds and GalU, which was purified using affinity chromatography (Fig. 2c).Surprisingly, MST experiments revealed that only Stattic could bind to GalU with a well-fit curve, and the dissociation constant (K d ) obtained by MST was 20 ± 0.35 µM (Fig. 2d), while the remaining eight compounds could not bind to GalU (Fig. S2).GalU is highly conserved in gram-negative bacteria.We then constructed and expressed GalU homolog proteins from four other species, including Klebsiella oxytoca, Acinetobacter baumannii, Burkholderia cenocepacia, and Pseudomonas aeruginosa, which share 94%, 58%, 54%, and 40% identity with GalU of S. sonnei, and tested their binding ability to Stattic (Fig. S3a).The results showed that the GalU protein of K. oxytoca also bound to Stattic with a K d of 50 ± 0.25 µM (Fig. S3b through e).

Glutamine 109 is the key site for binding and activity
GalU functions as a nucleotidyltransferase, catalyzing the formation of UDP-α-D-glucose and diphosphate (Fig. 3a).We tested whether the binding of Stattic to GalU affects the catalytic activity, and the malachite green reagent method was employed for detection (44).After incubating the protein for 10 min with various concentrations of Stattic, α-D-glucose 1-phosphate was added to initiate the reaction.The results demonstrated that the enzyme activity of GalU gradually decreased with increasing Stattic concentration and decreased by 60% when the concentration of Stattic was 50 µM (Fig. 3b).Then, we attempted to identify the Stattic-binding sites in S. sonnei GalU.The structure of E. coli homologous protein GalU was used as the docking template to predict the binding mode of Stattic and GalU by autodocking analysis.It was revealed that five amino acid residues, Arg21 (R21), Lys31 (K31), Lys65 (K65), Gln109 (Q109), and Gly179 (G179), might be critical for the interaction between GalU and Stattic.The docking model showed that the five residues are distributed near the active pocket of GalU, which may also contribute to the catalytic center (Fig. 3c).We then generated five single-point mutants (R21A, K31A, K65A, Q109A, and G179A).MST analysis showed that mutations at R21 and K65 markedly weakened the binding between GalU and Stattic, and the mutation of Q109 led to an absence of binding affinity between them (Fig. 3d).Moreover, GalU Q109A enzyme activity was reduced by 35% compared to wild-type GalU (Table S2), and enzyme activity was unaffected when 10 µM Stattic was added, indicating that Q109 was not only the key site for binding but also for enzyme activity.In addition, R21 and G179 also affected the activity of GalU, which was consistent with previous findings (44,45).

Stattic inhibits pathogenic phenotypes and shows a good synergistic effect with antibiotics against S. sonnei
Stattic inhibited the activity of GalU in a dose-dependent manner (Fig. 3b).Thus, we then examined the effects of different concentrations from 1 to 25 µM of Stattic on the pathogenic phenotypes in S. sonnei, including biofilm formation, EPS production, and cell cytotoxicity.The results showed that biofilm formation, EPS production, and cell cytotoxicity were inhibited by Stattic in a dose-dependent manner, and exogenous addition of Stattic at 25 µM resulted in decreases in biofilm formation, EPS production, and cell cytotoxicity by 42%, 46%, and 32%, respectively, compared to those of the group without Stattic (Fig. 4a through c), and under this conditions, Stattic displayed slight toxicity to HeLa cells (Fig. S4).In addition, GalU Q109A could only slightly restore the pathogenic phenotypes of galU deletion mutants, and the addition of Stattic did not inhibit the phenotypes of ΔgalU(galU Q109A ) (Fig. 4d through f), which suggested that Q109 of GalU is a crucial amino acid for Stattic binding and the biological functions of GalU in S. sonnei.
Synergistic therapy is a common treatment strategy for severe and complex infections as well as for drug-resistant bacterial infections (46)(47)(48).We then tested the synergis tic effect between Stattic and the commonly used antibiotics kanamycin, ampicillin, chloramphenicol, gentamicin, and spectinomycin on the treatment of Shigella infection (49).The results showed that when antibiotics were combined with Stattic, the inhibitory effect was enhanced, and the minimum inhibitory concentration (MIC) values were reduced to 6.25, 1.56, 1.56, 1.56, and 12.5 µg mL −1 from 12.5, 12.5, 3.125, 3.125, and 25 µg mL −1 , respectively (Table 1), indicating that the Stattic and antibiotic combinations had a significant synergistic antibacterial effect.The determination of the total number of CFUs also suggested that Stattic had a good synergistic antibacterial effect (Fig. S5a  and b).Furthermore, we continued to investigate whether Stattic has additive activity with antibiotics when used to treat HeLa cells infected by S. sonnei.The results showed that the combination of Stattic with antibiotics showed a stronger inhibitory effect on the virulence of S. sonnei than the antibiotics alone.Among them, the combination of ampicillin and Stattic exhibited a notable reduction of the virulence of S. sonnei by more than 65% (Fig. 5).

Bactericidal activity of Stattic at high concentrations
Stattic did not affect the normal growth rate of S. sonnei cells when the concentration was less than 25 µM.However, we found that Stattic exhibited an obviously bactericidal activity when the concentration exceeded 100 µM (Fig. S6).The time-kill assay revealed that the number of bacterial cells decreased dramatically over time when S. sonnei was exposed to high concentrations of Stattic, and the deletion of galU did not affect the bactericidal efficacy of Stattic (Fig. S7a and b).Changes in the protein and nucleic acid content of bacterial culture medium supernatants can reflect bacterial damage.After 10 h, the protein and nucleic acid content in the supernatant of S. sonnei cells treated with Stattic increased significantly compared with the group without Stattic (Fig. S8a and b).Furthermore, when the bacterial cells were centrifuged and analyzed by SDS-PAGE, the protein bands from S. sonnei cells treated with DMSO were significantly richer (Fig. S8c).The LIVE/DEAD BacLight Bacterial Viability Kit, which includes the fluorescent dyes SYTO 9 and propidium iodide (PI), could also be used to assess the antibacterial properties of Stattic.The cells without Stattic treatment were stained with green fluorescence, and no red fluorescence was observed (Fig. S8d); however, when a high concentration of Stattic was added, the red fluorescence gradually increased (Fig. S8e and f ), indicating that the bacterial cell membrane was damaged, which resulted in the nucleus being stained by PI.These results indicated that Stattic could destroy bacterial cell membranes, allowing bacterial contents to escape and causing irreversible damage.Furthermore, after exposure to Stattic, the bacterial cell membrane wrinkled, perforated, and even collapsed (Fig. S8g through i), and deletion of galU showed no effect on the demolishment of Stattic (Fig. S8j through o).Taken together, these results show that Stattic exerted good bactericidal activity at high concentrations.

Development of drug-resistance of S. sonnei to Stattic
To reveal the bactericidal mechanism of Stattic, we tried to identify the target of Stattic by chemical mutagenesis.S. sonnei grew slowly in the presence of 50 µM Stattic (Fig. S5).After long-term induction, the growth rate of S. sonnei exposed to 50 µM Stattic increased significantly (Fig. S9a), indicating that it had developed resistance to Stattic.We then sequenced the genome of the mutant strain (named SY1) and compared it with that of the wild-type strain.There were three important variations in the mutant strain, including insertion or deletion (InDel), single nucleotide polymorphism (SNP), and copy number variation (CNV).The results of InDel and SNP showed that nine genes were mutated (Tables S3 and S4), and according to the base types before and after mutation,  the mutant had no base preference, and the length of InDel was also within a reasonable range (Fig. S9b and c).The results of the functional analysis showed that four of them were transposases, and we tested whether the other five genes were associated with Stattic drug resistance.We overexpressed the five mutant genes in wild-type S. sonnei, which were named WT(rpoS), WT(ipaH_1), WT(rhsA), WT(sson_1673), and WT(sson_0750), respectively.Unfortunately, the results of the growth curve showed that the overexpres sion strains could not restore the growth rate to the S. sonnei SY1 strain when they were exposed to 50 µM Stattic (Fig. S9d).MST experiments also proved that the proteins encoded by these genes could not bind with Stattic (Fig. S10), indicating that these five proteins were not the key targets of Stattic.Therefore, we turned our attention to the last variant.A considerable increase or reduction in the copy number of large portions of the genome with a length of more than 1 kb is referred to as CNV, a complex phenomenon that is typically brought on by genomic rearrangement.The results of resequencing showed that there was a copy number increase in the gene segment found in the mutant SY1, and the gene length was 24,500 bp.By analyzing this gene fragment, we found that several genes encoding multiple antibiotic-resistance proteins were distributed in this region, mainly including MarR, MarA, MarB, and MarC (Fig. S9e).We overexpressed these genes in the wild-type strains, and the strains had increased resistance to Stattic (Fig. S9f ).Therefore, we speculated that the resistance of the drug-resistant strain SY1 to Stattic is not caused by base mutations but by additional regulation of resistant proteins.

Stattic inhibits the pathogenic phenotypes of E. coli and K. pneumoniae
As the intestinal pathogens E. coli and K. pneumoniae, which belong to Enterobacteria ceae, have developed multiple drug resistance (50, 51), we then measured the effect of Stattic on E. coli and K. pneumoniae because the homologous protein of GalU in S. sonnei shares 100% and 94% identities with those in E. coli and K. pneumoniae.The results showed that Stattic inhibited the virulence-related phenotypes of E. coli and K. pneumoniae in a dose-dependent manner.Exogenous addition of 25 µM Stattic resulted in decreases in biofilm formation, EPS production, and cell cytotoxicity of E. coli by 41%, 40%, and 37% (Fig. 6a through c) and decreases in those of K. pneumoniae by 32%, 38%, and 30% (Fig. 6d through f).These results suggest that Stattic is a potential antimicrobial compound with a broad spectrum.

DISCUSSION
Shigella is one of the main pathogens of intestinal infection that can cause shigellosis (6,8).Shigellosis is a global human health problem and a major cause of diarrhea that causes approximately 700,000 deaths per year worldwide and millions of hospital izations (52,53).However, drug resistance in S. sonnei is a serious problem, so new antibiotics or therapies for disease treatment are needed (54,55).The highly conserved protein GalU participates in the LPS core region biosynthetic process (22,24), which is a critical virulence factor of bacteria.It has been reported that galU mutants in Francisella tularensis showed reduced virulence in a murine pulmonary model (56).In addition, Guo et al. showed that GalU affects the formation of biofilm in planta of Xanthomonas citri subsp., while mutation of galU resulted in a loss of pathogenicity in grapefruit (57).
Listeria monocytogenes is a foodborne gram-positive pathogen, and studies have shown that the product encoded by galU is involved in wall teichoic acid galactosylation (58).
The galU mutant was also severely reduced in a mouse oral-virulence model and shown to have defects in listeria actin-based motility; additionally, it even developed sensitivity to the antibiotic cefotaxime, which acts on the cell wall (35,59).Previous studies have reported that the GalU protein functions as an EPS synthase involved in LPS biosynthesis in Shigella (22,31).Our findings indicated that Stattic can bind to GalU with a K d of 20 ± 0.35 µM, and we also discovered that the GalU homolog of K. oxytoca can also bind to Stattic with a K d of 50 ± 0.25 µM, suggesting that GalU can not only be used as an effective antibacterial target for compounds to block the virulence of S. sonnei but also as a target for other bacterial species (Table S5).Biofilm formation is an important pathogenic phenotype in S. sonnei and is one of the mechanisms of resistance used by bacteria.In addition, EPS is also an impor tant virulence factor that promotes bacterial colonization and creates a pathological environment (60,61).In this study, we discovered that Stattic has an excellent inhibitory effect on pathogenic phenotypes at low concentrations, which could inhibit its activity by directly binding to GalU and thus decreasing the yield of EPS as well as the formation of biofilm in a dose-dependent manner, ultimately reducing the virulence of S. sonnei.It was found that Q109 of GalU is the key binding site of Stattic, and mutation of it eliminated the inhibitory effect of Stattic.In addition, the residues of R21, K31, K65, and R179 also contribute to the activity of GalU.Our results demonstrated that Stattic has antimicrobial efficacy against S. sonnei by inhibiting GalU activity to interfere with the formation and secretion of virulence factors.This antibacterial mechanism is different from the direct germicidal efficacy of traditional antibiotics, which is beneficial to avoid the emergence of drug-resistant strains.
Controlling the use of antibiotics and reducing the pressure of antibiotic selection are the mainstream requirements to control the development of bacterial drug resistance.In gram-negative bacteria, galU was reported to be involved in the synthesis of polysac charides, including LPS and CPS (22), which are important components used by bacteria to resist antibiotics (62,63).Our results also showed that Stattic can be used in combina tion with many antibiotics to reduce their dosage (Table 1).One of the reasons is that Stattic can bind to GalU and inhibit LPS and CPS biosynthetic processes, thus increasing the sensitivity of S. sonnei to antibiotics.In addition, the combination of Stattic and antibiotics also reduced the toxicity of S. sonnei toward human cells (Fig. 5), suggesting that Stattic could potentially be developed as a new antimicrobial agent to treat S. sonnei infections and reduce the use of other antibiotics to prevent the development of drug resistance.
In addition, we found that Stattic also has germicidal effects at high concentrations.When S. sonnei is exposed to high concentrations of Stattic (>100 µM), its cell membrane will be destroyed, leading to content outflow, space collapse, and irreversible damage to S. sonnei.At the same time, we also discovered that S. sonnei can develop resistance to Stattic by upregulating the expression of multiple antibiotic-resistance genes, which proves that direct sterilization makes it easy to develop drug-resistant strains again.Stattic is a potent STAT3 inhibitor for tumor treatment (37,38), but its antimicrobial activity has not yet been reported.Our results showed that Stattic has at least dual antibacterial mechanisms at different concentrations and a broad spectrum of antimi crobial activity, suggesting that it has better development prospects than traditional antibiotics.

Strains, culture, and agents
Table 2 contains a list of the bacterial strains employed in this study.The plasmids and bacterial strains used in this study were all sequenced.Unless otherwise specified, S. sonnei and E. coli strains were cultured in Luria-Bertani (LB) medium (5 g/L yeast extract, 10 g/L tryptone, 10 g/L NaCl; pH 7.4) or LB medium with 15 g/L agar at 37°C.Antibiotics were added to the medium based on the needs of the experiment, and the following antibiotics were employed in this work: ampicillin (100 µg/mL), chloramphenicol (50 µg/ mL), kanamycin (100 µg/mL), gentamicin (50 µg/mL), or spectinomycin (100 µg/mL).Stattic (CAS 19983-44-9) was dissolved in DMSO before dilution, which was purchased from Solarbio.

Autodocking
The target GalU structure (PDB: 2E3D) was obtained from the Protein Data Bank (https:// www.rcsb.org/),and the protein contains four chains, A, B, C and D, which form a homotetramer.Chain A was used for the dock model.AutoDock 4.2.1 was used for all dockings in this study, which was supported by Autodock tools and MGL tools.It was also used for protein optimization, including deleting water, adding the polar hydrogen group, and Computing Gasteiger.Torsion bonds of the compounds were selected and defined.The docking parameters were kept to their default values.The receptor grid was centered at x = −9.176,y = 42.570, and z = 13.969, and the grid spacing changed from 0.375 to 0.44.Ligand tethering of the protein was accomplished by adjusting the genetic algorithm (GA) parameters throughout the course of 10 iterations.The dock results of complex conformation ranked by energy were displayed in PyMOL, which is the most popular protein visualization software.

Protein expression and purification
After cloning the intact galU gene into the pET28a (+) vector, it was transformed into competent E. coli BL21 (DE3) cells.Then, the LB medium with 100 g/mL kanamycin was used for protein expression.Up to an OD 600 of 0.4-0.6, cells were grown at 37°C and 220 rpm. 1 mM isopropyl-D-1-thiogalactopyranoside was added to induce recombinant gene expression overnight at 16°C.Cells were collected by centrifugation at 4,000 × g for 30 min after induction.Cells were disturbed by sonication after being resuspended in phosphate buffered saline (PBS) (pH 7.4).A 1 mL His-Tag column (GE Healthcare) that had previously been equilibrated with PBS was loaded with the supernatant after the resulting suspension had been centrifuged for 30 min at 4°C and 12,000 g.The PBS containing 250 mM NaCl were used to wash the column.Finally, the target proteins were eluted with 250 mM imidazole in PBS buffer, which was placed into gels (GenScript SurePAGE), and the purity was evaluated by SDS-PAGE.

MST assays
The binding affinity of compounds and GalU was detected by using the MST.Briefly, the purified GalU was labeled using the Protein Labeling Kit.Fluorescence was measured after labeled protein, and ligand concentrations were loaded onto silicon capillaries that had undergone standard treatment (MO-K025, NanoTemper Technologies).The binding affinity was measured and carried out at automatic LED power and 40% MST power on a Monolith NT.115 instrument.

Construction of the S. sonnei mutant and complemented strains
The galU mutants were generated by an in-frame deletion strategy using the λ Red recombinase system as described previously (40).In brief, the PCR product, which contains the 39 bp homologous arm sequences and the chloramphenicol-resistant gene amplified from pKD3, was transferred into the competent S. sonnei containing pKD46.Besides, the pUC18 was used to construct the complemented plasmid.The mutants and complemented strains were identified by PCR and Sanger sequencing, and the primers used in this study were all listed in Table 3.

Biofilm formation assays
The biofilm formation assays were carried out by using crystal violet staining as previously described (40).In brief, the OD 600 of fresh S. sonnei was diluted to 0.05 and added to each well of a 96-well polystyrene plate in the absence or presence of different concentrations of Stattic, and the samples without shaking were incubated at 37°C for 12 h.Then, remove the supernatant carefully and gently rinse with PBS three times.The samples were fixed with methanol and dried in an oven at 45°C, 150 µL of 0.5% crystal violet was added for staining and then washed with PBS to fully remove the crystal violet.Finally, the samples were dissolved by adding 150 µL of 95% ethanol.The OD 570 was measured using the Multiskan Spectrum.

Quantification of EPS
The same procedures as reported above were used to prepare the S. sonnei suspensions (OD 600 = 0.05).The samples with or without different concentrations of Stattic were cultured at OD 600 = 3.0 at 37°C with shaking, centrifuged at 12,000 × g for 30 min at 4°C, the supernatant was collected and incubated with two times the volume of absolute ethanol for 12 h at 4°C and then centrifuged at 12,000 × g for 10 min again.After removing the supernatant, the precipitate was air-dried, and the weight was calculated by using a one-hundred-thousandth analytical balance.

Cell cytotoxicity assays
Cytotoxicity assays were carried out in accordance with previously described procedures (40).In brief, HeLa cells (1 × 10 5 cells/well) were co-incubated with fresh S. sonnei cells at 10 9 colony-forming units (CFU)/mL for 8 h in dulbecco's modified eagle medium (DMEM) (1% fetal bovine serum).The CytoTox 96 Kit (G1780, Promega) was used to measure the content of LDH released from infected cells.The cytotoxicity experiment results were assessed by detecting the OD 490 and calculating the cytotoxicity compared to an uninfected control.

Growth rate analysis
The same procedures as reported above were used to prepare the S. sonnei suspensions (OD 600 = 0.05).The samples were incubated at 37°C with shaking, and the OD 600 was detected every 4 h.Each experiment was repeated three times.

Time-kill assay
The time-kill assay was carried out as described previously (65).Antibiotics with or without Stattic were added to S. sonnei suspensions (7 log CFU/mL).LB broth containing DMSO was used as the control group.The samples were incubated at 37°C, and the medium was taken at 0, 4, 8, 12, 18, and 24 h, which was diluted with sterile water and spread on a non-resistant LB plate overnight.Finally, the number of bacteria was calculated.

Determination of MICs
The MICs of antibiotics were estimated using the microdilution method described previously (65).In brief, antibiotics were added to 96-well plates with 150 µL of S. sonnei suspensions (OD 600 = 0.05) at seven different final concentrations (50 µg/mL, 25 µg/mL, 12.5 µg/mL, 6.25 µg/mL, 3.125 µg/mL, 1.56 µg/mL, and 0.78 µg/mL).The MIC was determined by measuring the OD 600 nm on a multimode microplate reader.For the combination therapy assay, Stattic was added to the mixture at a final concentration of 10 µM.The plates were then incubated at 37°C for 24 h without shaking, and the follow-up experimental steps were consistent with the time-kill assay.

Determination of nucleic acid/protein leakage
The leakage of nucleic acids and proteins through the S. sonnei cell membrane was measured according to the method described previously (65).Briefly, S. sonnei cells suspended in sterile water were treated with different concentrations of Stattic (0, 100 µM, and 200 µM) at 37°C for 10 h.By centrifuging at 12,000 × g for 15 min at 4°C, the extracellular nucleic acid and protein content of the supernatants were determined using a NanoPhotometer (Implen N60 Touch).Intracellular protein analysis was performed by SDS-PAGE.

Determination of membrane integrity by confocal laser scanning microscopy
Bacterial suspensions (OD 600 = 0.5) treated with different concentrations of Stattic were incubated at 37°C overnight.Before staining, samples were washed twice with PBS.Then, fluorescent dyes (SYTO 9 and PI) were added, and the mixtures were incubated in the dark for 10 min.Samples were washed twice with PBS and finally resuspended in PBS.Eightmicroliters of the suspension stained with SYTO 9 and PI were dropped on a microscope glass slide and observed using a confocal laser scanning microscopy (CLSM).

Determination of morphological changes by scanning electron microscopy
Bacterial suspensions (OD 600 = 0.5) treated with different concentrations of Stattic (0, 100 µM, and 200 µM) were incubated at 37°C for 4 h.After incubation, the samples were centrifuged (4°C, 4,000 × g, 5 min) and washed twice with PBS.Then, 2.5% glutaralde hyde in PBS was added to resuspend the cell pellets, and the resulting samples were incubated at 4°C for 10 h.The cell pellets were collected and centrifuged at 4,000 × g for 5 min at 4°C and then washed twice with PBS.The 30%, 50%, 70%, 80%, 90%, and 100% water-ethanol solutions were used for the gradient dehydration of the samples.Finally, samples were treated with critical point drying and gold spray and observed using a SEM.

Screening of the Stattic-resistant S. sonnei mutant strain and whole-genome resequencing
The method of successive generations was used to develop the resistance of S. sonnei to Stattic.In brief, Stattic at a final concentration of 50 µM was added to the S. sonnei suspensions (OD 600 = 0.05) and then incubated for 24 h at 37°C.Fresh S. sonnei suspensions with or without 50 µM Stattic were used as negative or positive control groups, respectively.Successive generations were performed until there was a significant difference in growth rates between the experimental group and the negative control group.Finally, whole-genome resequencing of the Stattic-resistant S. sonnei mutant strain and S. sonnei wild-type strain was carried out by Shanghai Personalbio Biotechnology (China) using the Illumina HiSeq sequencing platform (paired-end, 2 × 150 bp).The genome sequences are deposited in the National Center for Biotechnology Information (NCBI) BioProject repository under the accession number PRJNA1025100, and the BioSample numbers of S. sonnei wild-type strain and mutant SY1 strain are SAMN37717590 and SAMN37724331, respectively.

Statistical analysis
Statistical analyses were performed using GraphPad Prism 8.All experiments were performed at least three times independently.Data are presented as the mean ± SD.Unpaired t tests between two groups, one-way analysis of variance, or two-way analysis among multiple groups were used to calculate P values.P values are reported using the following symbolic representations: ns (no significance), *P < 0.05, **P < 0.01, and ***P < 0.001.

FIG 1
FIG 1 Effects of compounds on the virulence of S. sonnei.All compounds were dissolved in Dimethyl sulfoxide (DMSO).LDH release was used to assess cytotoxicity, and the amount of LDH released by S. sonnei infectious assays treated with the same volume of DMSO was used as a control and defined as 100%, aimed to normalize the LDH production of HeLa cells infected by S. sonnei in the presence of compounds.The data are presented as the mean ± SD (n = 3, independent measurements).One-way ANOVA was used to determine the significance of the results (*P < 0.05; **P < 0.01; ***P < 0.001; ns = no significance).

FIG 2
FIG 2 Stattic binds to GalU.(a) Genomic organization of the galU region in S. sonnei Ss046.(b) Analysis of the domain structure of GalU.(c) SDS-PAGE analysis of the GalU protein.(d) MST analysis of the binding of Stattic to GalU.

FIG 3
FIG 3 The effect of Stattic on the enzyme activity of GalU.(a) Catalytic reaction of GalU.(b) The change in GalU activity in the presence of different concentrations of Stattic.(c) The structure model of the binding of Stattic to GalU.Cyan represents the solid surface of GalU protein, yellow represents the stick model of Stattic, and magenta represents the amino acid residues that may interact with Stattic.(d) MST analysis of the binding of Stattic to GalU mutants.The data are presented as the mean ± SD (n = 3, independent measurements).One-way ANOVA was used to determine the significance of the results (*P < 0.05; **P < 0.01; ***P < 0.001; ns = no significance).ND, not detected.

FIG 4
FIG 4 Effects of Stattic on the virulence-related phenotypes of S. sonnei.The virulence-related phenotypes of (a) biofilm formation (n = 5, independent measurements) and (b) EPS production (n = 5, independent measurements) in S. sonnei in the presence of different concentrations of Stattic were examined.(c) The cell cytotoxicity of S. sonnei in the presence of different concentrations of Stattic was evaluated by LDH assay (n = 3, independent measurements).The effects of Stattic on biofilm formation (d), EPS production (e), and cell cytotoxicity (f) of S. sonnei wild-type, galU mutant, galU complemented, and galU mutant complemented with galU Q109A were also examined (n = 3, independent measurements).S. sonnei was treated with different concentrations of Stattic and incubated statically at 37°C.The data are presented as the mean ± SD.Error bars indicate the SDs.One-way ANOVA was used to determine the significance of the results (*P < 0.05; **P < 0.01; ***P < 0.001; ns = no significance).

FIG 6
FIG 6 Effects of Stattic on the virulence-related phenotypes of E. coli and K. pneumoniae.The virulence-related phenotypes of biofilm formation (n = 3, independent measurements) (a), EPS production (n = 3, independent measurements) (b), and cell cytotoxicity (c) of E. coli, and those phenotypes (d-f) of K. pneumoniae, are presented with the addition of different concentrations of Stattic.The cell cytotoxicity of E. coli and K. pneumoniae in the presence of different concentrations of Stattic were evaluated by LDH assay (n = 3, independent measurements).The data are presented as the mean ± SD.Error bars indicate the SDs.One-way ANOVA was used to determine the significance of the results (*P < 0.05; **P < 0.01; ***P < 0.001; ns = no significance).

TABLE 1
Synergistic activities of Stattic at 10 µM with Kan, Amp, Chl, Gen, and Spec against the S. sonnei wild-type strain a

TABLE 2
Bacterial strains and plasmids used in this study

TABLE 3
PCR primers used in this study a

TABLE 3
PCR primers used in this study a (Continued) a Restriction enzyme sites are underlined.F, forward; R, reverse.