Colistin (88%) was purchased from Hebei Shengxue Dacheng bio-pharmaceutical Co., Ltd. (Hebei, China), niclosamide (98%), cremophor EL (EL), ethyl oleate (EO), isopropyl myristate (IPM), Tween-80, Tween-20, and Poloxamer 188 were commercially available from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). Poloxamer 407 (P407) was purchased from Accela ChemBio Co. Ltd. (Shanghai, China). Castor oil, octyl phenol, and polyoxyethylene ether-10 (OP) were purchased from Shanghai Yuanye Biochemical Co., Ltd. (Shanghai, China). Cremophor RH-40 was provided by Jiangsu Haian Petrochemical Plant (Jiangsu, China). Isopropanol, ethanol, propylene glycol, and polyethylene glycol 400 (PEG400) were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All other chemicals and reagents were of analytical grade. Ten non-duplicate clinical COL-resistant Salmonella strains were isolated from swine, chickens, and humans, specifically and preserved in the Laboratory of Pharmacology, Henan Agricultural University.

The checkerboard assays were conducted to explore the synergistic activity between COL and NIC by determining the fractional inhibitory concentration index (FICI) (14). Briefly, COL was continuously diluted 2-fold along row by row, while NIC was continuously diluted 2-fold along the column by column to establish a matrix in 96-well plates, where each well contained different concentrations of two drugs. The bacterial suspension (100 μL) was added to each well to reach an ultimate bacterial concentration of approximately 5 × 10⁵ colony-forming unit (CFU)/mL. The 96-well plates were then incubated at 37°C for 18 h and examined for visible turbidity.

The FICI was calculated according to the following formula:

FICI=FICI_A + FICI_B = (MIC_A in combination/MIC_A alone) + (MIC_B in combination/MIC_B alone).

FICI was interpreted as follows: FICI ≤0.5 (synergy), 0.5 < FICI ≤1 (additive effect), 1 < FICI≤2 (irrelevance), FICI >2 (antagonism) (28).

The solubility of NIC in a range of oils (EO, IPM, and castor oil), surfactants (RH-40, EL, OP, Tween-80, and Tween-20), and co-surfactants (ethanol, isopropanol, propylene glycol, and PEG400) were screened for the development of nanoemulsion. An excess amount of NIC was added to a clean glass vial with 1 mL of each oil, surfactant, and co-surfactant and vortexed for 30 s (25). Then, the sample was mechanically shaken in a water bath shaker at 37°C for 72 h to reach equilibrium (29). The equilibrated samples were centrifuged at 10000 rpm for 15 min. The amount of soluble NIC of each sample was determined using an HPLC (high-performance liquid chromatography) method and UV–vis spectroscopy as described below. The oils, surfactants, and co-surfactants with the highest NIC solubility were selected as the components of nanoemulsions. All excipients used in this study are injectable and comply with an FDA approval.

Nanoemulsion was prepared by low energy (spontaneous) emulsification method. Based on the solubility results above, the selection was made for oil, surfactants, and co-surfactants to construct a pseudo-ternary phase diagram using the water titration method at room temperature. The specific experimental steps are as follows: first, the effect of surfactants on the formation of nanoemulsion was observed. Co-surfactant (ethyl alcohol) is separately mixed with surfactant (EL, Tween-20, and OP) to form a surfactant/co-surfactant mixture (Smix), and the Smix ratio is 3:1. Pseudo-ternary phase diagram was constructed with different ratios of oil to Smix (1:9, 2:8, 3:7, 4:6, 5:5, 6:4, 7:3, 8:2, 9:1) with EO as an oil phase. Similarly, the effects of different co-surfactants on the formation of nanoemulsion were analyzed. Then, EL was used as the surfactant, and it was mixed with ethyl alcohol at different Smix (1:1, 2:1, 3:1, and 3:2) with EO as the oil phase to determine the maximum microemulsion region from four different ratios of Smix. Then, these samples were checked for transparency and uniformity after the gradual addition of water (30).

The microemulsion technique was optimized to produce blank nanoemulsions by adjusting the ratio of EO (oil phase), EL (surfactant), and ethyl alcohol (co-surfactant). The optimization process was guided by particle size and polydispersity index (PDI) as key indicators. Six different blank nanoemulsions (NE1 to NE6) were prepared (Supplementary Table S1). Particle size and PDI were measured by zeta-sizer Nano (ZS-90, Malvern Instruments, Malvern, UK). Using the optimized conditions of NE6, COL/NIC-NEs were prepared based on blank nanoemulsion formulations. Briefly, NIC (0.2% w/w) dissolved in EO (5%) was added to Smix (45% w/w) consisting of EL/ethyl alcohol (3:1) and stirred using a magnetic stirrer for 5 min. Finally, 5 mL of deionized water containing (0.4%) COL was added slowly with stirring to obtain a clear solution. A schematic diagram for the fabrication of COL/NIC-NEs and COL/NIC-NEGs is shown in Figure 1A.

The incorporation of nanoemulsion into a mixture of 30% P407 and 8% P188 was carried out by first preparing the poloxamer gel base. To do this, 30% P407 and 8% P188 powder were added to distilled water while stirring at a low speed. The mixture was then allowed to hydrate for 10 min before increasing the stirring speed too high for an additional 30 min to ensure proper gel formation. The prepared nanoemulsion was then added to the poloxamer gel at a concentration of 50% (v/v) and mixed until a homogenous gel was formed. We then allowed the mixture to sit for 24 h to allow for proper hydration and formation of the gel. Finally, we checked the gelation temperature and pH of the final product to ensure it met the desired specifications.

Fourier transform infrared spectrometer (Thermo Scientific Nicolet, iS20 FTIR, USA) was performed to check the chemical composition of the formulation. The specific parameters are based on previous studies (25). The measurement range is 350 to 4,000 wave numbers, the data interval is 0.5 cm⁻¹, and the resolution is 2 cm⁻¹. The figure is expressed as a relation between relative transmittance (%) and wave number (cm⁻¹).

A high-performance liquid chromatography system (Agilent Technology Co., Ltd) was used to detect the content of NIC according to the previously described method (32). NIC was separated by an Agilent TC-C18 column (5 mm, 4.6 mm–250 mm) guarded with a precolumn at 30°C and detected at 330 nm. The injection volume was set to 20 μL. The mobile phase consisted of 85% methanol and 15% water (containing 0.1% formic acid) pumped at a flow rate of 1.0 mL/min. An NIC stock solution was prepared in methanol (1 mg/mL) and sequentially diluted with methanol to 100, 80, 60, 40, and 20 μg/mL. A standard curve was established, and the content was calculated using the peak area of the sample at the retention time (6.8 min).

COL content was determined using UV–vis spectroscopy. COL solution (1 mg/mL in acetonitrile) was diluted with distilled water to 500, 250, 125, 62.5, 1, 0.5, and 0.25 μg/mL, and the absorption of the sample at 230 nm was determined using a UV–vis spectrometer. A standard curve was established, and the content was calculated using the absorption value of the sample.

In vitro release studies of COL and NIC from COL/NIC-NEs and COL/NIC-NEGs were carried out at 37°C via dialysis as previously described (33). The COL/NIC-NEs and COL/NIC-NEGs (1 mL) were transferred to dialysis bags (10 kDa MWCO), respectively, and then immersed in 20 mL of PBS (pH 7.4) and kept in a shaking water bath at 37°C and 100 rpm. At a predetermined time (0.5, 1, 1.5, 2, 4, 6, 8, 10, 12, 14, 24, 36, 48, 60, 72, 84, and 96 h), the 0.5 mL COL/NIC-NEs and COL/NIC-NEGs in the dialysis bag was taken out and refilled with PBS of equal volume after each sampling. The accumulative release of NIC and COL from COL/NIC-NEs and COL/NIC-NEGs was determined by HPLC and UV–vis spectroscopy. All experiments were performed in triplicate.

The MICs of the COL/NIC-NEs and COL/NIC-NEGs to the 10 COL-R strains were measured by broth microdilution method according to the Clinical and Laboratory Standards Institute guidelines (34), and E. coli ATCC 25922 was used as the quality control. All drugs were diluted using 2-fold serial dilutions in a sterile 96-well microtiter plate with Mueller-Hinton broth (MHB), and then, 100 μL of the bacterial suspensions (1 × 10⁶ CFUs/mL) was added to each well. MIC values were defined as the lowest concentrations of drugs with no visible growth of bacteria after 18 h of incubation at 37°C.

To further investigate the bactericidal activity of the COL/NIC-NEs and COL/NIC-NEGs, the time–kill curves of randomly selected clinical COL-resistant Salmonella SH134 (mcr-1 positive) isolate and S2a (mcr-1 negative) were determined during the exponential growth phase as described with some modifications (14). Bacterial cells at a final concentration of 1 × 10⁵ CFUs/mL were cultured with MHB in NIC (1 μg/mL), COL (2 μg/mL), blank-NEs, and the COL/NIC-NEs and COL/NIC-NEGs (containing COL and NIC at 2 and 1 μg/mL, respectively). Samples of the bacterial cells were collected after culturing for 0, 2, 4, 8, 12, and 24 h and spread on LB agar plates after dilution. CFUs were calculated after incubation at 37°C overnight. All experiments were repeated three times on different days.

To investigate the antibacterial mechanisms of COL/NIC-NEs and COL/NIC NEGs, we conducted fluorescence experiments. Bacterial pretreatments in all measurements were performed with similar protocols as follows. In the fluorescence assay, the mcr-1-positive strain SH134 was chosen as the indicator strain. Briefly, bacteria were grown overnight at 37°C with shaking at 200 rpm. Then, the cultures were washed and suspended with 5 mmol/L HEPES (pH 7.0, plus 5 mmol/L glucose). The OD₆₀₀ of bacteria suspension was standardized to 0.5 in the same buffer, and fluorescent dye was added. After incubation at 37°C for 30 min, an aliquot of 1 mL of bacterial suspension was mixed with the same concentration of free COL, free NIC, free COL and NIC, blank-NEs, COL/NIC-NEs, and COL/NIC-NEGs. After incubation for 1 h, 200 μL of bacterial cells was added to the 96-well plate. Subsequently, fluorescence intensity was measured on a Spark 10 M Microplate reader (Tecan) (35).

Fifty-six KM mice were randomly divided into seven groups (n = 8) for survival experiments. All mice were treated with streptomycin by gastric administration 3 days before the formal experiment. Each group was infected with Salmonella SH134 at 1.0 × 10⁷ CFUs by intraperitoneal injection administration in addition to the blank control group. Each group was intraperitoneally injected once a day for 3 days with the following methods: PBS, model, free COL (5 mg/kg), free NIC (5 mg/kg), free COL+NIC (COL 5 mg/kg, NIC 2.5 mg/kg), and COL/NIC-NEs (COL 5 mg/kg, NIC 2.5 mg/kg). Particularly worth mentioning is that the COL/NIC-NEGs (COL 10 mg/kg, NIC 5 mg/kg) group was treated by injection only on the first day. The survival rate of mice in each group was observed for up to 5 days.

Another six groups (n = 6) with a sublethal dose of Salmonella SH134 at 1.0 × 10⁶ CFUs received the same treatment as described above. Colony colonization and histopathology of the liver and spleen were studied. All mice were killed by cervical dislocation on the second day. The tissue samples were ground for continuous dilution and plated on SS agar plates to observe the number of colonies. The typical lesions of livers or spleens of different groups of mice were prepared by hematoxylin–eosin (HE) staining, and histopathological analysis was performed.

All experiments were performed in triplicate. Statistical analysis was performed using GraphPad Prism 8.0. All values are stated as mean ± SD. Unpaired t-tests (normally distributed data) between two groups and one-way ANOVA among multiple groups were used to calculate p-values. Statistical significance was determined at *<p 0.05, **p<0.01, and ***p<0.001; ns denotes no significant difference.

Checkerboard assays were used to evaluate the antibacterial activity of a combination of NIC and COL against COL-resistant Salmonella (13). As shown in Figure 2A, a synergistic effect was observed in mcr-1-positive and mcr-1-negative strains. NIC robustly reversed COL resistance, with the MIC of COL decreased to a sensitivity of no more than 2 μg/mL, with FIC indices lower than 0.5 for all the tested resistant strains (FICI = 0.0195–0.129) (Figure 2B and Table 1). Importantly, the MIC of COL in combination ranging from 0.0625 to 1.0 μg/mL was lower than the resistance breakpoint, which was set by the Clinical and Laboratory Standards Institute at 4 to 8 μg/mL for Gram-negative bacteria. These results demonstrated that the NIC can reverse the resistance phenotype of COL-resistant Salmonella.

The drug absorption and bioavailability of NIC when administered in combination with COL in vivo may be limited by its poor solubility. The selection of oil phase, surfactant, and co-surfactant with high solubility is the basis of the successful preparation of nanoemulsion. Oil phase drug solubilizing capacity is a crucial characteristic as it affects the potential of formulation for drug loading. As shown in Figure 3A, among three oil phases, the solubility of NIC in EO (1.54 ± 0.04 mg/mL) is a little more than that in IPM (1.49 ± 0.06 mg/mL) and castor oil (1.52 ± 0.06 mg/mL). Moreover, due to the advantages of strong permeability, easy absorption, high stability, good safety, and low toxicity of EO, EO was selected for further investigation for nanoemulsion studies as oil bases (38). Surfactant EL had the highest solubility (10.92 ± 0.64 mg/mL) of NIC among the different surfactants, followed by Tween-20 (2.63 ± 0.05 mg/mL), OP (1.77 ± 0.1 mg/mL), and Tween-80 (1.35 ± 0.05 mg/mL). Among the different co-surfactants, PEG400 and ethyl alcohol showed higher solubility than the isopropanol and propylene glycol. Thus, we selected oil phase (EO), surfactant (EL, Tween-20, and OP), and co-surfactant (PEG400, ethyl alcohol, and isopropanol) for subsequent pseudo-ternary phase diagram screening.

The pseudo-ternary phase diagram, which compares the microemulsion region, is an effective evaluation technique (39). As shown in Supplementary Figure S1, surfactant EL had the largest microemulsion area compared with Tween-20 and OP, indicating that EL is a surfactant suitable for NIC nanoemulsions. According to Supplementary Figure S2, a comparison of the microemulsion region of ethyl alcohol, PEG400, and isopropanol, with the biggest microemulsion region, was chosen as the co-surfactant for NIC nanoemulsions. Surfactant/co-surfactant (Smix) ratio plays a very important role in preparing more stable and effective nanoemulsions. As shown in Figure 3B, pseudo-ternary phase diagrams were constructed by using different Smix ratios (1:1, 2:1, 3:1, and 3:2). The results show that the microemulsion region is the largest when the Smix ratio is 3:1, which indicates that the formation of nanoemulsion is the strongest, so it can be used in the final formulation. Based on the pseudo-ternary phase diagrams, EO, EL, and ethyl alcohol were chosen, respectively, as oil phase, surfactant, and co-surfactant of NIC nanoemulsions.

Blank nanoemulsions were optimized based on different proportion factors in the prescription. Significant changes were observed by changing these prescriptions in different proportions. The final formulation was prepared based on the optimized blank nanoemulsion formula (NE1 to NE6). As shown in Figure 3C and Supplementary Table S2, NE6 has the characteristics of small particle size and small PDI, compared with other formulations, so NE6 is selected as the final formulation of the COL/NIC-NEs.

The COL/NIC-NEs were prepared according to the optimized formulation NE6. The preparation process of COL/NIC-NEs and COL/NIC-NEGs is illustrated in Figure 1A. The appearance of the COL/NIC-NEs as shown in Figure 1B was yellow, clear, and transparent liquid. The mean particle sizes, PDI, and zeta potentials of the COL/NIC-NEs were measured using DLS. The average hydrodynamic diameter of the COL/NIC-NEs droplets was about 19.86 ± 2.21 nm, with a PDI of 0.103 ± 0.01 (Figure 1C), and the zeta potential of the COL/NIC-NEs was −1.25 mv (Figure 1D), indicating that the nanoemulsions had satisfactory quality results. From the images obtained by TEM analysis, it is evident that COL/NIC-NEs showed spherical, non-adhesive, non-aggregating, and evenly distributed in space, the sizes of the nanodroplets corresponded to the ones measured with DLS, as evident from Figure 1E.

To investigate the stability of the COL/NIC-NEs, we conducted centrifugation stability, temperature stability, and long-term storage stability tests. The centrifugal stability study found that after centrifugation, the COL/NIC-NEs did not layer or precipitate, and remained a yellow clear transparent liquid with no significant changes in particle size and PDI, indicating that the formula has good centrifugal stability (Figure 1F). In the study of temperature stability (Figure 1G), the COL/NIC-NEs remained stable at low temperatures of −20°C, refrigerated temperature of 4°C, room temperature of 25°C, and 60°C, exhibiting minimal changes in particle size and PDI. However, after treatment at 80°C, the particle size of COL/NIC-NEs significantly increased from 19.34 nm to 112.6 nm, indicating that the COL/NIC-NEs are unstable at 80°C and high temperatures can disrupt the dispersion system of the nanoemulsion. Long-term storage stability is a necessary precondition for the practical application of nanoemulsions. As shown in Figure 1H, the long-term stability experiment found that the COL/NIC-NEs system remained relatively stable within 1–7 days, with a particle size of around 20 nm. Over longer storage periods, it was found that the particle size slowly increased. However, despite the rise, the sizes of COL/NIC-NEs were still in the acceptable range of nanoemulsion.

Following the optimization studies, a thermal sensitivity situ gel loaded with COL/NIC-NEs was prepared. Then, the evaluation of the COL/NIC-NEGs of poloxamer 407 and 188 for temperature was carried out at two different temperatures, i.e., 2–8°C and 36–38°C, respectively, to verify the thermosensitivity of the in situ gels. As shown in Figure 1I, the COL/NIC-NEGs remained at a transparent solution state at 2–8°C and the solution state of the COL/NIC-NEGs transformed into the gel state at 36–38°C. The pH of the COL/NIC-NEGs measured by the PH meter is 6.51, which meets the requirements of injection preparations. The results show that the prescription conditions of the COL/NIC-NEGs are reasonable and feasible.

As shown in Figure 4A, the COL spectrum (black) showed characteristic peaks at 1657 cm⁻¹ (−C=O amide bonds) and 2,958 cm⁻¹ (−CH stretching). The characteristic FTIR spectrum of NIC (red) signaling includes bending vibration of phenolic hydroxyl groups (−C − O), −NO2 bending, and − C=O stretching bending, which occur at 1327–1329, 1,508–1,510, and 1,652–1,653 cm⁻¹, respectively. Major peaks of the COL/NIC-NEs and COL/NIC-NEGs appeared at wave numbers 2,924–2,925, 1,641–1,643, 1,456–1,460, and 1,349 cm⁻¹. The FTIR spectrum of COL/NIC-NEs (blue) and COL/NIC-NEGs (green) showed lower peak intensities, which suggests a good blend of COL and NIC with the selected excipients without the presence of chemical interactions.

To investigate the extended-release effect, we evaluated the NIC and COL-release from the COL/NIC-NEs and COL/NIC-NEGs at 37°C in PBS (pH 7.4) (Figure 4B). The release profile demonstrated an initial burst release of free COL and NIC within the 6 h (96.63 ± 2.31% and 61.82 ± 1.85%, respectively). The release rate of COL in COL/NIC-NEs and COL/NIC-NEGs was 88.35 ± 0.38% and 82.88 ± 1.87%, respectively, at 96 h. Similarly, the release rate of free NIC was 75.89 ± 0.56%, that of COL/NIC-NEs was 64.31 ± 0.52%, and that of COL/NIC-NEGs was 56.92 ± 3.19%. After fitting, COL and NIC in COL/NIC-NEs conform to the first-order release model (adjusted R² = 0.91923 and 0.98597, respectively), and the equation that COL and NIC in COL/NIC-NEGs conform to is the Higuchi model (adjusted R² = 0.98644 and 0.97862, respectively) (Supplementary Figure S3). As expected, the release of drugs NIC and COL from COL/NIC-NEs and COL/NIC-NEGs was slower than the free NIC and COL. It shows that NIC- and COL-loaded nanoemulsions and nanoemulsion gels have slow-release effects, achieving the effect of reducing the amount and increasing the efficiency.

To study the antibacterial effects of COL/NIC-NEs and COL/NIC-NEGs, we examined the MIC of the COL, NIC, COL/NIC-NEs, and COL/NIC-NEGs, respectively (Supplementary Table S2). Compared with alone COL, the MIC of COL in the combination of the COL/NIC-NEs and COL/NIC-NEGs was decreased significantly, which was reduced by 16-128-fold (Figure 5A). Excitingly, a synergistic effect in the COL/NIC-NEs and COL/NIC-NEGs was observed in all the tested mcr-1-positive and mcr-1-negative COL-resistant strains. To further evaluate the synergetic effect of the COL/NIC-NEs and COL/NIC-NEGs, we performed the time–kill curves assays using mcr-1-positive strain SH134 and mcr-1-negative strain S2a as a pattern strain. The results suggested that the COL/NIC-NEs and COL/NIC-NEGs exhibited synergistic bactericidal efficacy. Specifically, neither alone NIC nor COL killed Salmonella SH134 and Salmonella S2a in the medium. However, the COL/NIC-NEs and COL/NIC-NEGs decreased CFUs of Salmonella SH134 and Salmonella S2a approximately by >2-log₁₀ in 24 h compared to using any drug alone (Figure 5B). Furthermore, the COL/NIC-NEs and COL/NIC-NEGs can enhance the bactericidal activity of COL compared with free COL+NIC (Figure 5B; Supplementary Figure S4). The enhanced antibacterial activity of the COL/NIC-NEs and COL/NIC-NEGs was further confirmed by examining the bacterial surface morphology using scanning electron microscopy (SEM) (Figure 5C). SEM images showed rough bacterial surfaces or lytic bacterial cells destroyed and damaged by the COL/NIC-NEs and COL/NIC-NEGs treatment (arrows in Figure 5C). In contrast, the cell membrane of bacterial cells remained relatively intact after treatment with the free drug combination or alone. In conclusion, both the COL/NIC-NEs and COL/NIC-NEGs significantly enhance the antibacterial activity of the drug combination.

The enhanced antibacterial activity of the COL/NIC-NEs and COL/NIC-NEGs delivery system may be due to the enhanced interaction of the encapsulated drugs with the bacterial cell membrane (40). To verify whether the enhanced antimicrobial activity is related to cell membrane interactions, we performed fluorescent probe assays. First, we investigated the changes in bacterial membrane biochemical parameters treated with the COL/NIC-NEs and COL/NIC-NEGs. NPN and PI were used to evaluate bacterial outer membrane permeability and bacterial cell membrane integrity, respectively. DiSC3 (5) and BCECF-AM were used to detect the electric potential (Δψ) and transmembrane proton gradient (ΔpH) (41). We found that the COL/NIC-NEs and COL/NIC-NEGs increased outer membrane permeability, and whole membrane permeability, and led to the dissipation of PMF by disrupting Δψ and ΔpH. In contrast, free COL did not induce significant changes in the fluorescence intensity of NPN, possibly attributed to the weakened interaction between modified lipopolysaccharide and COL in COL-resistant bacteria (mcr-1 positive). In addition, the fluorescence intensity of NPN and PI was significantly enhanced in the groups of COL/NIC-NEs and COL/NIC-NEGs compared to the free COL and NIC groups. This suggests that the COL/NIC-NEs and COL/NIC-NEGs can improve the fusion with cell membranes, thereby enhancing cell membrane permeability (Figures 6A,B). The PMF is that it consists of two components including electric potential (Δψ) and transmembrane proton gradient (ΔpH), which continually work together to maintain homeostasis (42). We measured the fluorescence intensity of the fluorescent probes DiSC3 (5) and BCECF-AM in the tested strains, respectively. The fluorescence intensity of the COL/NIC-NEs and COL/NIC-NEGs is significantly enhanced, indicating that the dissipation of PMF is accelerated by destroying the membrane potential and ΔpH (Figures 6C,D).

It has been shown that NIC can act as an efflux pump inhibitor to reverse the resistance of Klebsiella pneumoniae and Acinetobacter baumannii (21), thus promoting the retention of COL in the cell and enhancing their antibacterial activity. In order to investigate whether NIC inhibits drug efflux and thus improves the antibacterial activity of NIC against COL-resistant bacteria, the efflux effect of COL-resistant Salmonella on EtBr was studied. Compared to the control group without any treatment, both free NIC and free combination of NIC and COL increased fluorescence intensity, indicating an inhibitory effect on the efflux pump of EtBr (Figure 6E). More importantly, the fluorescence intensity of the COL/NIC-NEs and COL/NIC-NEGs was significantly increased, suggesting that they can enhance antimicrobial activity by inhibiting the activity of the efflux pump.

Studies have reported that various antibiotics can affect different cellular targets by causing a reactive oxygen species (ROS)-mediated disturbance in metabolism and respiration (43, 44). ROS is a common mechanism of antibacterial agents, and it can promote cell oxidative damage by increasing ROS generation. Therefore, we sought to detect changes in ROS levels in different drug combinations. The results indicated that the free NIC and COL combination group led to increased ROS levels better than the COL monotherapy group. Interestingly, the ROS levels significantly increased in the COL/NIC-NEs and COL/NIC-NEGs compared with the free NIC and COL combination group (Figure 6F). Thus, the above results showed that oxidative stress of bacteria cells exposure to the COL/NIC-NEs and COL/NIC-NEGs led to increasing intracellular ROS levels.

The gastrointestinal tract serves as a critical site for the colonization of mcr-1-positive Salmonella, and these bacteria have been isolated from patients with intra-abdominal infections. To investigate the synergistic antibacterial activity and therapeutic potential of COL/NIC-NEs and COL/NIC-NEGs in vivo, a mouse model of abdominal infection was established through the injection of clinically isolated COL-resistant Salmonella (Figure 7A).

To that end, we tested the in vivo efficacy of this combination in mouse infection models with Salmonella SH134. As shown in Figure 7B, treatment with alone NIC and COL did not exert adequate protection in the infected mice. Excitingly, administration of COL/NIC-NEs and COL/NIC-NEGs significantly improved the survival rate (p < 0.001), with survival increasing from 37.5% (COL) to 87.5 and 75%, respectively. In particular, when compared to the free COL+NIC group, COL/NIC-NEs and COL/NIC-NEGs were able to further boost the survival rates of mice at an equivalent dose, thus further corroborating that COL/NIC-NEs and COL/NIC-NEGs can potentiate the therapeutic efficacy of COL and NIC. Similarly, the expected efficacy of combination therapy in vivo was also observed in the bacterial load in the liver and spleen of mice (Figure 7C). More specifically, the combinational therapy with COL/NIC-NEs and COL/NIC-NEGs exhibited approximately a 2-log₁₀ reduction in colony-forming units (CFUs) compared to treatment with COL and NIC alone.

Finally, the therapeutic effectiveness of this combination was evaluated through histopathological examination of liver and spleen tissues (Figure 7D). HE staining demonstrated prominent cell necrosis in hepatocytes of the model group and in the groups treated exclusively with COL and NIC, respectively. However, a protective effect on the liver was observed in the treatment groups of COL/NIC-NEs and COL/NIC-NEGs. In the spleen sections, the demarcation between the white and red pulp was unclear in the model group, the NIC group, and the COL group. In particular, compared to the model group, the NIC group, and the COL group, the spleens of both the COL/NIC-NEs and COL/NIC-NEGs groups exhibited normal histological features. Overall, these findings indicated that COL/NIC-NEs and COL/NIC-NEGs significantly safeguarded mice against COL-resistant Salmonella infection.