Salmonella Typhimurium strain SL1344 is from a laboratory strain collection. High-glucose Dulbecco’s modified Eagle’s medium (DMEM), foetal bovine serum (FBS), trypsin, and penicillin–streptomycin solution were purchased from Sangon Biotech (Shanghai) Co., Ltd. Nar was purchased from Chengdu Herbpurify Co., Ltd. Luria–Bertani (LB) medium was obtained from Beijing Solarbio Science & Technology Co., Ltd. SL1344 was cultivated in LB medium at 37 °C with shaking. The strains and plasmids used for this work are shown in Supplementary Table S1.

HeLa human cervical cancer cells and J774A.1 mouse mononuclear macrophages were cultured in DMEM supplemented with FBS (10%), penicillin (100 U/mL) and streptomycin (0.1 mg/mL) at 37 °C with 5% carbon dioxide.

HeLa cells were seeded into 24-well plates (1×10⁵ cells/well) and cultured overnight. SL1344 was cocultured with various concentrations of Nar (0, 16, and 32 μg/mL) for four hours, after which the bacteria were harvested and suspended in DMEM after being washed three times with sterile phosphate-buffered saline (PBS), which was used to treat the cells (multiplicity of infection, MOI = 40) for 30 minutes. Afterwards, the culture medium was discarded, and the cells were lysed with 0.2% saponin after being washed three times with sterile PBS. After dilution, the samples were plated onto LB agar medium and cultured at 37 °C overnight to count the CFUs to investigate the effect of Nar on bacterial adhesion (Shi et al., 2022a). The ratio of adhesion was calculated according to the following formula: (CFU_n1/CFU_n0) ×100, where CFU_n1 and CFU_n0 represent the bacterial numbers obtained from samples with and without Nar treatment, respectively. For the invasion assay, HeLa cells were treated with SL1344 and various concentrations of Nar (0, 16, and 32 μg/mL) for one hour, after which the medium was discarded, and gentamicin (100 μg/mL) was used to kill the extracellular bacteria. One hour later, the cells were harvested and diluted after being washed with PBS three times. Equal-volume samples were plated onto LB agar medium and cultured overnight. The colonies were counted to analyze the effect of Nar on the invasion of SL1344 (Li et al., 2024). The relative invasion was calculated according to the following formula: (CFU_s1/CFU_s0) ×100, where CFU_s1 and CFU_s0 represent the bacterial numbers obtained from samples with and without Nar treatment, respectively.

The minimum inhibitory concentration (MIC) of Nar against SL1344 was determined on the basis of the American Society for Clinical and Laboratory Standards (CLSI) and methods described previously with some modifications (Lambert and Pearson, 2000; Andrews, 2001). Specifically, LB media containing a series of different concentrations of Nar (0–128 μg/mL) were prepared in a 96-well plate, SL1344 was added to reach a final concentration of 5 × 10⁵ colony-forming units per milliliter (CFUs/mL), and the plates were cultured at 37 °C for 24 hours. The minimum concentration with no bacterial growth was defined as the MIC. For the growth curve assay, various concentrations (0, 32, and 64 μg/mL) of Nar were cocultured with the logarithmic growth of SL1344, the optical density at 600 nm (OD600) was detected at the specified time points, and the effect of Nar on the growth of SL1344 was analyzed via statistical analysis.

HeLa and J774A.1 cells were separately seeded into 96-well plates (2 × 10⁴ cells/well) and cultured overnight. Nar (0, 32, and 64 μg/mL) was cocultured with these cells for six hours, the supernatant was mixed with an equal volume of lactate dehydrogenation agent (LDH; Beyotime, Shanghai, China), and the plates were incubated in the dark for 30 minutes after centrifugation (1000 rpm, 10 minutes) (Rho et al., 2020). Afterwards, the OD490 was determined to analyze the cytotoxicity of Nar. Cells treated with 0.1% Triton X-100 or DMEM alone were used as positive (PCs) or negative (NCs) controls, respectively.

HeLa cells were seeded into 24-well cell culture plates (5 × 10⁴ cells/well). The next day, SL1344 treated with or without Nar (32 μg/mL) was used to infect the cells (MOI = 20). After one hour, the medium was removed, the cells were washed twice with sterile PBS, and gentamicin (100 μg/mL) was added to treat the cells for one hour. The cells were subsequently fixed with 4% paraformaldehyde and blocked with 4% goat serum, then Salmonella antibody (1:3000, Abcam) and a goat anti-rabbit secondary antibody conjugated to Alexa Fluor 488 were used to detect extracellular Salmonella. After this, samples were treated with 0.2% Triton X-100 and blocked with goat serum, the intracellular Salmonella were examined by using Salmonella antibody and Alexa Fluor 594 conjugated secondary antibody. The nuclei were stained with Hoechst (Beyotime) (Li et al., 2024). Images were subsequently obtained using a fluorescence microscope (Olympus, IX83) with a 60x oil immersion lens to analyze the effect of Nar on Salmonella invasion. The ratio of invasion was calculated according to the following formula: (Numbers_s1/Numbers_s0) × 100, where Numbers_s1 and Numbers_s0 represent the bacterial numbers obtained from the samples with and without Nar treatment, respectively.

This assay was performed on the basis of a method described previously (Cheng et al., 2017). Specifically, the density of SL1344 was adjusted to 2 × 10⁸ CFUs/mL, and 5-μL samples were dropped vertically onto soft LB agar medium (0.3%) supplemented with various concentrations of Nar (0, 32, and 64 μg/mL) and incubated at 37 °C for 12 hours. The diameters of the clones were measured to evaluate the effect of Nar on the motility of the bacteria.

SL1344 was cultured with different concentrations of Nar (0, 16, and 32 μg/mL) for four hours, and the bacteria were harvested after centrifugation (12,000 rpm, 5 minutes). The total RNA of each sample was obtained using a total RNA extraction kit (Sangon Biotech) following the manufacturer’s instructions. RT–PCR was carried out using reverse transcription kits (Solarbio, Beijing, China) to obtain cDNA, after which the cDNA was used as the template to perform a qPCR assay by using SYBR fluorescent reagent (KTSM1401; AlpalifeBio) (Shi et al., 2022a); gyrb was used as the housekeeping gene. The data were analyzed via the ΔΔCt method. The primers used here are shown in Supplementary Table S2, and the RT–qPCR conditions are shown in Supplementary Table S3.

Molecular docking and dynamics simulations were carried out according to methods reported previously (Gao et al., 2020; Yang et al., 2021; Niu et al., 2022). Briefly, the FimZ protein structure was obtained from AlphaFold (AF-P26319-F1), and the Nar structure file was obtained from PubChem. FimZ was used as the receptor, and Nar was used as the ligand. A docking box was generated after both the acceptor and the ligand were treated with AutoDock Tools, and the docking calculation was carried out with AutoDock Vina (Trott and Olson, 2010; Tang et al., 2022). Molecular dynamics simulation was performed on the basis of the conformation obtained from the docking calculation using GROMACS version 2020.6 (Van Der Spoel et al., 2005). The binding free energy was analyzed to predict the potential binding sites during the binding process.

EMSAs were performed according to a previously described method with some modifications (Hellman and Fried, 2007). The promoter region of fimA was amplified by using the primers shown in Supplementary Table S4, and the DNA of SL1344 was used as the template. Afterwards, the DNA (5 ng) was co-incubated with recombinant FimZ protein (500 ng) with or without various concentrations of Nar in specific buffer for 30 minutes at room temperature. The samples were separated on a 6% native polyacrylamide gel, and the DNA was observed with Typhoon 7000.

This assay was carried out on a Biacore X100 system (GE Healthcare), and the sensor chips used for this system were nitrilotriacetic acid (CM5). Prior to the assay, the His antibody was captured on the sensor chips. The running buffer (PBS-P) used for this assay was filtered and degassed using 0.22-micron filters (Millipore, Billerica, MA, USA). Afterwards, samples of Flag-FimZ-His or its mutants were captured using a His-antibody-coated CM5 sensor chip with a density of approximately 1000 RU. To detect binding between Nar and FimZ, running buffer (50 mM Tris, 150 mM NaCl, 10 mM MgCl₂, 1 mM MnCl₂, and 5% DMSO, pH 7.5) containing Nar (1 μg/mL to 65 μg/mL) was moved through the flow cell at a flow rate of 20 μL/minute (Nguyen et al., 2015; Olaru et al., 2015). The collected data were analyzed using Biacore X100 evaluation software.

TEM was carried out according to a protocol reported previously (Cheng et al., 2018). Specifically, SL1344 treated with different concentrations of Nar (0 and 32 μg/mL) was harvested and suspended in PBS, and 10-μL samples were applied to carbon-coated copper grids and incubated for 5 minutes at room temperature. Afterwards, the bacteria were treated with 0.5% phosphotungstic acid for ten seconds after the excess liquid was removed. After drying, the samples were observed under a transmission electron microscope (AX-FNSP, Nikon) (Dingle et al., 2011).

J774A.1 mouse macrophages were seeded into 6-well cell culture plates (1 × 10⁶ cells/well) and cultured overnight. SL1344 with or without Nar (32 μg/mL) was used to infect cells (MOI = 20) for four hours. The supernatant was collected after centrifugation (12,000 rpm, 5 minutes). The levels of tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β) and interleukin-6 (IL-6) were detected using an ELISA kit (Sangon Biotech, Shanghai, China) to measure the ability of Nar to mitigate bacteria-mediated inflammatory responses (Tabatabaei and Ahmed, 2022).

SL1344 was seeded into 96-well culture plates to a final concentration of 2 × 10⁷ CFUs/mL; after different concentrations of Nar (0, 16, and 32 μg/mL) were added, the plates were incubated statically at 30 °C for 48 hours. The medium was subsequently discarded, and the samples were dried after being washed with PBS and then treated with crystal violet (0.1%, 200 μL/well) for twenty minutes. The OD570 values were measured after the samples were treated with glacial acetic acid (33%, 150 μL/well) for fifteen minutes (Hassan et al., 2011).

The proteins used here were constructed, expressed and purified via methods described previously (Soleymani et al., 2017; Aburto et al., 2019). Briefly, the fimZ whole fragment of SL1344 was obtained by polymerase chain reaction (PCR), after which the fragment was digested with restriction enzymes. Subsequently, the target gene DNA was inserted into pET-28a by using T4 DNA ligase (TransGen Biotech, Beijing, China) and transferred to Escherichia coli BL21 (DE3) strains (TransGen Biotech, Beijing, China). Protein expression was induced with 0.2 mM isopropyl-D-1-thiogalactopyranoside (Sigma–Aldrich). The purified proteins were harvested with 200 mM imidazole (Sigma–Aldrich). The FimZ mutants were obtained with the same method. The restriction enzymes used for this assay were BamHI and SalI. The primers used are listed in Supplementary Table S4.

All mouse infection models used in this work followed the rules of the Laboratory Animal Ethics Committee of Jilin University. This assay was performed as previously described with some modifications (Shu et al., 2021; Sheng et al., 2023). Female BALB/c mice (approximately 20 g, aged 6–8 weeks) were obtained from Liaoning Changsheng Biotechnology Co., Ltd. Before infection, the mice were treated with 5 mg/mL streptomycin in water for three days to ensure a clean intestinal environment. For the survival assay, each mouse received 100 μL of SL1344 suspension (1 × 10⁸ CFUs/mL) by gavage, and ten mice were assigned to each group. Two hours after infection, 100 mg/kg Nar was administered to each mouse via subcutaneous injection (twice per day), and the infection group or the blank control group received an equal volume of solvent or sterile PBS. The survival of the mice was observed every day to evaluate the protective effect of Nar against infection by SL1344. For other analytical indicators, each mouse was treated with 5 × 10⁶ CFUs of bacteria via the same therapeutic method. Four days later, the blood was harvested, and the mice were euthanized (cervical dislocation) after anaesthesia via pentobarbital sodium injection (30 mg/kg). Afterwards, the liver, spleen and caecum were harvested. The clones in the liver and spleen were statistically analyzed after culture on LB agar plates for 18 hours. Serum levels of cytokines were detected to evaluate the remission effect of Nar.

The data are presented as the means with standard deviations (SDs) of three independent experiments. Statistical analysis was carried out via an unpaired t test in GraphPad Prism 9.5.0. Significance was defined as p ≤ 0.05.

After treatment with 16 μg/mL or 32 μg/mL Nar (Figure 1a), the percentage of SL1344 invading HeLa cells decreased to 75.82% and 43.41%, respectively (Figure 1b). The MIC of Nar against SL1344 was greater than 128 µg/mL (Supplementary Table S5). The growth curves revealed that Nar does not affect the growth of SL1344, as the bacteria showed similar growth when treated with or without Nar (Figure 1c). The amount of LDH did not differ when HeLa or J774A.1 cells received different concentrations of Nar (Figures 1d, e), suggesting that Nar does not have cytotoxic effects. The results of the immunofluorescence assay revealed that the number of bacteria observed in the Nar treatment groups was much lower than that in the control group (Figures 1f, g), further confirming the inhibitory effect of Nar against SL1344 invasion of HeLa cells. These results suggest that Nar has no antibacterial properties or cytotoxicity but significantly inhibits SL1344 invasion of HeLa cells.

Flagella-mediated motility and TIFs affect bacterial invasion of host cells. To clarify the specific target of Nar, a motility assay was carried out, and the diameter of the samples did not significantly differ when SL1344 was treated with or without Nar (Figures 2a, b), indicating that invasion was not reduced by changes in the flagella. We subsequently evaluated the effect of Nar on the gene expression of TIFs by RT–qPCR, and we found that the levels of fimA, fimH, fimC and fimD decreased to 43.66%, 34.67%, 53% and 56%, respectively, when SL1344 was subjected to 32 μg/mL Nar treatment, but the levels of fimY and fimZ, which are regulatory genes, did not change (Figures 2c, d). These results indicate that Nar may affect TIFs to reduce the SL1344 invasion of HeLa cells.

The transcription levels of TIF component genes were decreased after Nar treatment, but those of regulatory genes did not change. We speculated that Nar may target regulatory proteins to affect the function of TIFs. Therefore, we performed molecular docking to explore the potential target of Nar and found that it bound to the binding pocket of FimZ with an affinity of -7.1 ± 0.39 kcal/mol. To confirm the reliability of the combination, we carried out a dynamic simulation assay and found that the RMSD values of FimZ and Nar fluctuated around 0.196 ± 0.02 nm and 0.056 ± 0.02 nm, respectively, during the simulation, which revealed that they maintained stable structures during the process (Figure 3a). The RMSD of Nar relative to the backbone of FimZ fluctuated around 0.485 ± 0.12 nm (Figure 3a), indicating that it remained in the original binding pocket, which was evidenced by the structural overlay of different frames in the trajectory (Figure 3b). The distance between Nar and FimZ fluctuated around 0.32 ± 0.02 nm over time (Figure 3c), confirming that the binding was stable. The binding free energy between Nar and FimZ was -79.08 ± 3.90 kJ/mol, which was attributed to the electrostatic force (ele; -17.74 ± 3.56 kJ/mol), van der Waals force (vdw; -133.77 ± 1.17 kJ/mol) and solvation energy (sol, 72.43 ± 2.52 kJ/mol), suggesting that vdw plays a critical role in promoting binding, which was confirmed by the results of the weak interaction analysis (Figure 3d). Residue energy decomposition revealed that 97ARG, 137ASN, 134PHE, 138THR, 135ILE, 94ARG, 142LYS, 96ILE, and 139ARG in FimZ contributed more energy to binding (Figure 3e). 97ARG, 137ASN, 134PHE and 138THR contributed more vdw and ele and presented shorter distances to Nar (Figure 3f). Taken together, these results indicate that vdw and ele are the main interactive forces holding Nar in the binding pocket of FimZ; this interaction originates mainly from 97ARG, 137ASN, 134PHE and 138THR.

To confirm whether the FimZ fusion protein maintained its active construct, an EMSA was carried out, and we found that the FimZ fusion protein bound to the promoter of fimA, but the binding was inhibited by Nar (Supplementary Figure S2). These results confirmed the activity of the FimZ fusion protein, and Nar inhibited its binding to the fimA promoter. To identify the critical binding sites, we carried out residue mutation and SPR assays (Figure 4a). The equilibrium dissociation constant (KD) values for the wild-type (WT), I135A and T138A strains were 3.392 × 10–⁵ mol/L, 1.523 × 10–⁴ mol/L and 4.120 × 10–⁴ mol/L, respectively (Figures 4b–g), suggesting that the mutation of 135ILE and 138THR affects the affinity between Nar and FimZ. However, the response signals were too weak to calculate KD values between Nar and F134A, N137A or R97A (Figures 4h–j). These results confirmed that ILE135, THR138, 97ARG, 137ASN and 134PHE in FimZ are important for promoting its binding to Nar, especially 97ARG, 137ASN and 134PHE.

Many TIFs formed on the SL1344 strains, while the bacteria changed to oval shapes and were sterile after receiving the 64 μg/mL Nar treatment (Figure 5a). The adherence of SL1344 to HeLa cells decreased to 64.43% and 35.22% when 16 μg/mL or 32 μg/mL Nar was used, respectively (Figure 5b). The levels of IL-1β, IL-6 and TNF-α decreased to 410.69 pg/mL, 175.70 pg/mL and 790.4 pg/mL, respectively, when each sample received 32 μg/mL Nar treatment (Figures 5c–e), and the formation of the biofilm was reduced to 42.24% (Figure 5f). These results suggest that Nar binds with FimZ to inhibit the formation of TIFs and affects the underlying phenotype.

SL1344-infected mice that did not receive Nar treatment died three days after infection, and when the infection time increased to seven days, the survival decreased to zero. However, in the Nar treatment group, no dead mice were detected until 5 days after infection, and the final survival of this group was 43.33% (Figure 6a). The logarithm of the bacterial load in the liver and spleen of the mice decreased by 1.49 and 0.79, respectively, in the Nar treatment group (Figure 6b). The serum levels of TNF-α and IL-1β decreased by 312.95 pg/mL and 127.85 pg/mL, respectively, after treatment with Nar (Figure 6c). The intestinal tissues harvested from the control group showed a normal tissue status, but those from the infection group were filled with haemorrhages and were tenuous; although the intestinal tissues from the Nar treatment group did not recover to the status those of the control group, the haemorrhages disappeared, and the status was much better than that of the infection group (Figure 6d). In the control group, the liver and spleen samples showed normal coloration and dense tissues; samples from the infection group were swollen and less dense, and the spleen was black, whereas these symptoms were alleviated in samples from the Nar treatment group (Figure 6e). These results indicate that Nar can protect mice against SL1344 infection.