As shown in Fig. 2, we first synthesized the intermediate 3 from 9,10-phenanthroquinone (1) and 4-benzenedicarboxaldehyde (2) in the presence of NH₄OAc by the cyclization reaction, followed by the reduction of aromatic nitro group to amine and then the quaternization reaction of compound 4 with methyl iodide to obtain N,N,N-trimethyl-4-(1H-phenanthro[9,10-d]imidazol-2-yl)benzenaminium iodide (5a). Subsequently, to increase the hydrophobicity of 5a, we carried out an alkyl substitution at the N-1 position of intermediate 3 and then obtained compounds 8a and 8b using a similar protocol for the synthesis of 5a. Interestingly, we found that the use of acetonitrile as a solvent in the quaternization of compounds 7a and 7b gave products 9a and 9b instead of 8a and 8b. It has been widely reported that amino groups have a great influence on the antibacterial activity of compounds [39,40].

To further investigate the effect of an amino group on phenanthro[9,10-d]imidazole compounds as well as to increase their structural diversity, we also prepared compound 12a using 9,10-phenanthroquinone (1) and ethyl 4-formylbenzoate (10) as the raw material by cyclization and amination reactions. Similarly, we further synthesized compounds 14a and 14b with alkyl substitution at the N-1 position of the imidazole ring (Fig. 3).

Additionally, to increase the water solubility of compounds 12a, 14a, and 14b in favor of evaluation of their antibacterial activities, we further prepared hydrochlorides 18a, 20a, and 20b of compounds 12a, 14a, and 14b (Fig. 4). The structures of all novel phenanthro[9,10-d]imidazole compounds 5a, 8a, 8b, 9a, 9b, 12a, 14a, 14b, 18a, 20a, and 20b were characterized by ¹H/¹³C nuclear magnetic resonance (NMR), and high-resolution mass spectrometry (HRMS) (see the Supplementary Materials).

In vitro antibacterial and hemolytic properties were assessed for the synthesized phenanthro[9,10-d]imidazole derivatives. Their antibacterial efficacy against a panel of gram-positive and gram-negative bacterial strains was initially examined using a 2-fold serial dilution method following Clinical and Laboratory Standards Institute guidelines [41]. The results indicated that compound 9b exhibited pronounced antibacterial activity against both G+ and G− bacteria, with minimum inhibitory concentration (MIC) values ranging from 0.5 to 64 μg/ml, showing particularly strong activity against S. aureus ATCC 29213. To further evaluate cytotoxicity and membrane selectivity, we conducted hemolytic assays to determine the half hemolytic concentration (HC₅₀) values of these compounds. Compound 9b displayed minimal hemolytic activity toward sheep erythrocytes and demonstrated excellent selectivity for S. aureus, yielding a high selectivity index of 744.2 (Table 1).

Encouraged by the good antibacterial activity of compounds 5a, 8a, 8b, 9a, and 9b against S. aureus, we further screened their antibacterial activities against 10 clinical MRSA isolates. Interestingly, compound 9b also displayed robust antibacterial activity against these clinical MRSA isolates with MIC values in the range of 0.5 to 1 μg/ml (Table 2), which is comparable to the first-line drug vancomycin. The antibacterial activity screening results of 5a, 8a, 8b, 9a and 9b (Table 2) showed that the MIC of 9a and 9b was 8-fold or 32-/64-fold higher than that of the reference compounds 8a and 8b, suggesting that quaternization occurs at position N-3 of the imidazole ring to yield a more active compound than at the amino position of the benzene ring.

Given the excellent in vitro anti-MRSA activity and low hemolytic effect, 9b was selected as a candidate for further studies. To evaluate its bactericidal properties, we first investigated the time-kill kinetics against S. aureus ATCC 29213 and clinical MRSA isolates. As a result, 9b was effective in killing bacteria, and it completely eliminated S. aureus ATCC 29213 and MRSA-16 in exponential growth phase within 0.5 and 2 h at 8× MIC, respectively, superior to the reference drug vancomycin (Fig. 5A), and also displayed impressive bactericidal properties even at 4× MIC on completely killing eradicated S. aureus ATCC 29213 and MRSA-16 within 0.5 and 2 h, respectively (Fig. 5A). Additionally, the development of resistance to compound 9b against S. aureus ATCC 29213 was investigated at sub-MIC (1/2 MIC). Figure 5B shows that the MIC values of 9b remained unchanged within 20 passages of S. aureus ATCC 29213 treated by the sub-MIC level of 9b, indicating that 9b is not susceptible to induce resistance to S. aureus ATCC 29213. In contrast, development of resistance to the reference drug norfloxacin was apparently observed, with a 128-fold increase in MIC values at the 20th passage. Here, we hypothesized that the less susceptibility of compound 9b to resistance may be due to its rapid bactericidal capacity and membrane-targeting mode.

The above mechanistic findings demonstrated that compound 9b specifically targeted the phospholipid component PG in bacterial membranes and subsequently interacted with bacterial DNA, exerting potent antibacterial activity through a dual-action mechanism. Given this mode of action, ensuring favorable in vitro and in vivo safety and plasma stability is critical for its therapeutic potential.

To provide exposure context for the in vivo efficacy evaluation, we determined the pharmacokinetic properties of 9b in rats following intravenous administration (Table 3). After intravenous dosing (2 mg/kg), 9b reached a maximum plasma concentration (Cmax) of 2,173.33 ng/ml within 5 min and exhibited a terminal half-life (t₁/₂) of 5.81 h, with an area under the plasma concentration–time curve from time zero to infinity (AUC₀–inf) of 1,273.87 h·ng/ml. The calculated volume of distribution (Vdss) of 3.03 l/kg suggested that 9b is not confined to the plasma compartment and undergoes appreciable tissue distribution. These results provided important supportive information by confirming that 9b remains chemically detectable in vivo and possesses the capacity for tissue distribution. To further evaluate the blood compatibility of 9b, the morphology of sheep red blood cells was examined under microscopy after incubation with the compound. Even at a high concentration of 128 × MIC against MRSA-16, red blood cells treated with 9b showed no observable morphological changes compared with the blank control, maintaining intact membranes without signs of hemolysis. In contrast, the positive control (0.1% Triton X-100) caused complete erythrocyte lysis (Fig. 9A). Quantitative hemolysis analysis was conducted to generate dose–response curves over concentrations encompassing clinically relevant exposure ranges centered on the MIC (Fig. S6). The data showed that 9b caused minimal hemolysis within this exposure window, supporting its acceptable hemocompatibility under clinically relevant conditions.

Cytotoxicity assessment revealed that LO2 and NIH3T3 cells maintained >90% viability at 64 μg/ml of 9b, with half maximal inhibitory concentration (IC₅₀) values of 37.9 μg/ml and 32.5 μg/ml, respectively, and high therapeutic index (>65) (Fig. 9B). Confocal fluorescence microscopy further indicated that 9b entered the cytoplasm but showed no nuclear localization, as only SYTO 9-stained nuclei displayed strong green fluorescence, while 9b exhibited weak blue fluorescence confined to the cytoplasmic region (Fig. 9C). Importantly, this intracellular distribution did not compromise membrane integrity, as confirmed by lactate dehydrogenase release assays, where lactate dehydrogenase activity in 9b-treated cells remained comparable to that of the blank control and obviously lower than the Triton X-100-positive control (Fig. 9D). These findings collectively demonstrated that 9b selectively targeted bacterial cells while preserving mammalian cell membrane integrity and avoiding nuclear DNA interaction. Due to the presence of various hydrolytic enzymes in plasma, such as cholinesterases, lipases, and phosphatases, many compounds undergo rapid enzymatic hydrolysis, leading to a sharp decline in their plasma concentration and consequently a loss of antimicrobial efficacy [52–54]. As illustrated in Fig. 9E, the minimum bactericidal concentration (MBC) of 9b remained unchanged after treatment with plasma for 6 h, suggesting that 9b is not easily hydrolyzed by hydrolases and possessed commendable plasma stability. Furthermore, as shown in Fig. S7, the chemical stability of 9b in rat plasma was evaluated by monitoring the percentage of intact compound over time. The results showed that 9b remained largely intact in plasma, with approximately 92% of the compound remaining after 4 h. Kinetic fitting using a first-order degradation model yielded an apparent degradation rate constant (ke) of 0.0173 h⁻¹, corresponding to an estimated half-life (t₁/₂) of 40.17 h. In contrast, the reference compound propantheline showed substantially faster degradation under the same conditions, with only 48.5% remaining after 4 h and a calculated half-life of 3.85 h. These results indicated that 9b possesses favorable chemical stability in plasma compared with the reference compound. To further clarify the behavior of 9b in plasma, its plasma protein binding was determined in rat plasma. At a concentration of 4 μg/ml, 9b exhibited a free fraction (Fu) of 2.47% and a corresponding protein-bound fraction (Fb) of 97.53%, indicating substantial plasma protein binding. These results suggested that a large proportion of 9b exists in a protein-bound form in plasma. Although 9b exhibited a relatively high plasma protein binding rate, measurable antibacterial activity was still observed in plasma-containing systems (Fig. 9E). This phenomenon may arise from several factors. First, plasma protein binding is a reversible equilibrium, allowing bound drug to continuously release free compound as the free fraction is consumed. Second, 9b appeared to have a strong affinity for bacterial membrane components such as PG, which may enable the compound to partition preferentially toward bacterial cells rather than remain protein-bound. Finally, even a small free fraction may remain pharmacologically active, particularly for membrane-active antibacterial agents. These findings demonstrated that 9b exhibited favorable safety in eukaryotic cells and can selectively compromise bacterial cell membranes and interact with bacterial DNA at therapeutic concentrations, thereby displaying its antibacterial activity.

Subcutaneous administration of compound 9b at various doses (40, 20, 10, 5, and 2.5 mg/kg) was performed on healthy mice to monitor general condition and survival over a 24-h period for preliminary in vivo toxicity assessment. Animals in all treatment groups remained in good health, and no visible abnormalities were detected relative to the blank control group. The survival rate was 100% for all tested doses (Table S2). Mice treated with 40 or 20 mg/kg exhibited mild redness and swelling at the injection site, whereas those receiving 10, 5, or 2.5 mg/kg showed no obvious skin irritation, and their appearance was comparable to that of the untreated controls. These observations suggest that compound 9b demonstrates acceptable in vivo tolerability at doses ≤10 mg/kg. Next, the blood and skin at the administration site of the mice receiving the maximum dose (10 mg/kg), as well as tissues of vital organs for blood biochemistry indexes and hematoxylin and eosin staining tests, respectively. Additionally, Fig. 9F showed no significant abnormalities in the tissue sections (skin, heart, liver, spleen, lungs, and kidneys) of mice in 9b group compared with those of the blank group. Figure 9G displays that the blood biochemistry and blood routine indexes of the mice treated with 9b (10 mg/kg) were within the normal range and almost the same as that of the blank control group.

Additionally, we investigated the survival status of mice following intramuscular injections of different concentrations of 9b (20, 15, 10, and 5 mg/kg) into the thigh muscles. The results showed that the mice in all groups survived (Table S3), but the mice in 9b (20 mg/kg) had swollen and hardened thigh muscles. The thigh muscles of the remaining groups of mice were elastic and did not differ from the control group, indicating that the injection dose without adverse effects should be ≤15 mg/kg. Mice with leg muscles injected with 15 mg/kg of 9b solution showed no pathological alterations in the major organs of mice (heart, liver, spleen, lungs, and kidneys), no inflammatory cells were observed in the thigh muscles, and the muscle fibers did not show any breakage (Fig. 9H). There was no significant difference in blood routine and blood biochemistry indexes compared with the control group (Fig. 9I).

Collectively, these results confirmed that 9b achieved a favorable safety profile both in vitro and in vivo. This was largely attributed to its mechanism-driven dual selectivity—preferential binding to bacterial membranes followed by DNA interaction—while sparing mammalian membranes and nuclear DNA, highlighting its promise as a safe and effective lead compound for MRSA treatment.

Considering its pronounced in vitro anti-MRSA activity and favorable safety profile, the in vivo antibacterial performance of compound 9b was further investigated using a murine MRSA skin abscess model established according to previously reported protocols (Fig. 10A). A 6× 10⁸ CFU/ml dose of MRSA-16 was injected subcutaneously into the dorsal skin of mice, followed by the administration of 9b (2.5 and 5 mg/kg) and vancomycin (2.5 mg/kg) for treatment 2 h postinfection. As displayed in Fig. 10B, extensive abscess formation was evident in the dorsal skin and subcutaneous layers of mice in the control group. It is noteworthy that no abscesses were observed in mice treated with a high dose of 9b (5 mg/kg), and the dorsal skin and subcutaneous tissues were indistinguishable from those of the blank group. By comparison, animals treated with vancomycin (2.5 mg/kg) or low-dose 9b (2.5 mg/kg) still exhibited visible abscesses and localized erythema in the subcutaneous tissue. Additionally, histopathological analysis of the infected skin revealed extensive infiltration of inflammatory cells in the subcutaneous tissue, with further spread into the dermis and epidermis, indicative of a septic inflammatory response (Fig. 10C). In contrast, skin tissues from mice treated with vancomycin or low-dose 9b (2.5 mg/kg) exhibited only mild infiltration of inflammatory cells, indicating moderate improvement of infection (Fig. 10C). Notably, high-dose 9b (5 mg/kg) treatment resulted in skin morphology nearly indistinguishable from that of the blank group (Fig. 10C).

Moreover, to further investigate the in vivo efficacy of 9b, bacterial load and survival rates in the infected skin tissues of mice were determined by plate counting. We quantified the bacterial load in infected skin tissues of mice. As shown in Fig. 10D, only a few bacteria were observed in the 9b-treated groups (2.5 and 5 mg/kg) compared with the model group, and at same dosage, 9b exhibited superior antibacterial activity in vivo compare to vancomycin. Quantitative analysis further demonstrated that the bacterial load in the infected skin of the control group was approximately 8.9 log₁₀ CFU/g (Fig. 10E). Treatment with 9b (2.5 mg/kg) and vancomycin (2.5 mg/kg) reduced bacterial load by 2.4 and 1.0 log₁₀ CFU/g (Fig. 10E), respectively, corresponding to decreases in bacterial survival rates of 92.6% and 84.9% (Fig. 10F). Notably, administration of 9b at 5 mg/kg produced a more pronounced effect, with the bacterial load reduced by 4.0 log₁₀ CFU/g and bacterial survival nearly abolished (Fig. 10E).

Furthermore, systemic pathological changes were evident in the model group, with the liver, lung, and kidney exhibiting signs of congestion, interstitial edema, alveolar wall thickening, inflammatory infiltration, and renal tubular epithelial cell swelling (Fig. S8A). Conversely, the mice in the other treatment groups had no aberrant alterations in major organs. The mice in vancomycin (2.5 mg/kg) group exhibited good improvement in pathological symptoms compared with the mice in the control group, especially the mice in the 9b (2.5 and 5 mg/kg) groups basically tended to be normalized with no obvious pathological abnormalities (Fig. S8A). Collectively, 9b with outstanding in vivo therapeutic efficacy has the potential to be a candidate for combating MRSA infections.

To further investigate the in vivo efficacy of 9b, a mouse thigh infection model was developed using a clinical MRSA-16 isolate (Fig. 10G). From Fig. 10H, the bacterial load in the thigh tissues of mice in the control group was approximately 7.8 log₁₀ CFU/g. In contrast, infected mice treated with low-dose 9b (5 mg/kg) and vancomycin (10 mg/kg) both achieved a reduction in bacterial load of about 1.5 log₁₀ CFU/g, resulting in bacterial survival rates of approximately 94.45% and 95.14%, respectively, indicating significant antibacterial efficacy. Notably, infected mice treated with 9b (10 mg/kg) exhibited a reduction in thigh bacterial load of approximately 2.6 log₁₀ CFU/g and a reduction in bacterial survival of about 99.36% (Fig. 10H and I), suggesting that 9b has strong in vivo anti-MRSA efficacy.

Finally, histopathological evaluation was conducted to assess the systemic toxicity and tissue protective effects of 9b in a thigh model of mice infected with a clinical MRSA-16 isolate. As shown in Fig. 10J, the control group (MRSA infection without treatment) exhibited extensive muscle fiber fragmentation, dissolution, and dense inflammatory cell infiltration, indicative of severe tissue damage. In contrast, vancomycin (10 mg/kg) treatment partially alleviated these pathological features, with reduced inflammation and partially preserved muscle structure. Notably, 9b at both 5 and 10 mg/kg effectively preserved the integrity of skeletal muscle fibers and obviously reduced inflammatory cell infiltration, suggesting robust antibacterial efficacy and tissue protection. For major organs (Fig. S8B), MRSA infection induced systemic injury, particularly in the spleen, lungs, and kidneys. The control group showed signs of splenic extramedullary hematopoiesis, pulmonary inflammation, and renal tubular degeneration, reflecting infection-associated systemic stress. Treatment with vancomycin improved these pathological changes to a certain extent. Importantly, mice treated with 9b displayed well-preserved tissue architecture across all examined organs, with no signs of necrosis or massive inflammatory infiltration, indicating no evident systemic toxicity at either dose. These findings support the therapeutic potential and safety profile of 9b in treating MRSA infections.

After confirming the favorable in vivo safety and antibacterial efficacy of 9b, we further evaluated its potential for real-time visualization of infection. Using a mouse skin infection model, imaging was performed at the characteristic excitation wavelength of 9b (260 nm). As shown in Fig. 11A, no detectable fluorescence signal was observed in normal mice treated with either saline or 9b, which is consistent with the low quantum yield of free 9b and indicates minimal background signal in the absence of bacterial binding. In contrast, in MRSA-11-infected mice, a distinct fluorescence signal was observed at the infection site as early as 1 h after 9b treatment, whereas no comparable signal was detected in saline-treated controls. This finding was consistent with our in vitro results (Fig. 5D), demonstrating that fluorescence intensity increases obviously upon bacterial binding. Moreover, the temporal changes in fluorescence intensity correlated with the reduction in bacterial burden as determined by CFU counting (Fig. 11B), suggesting that the fluorescence signal reflects dynamic changes in infection load. Collectively, these results indicated that 9b enables infection-site visualization with minimal background interference and exhibits potential as a theranostic agent for antibacterial monitoring.

Analytical grades of phenanthrene-9,10-dione, 4-benzenedicarboxaldehyde, methyl iodide, and various amines were acquired from Sigma-Aldrich, Energy-Chemical, and Aladdin Inc. Staphylococcus aureus ATCC 29213, Micrococcus luteu, Streptococcus suis, Escherichia coli ATCC 29213, Klebsiella pneumoniae 18-227, and Klebsiella pneumoniae 18-29 were sourced from the ATCC (American Type Culture Collection). Clinical MRSA isolates were obtained from the first affiliated hospital of Zhengzhou University/University of South China. A Bruker Avance instrument 400 MHz instrument was employed to measure ¹H/¹³C NMR of phenanthro[9,10-d]imidazole compounds. An Agilent 6545 mass spectrometer was used to determine HRMS of phenanthro[9,10-d]imidazole compounds. The purity of all the title phenanthro[9,10-d]imidazole compounds was larger than 95% by using high-performance liquid chromatography.

A solution of phenanthrene-9,10-dione (0.96 mmol) and 4-benzenedicarboxaldehyde (0.96 mmol) was dissolved in glacial acetic acid, followed by the addition of ammonium acetate (19.2 mmol). The reaction mixture was refluxed and agitated at 100 °C for 1 to 2 h. The reaction solution was then cooled to room temperature and poured into cold water, and the solids were recovered by filtration and washed with ice water. Subsequently, the solids were dried in a vacuum to obtain intermediate 3. Next, to an acetone solution of 3 (0.27 mmol) and 1-bromo n-anemonane or 3,3-dimethylallylbromide (1.91 mmol), K₂CO₃ (1.91 mmol) was added, and then the reaction was agitated at reflux. Upon completion of the reaction, the solution was concentrated, and the residue was purified by silica gel column chromatography (SG-CC) to provide compounds 6a and 6b.

To the ethanol solution of compounds 3 (0.44 mmol) or 6a/6b (0.44 mmol) was added stannous chloride dihydrate (4.40 mmol), then reacted under room temperature. Once the reaction finished, the solvent was concentrated, and EtOAc was added to dissolve the residue followed by adding saturated NaHCO₃ to adjusted the pH to neutral. Subsequently, the reaction mixture was filtrated and then extracted with EtOAc, followed by drying with anhydrous Na₂SO₄, concentration, and isolation to give the compounds 4, 7a, and 7b.

Finally, compound 4 (0.11 mmol) or 7a/7b (0.11 mmol) was dissolved with N-methyl-2-pyrrolidone or acetonitrile, followed by the addition of iodomethane (1.10 mmol). After stirring the reaction at 75 °C for 48 h, compounds 5a, 8a, 8b, 9a, and 9b were obtained by using SG-CC.

To a solution of phenanthrene-9,10-dione (0.96 mmol) and ethyl 4-formylbenzoate (0.96 mmol) in glacial acetic acid, ammonium acetate (19.2 mmol) was added with stirring at 100 °C for 1 to 2 h. Once the reaction finished, heating was stopped and the reaction was cooled down to room temperature and then poured into iced water, collecting the solids by filtration and washing with iced water. The solid was dried under reduced pressure to give intermediate 11 in 63% yield. Next, compound 11 (0.27 mmol) and 1-bromo n-anemonane or 3,3-dimethylallylbromide (1.91 mmol) were dissolved in acetone, after which K₂CO₃ (1.91 mmol) was added, and the mixture was refluxed at 50 °C. Upon completion of the reaction, the mixture was extracted with EtOAc, and the combined organic layers were dried over anhydrous Na₂SO₄. The solvent was then concentrated, and the mixture was purified to yield compounds 13a and 13b. Finally, compounds 11, 14a, and 14b were dissolved by ethylenediamine with stirring at 110 °C and monitored by thin layer chromatography until completion. Then separation was performed by SG-CC to obtain the title compounds 12a, 14a, and 14b. After completion of the reaction, extraction was performed using EtOAc, followed by the amalgamation of the organic phases, drying over anhydrous Na₂SO₄, concentration, and subsequent separation to yield compounds 13a and 13b.

To a dichloromethane solution of 4-formylbenzoic acid (0.72 mmol), N-Boc-ethylenediamine (0.87 mmol), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (1.81 mmol), 1-hydroxybenzotriazole (1.08 mmol), and triethylamine (1.81 mmol) were added, then stirring at room temperature under N₂ for 12 to 14 h. After that, the mixture was washed with saturated saline and extracted with dichloromethane, and then the organic phases were combined, dried with anhydrous Na₂SO₄, concentrated, and purified by SG-CC to give the intermediate 16 in a yield of 71%. Subsequently, compound 17 was synthesized in accordance with the approach used for the synthesis of compound 3. Next, we synthesized intermediate 19a and 19b using the same method as for compounds 6a and 6b. Finally, intermediates 17, 19a, and 19b (60 mg) were hydrolyzed with HCl/MeOH (4 mol/l) solution. The solvent was removed once reaction finished, and residues were washed by petroleum ether to obtain the title compounds 18a, 20a, and 20b. The structures of all the title compounds 5a, 8a, 8b, 9a, 9b, 12a, 14a, 14b, 18a, 20a, and 20b were determined using ¹H/¹³C NMR and HRMS. High-performance liquid chromatography analysis confirmed a purity of ≥95% for all title compounds. Spectral data of 5a, 8a, 8b, 9a, 9b, 12a, 14a, 14b, 18a, 20a, and 20b are presented in the Supplementary Materials.

The MICs of compounds 5a, 8a, 8b, 9a, 9b, 12a, 14a, 14b, 18a, 20a, and 20b were tested using microbroth dilution method in compliance with the Clinical and Laboratory Standards Institute guidelines. The detailed procedure follows the previously reported methods [33,58,59]. We repeated the determination of MICs 3 times.

The spectral properties of compounds 9b and 5a were investigated using a UV–visible spectrophotometer and a fluorescence spectrophotometer. Stock solutions of the compounds were prepared in PBS, 0.9% NaCl solution, and DMSO with a final concentration of 1 × 10⁻⁵ mol/l. The UV–visible absorption spectra were obtained by scanning the wavelength over an appropriate spectral range. The fluorescence emission spectra were subsequently recorded between 400 and 700 nm. To simulate physiological environmental changes, the pH, ionic strength, and ethanol content of the solution were adjusted, and the corresponding fluorescence spectra of compound 9b were recorded. The fluorescence quantum yield (Φ) of compound 9b was determined using quinine sulfate as a standard reference under excitation at 350 nm and calculated according to the relative quantum yield method [60].

Single MRSA-16 colonies were cultured in 2 ml Mueller–Hinton broth (MHB) at 37 °C, 200 rpm for 12 h. The culture was then diluted 1:1,000 and further incubated under the same conditions for 3 to 5 h. After incubation, cells were harvested by centrifugation (4,000 rpm, 4 °C, 4 min), washed 3 times with PBS, and resuspended in PBS containing 9b (32 μg/ml). The suspension was shaken at 37 °C, 180 rpm for 3 h, and samples were collected every 20 min in the dark. Fluorescence intensity of 9b was determined using a microplate reader with excitation at 260 nm and emission recorded from 300 to 700 nm. Bacterial suspensions without drug treatment served as blank controls, and each condition was tested in triplicate.

The bactericidal activity of compound 9b against S. aureus and MRSA-16 was evaluated using a plate-based colony counting assay [36]. Briefly, an individual bacterial colony was transferred into 1.0 ml of MHB and incubated at 37 °C with shaking at 200 rpm. The resulting culture was subsequently diluted with fresh MHB to achieve a final cell density of 1 × 10⁵ CFU/ml and allowed to grow for an additional 2.5 h. Compound 9b was then introduced at concentrations corresponding to 4× and 8× the MIC, followed by further incubation at 37 °C and 200 rpm for 8 h. Aliquots were collected at 0, 0.5, 1, 2, 4, 6, and 8 h, and viable bacteria were quantified by colony enumeration. Time-kill curves were constructed by plotting log CFU as a function of incubation time. All experiments were conducted in triplicate.

Single colonies of S. aureus ATCC 29213 were cultured in MHB medium supplemented with a subinhibitory concentration (1/2 MIC) of 9b or norfloxacin for 12 h. The resulting cultures were then transferred onto Mueller–Hinton agar plates containing the same subinhibitory concentration and incubated for 24 h. This passage procedure was repeated for 20 consecutive generations, and MICs were evaluated at each generation using the microbroth dilution method.

MRSA-16 in the logarithmic growth phase were coincubated with compound 9b (final concentration 8× MIC) for 1 h. The mixture was centrifuged (5,000 rpm, 4 °C) and washed with PBS buffer 2 to 3 times, the supernatant was discarded, then 2.5% glutaraldehyde (1.0 ml) was added to fix it, and the mixture was stored at 4 °C overnight. Dehydration was carried out with a gradient concentration of ethanol, and then the bacteria were subjected to critical point drying and gold spraying. Bacterial morphological changes were observed using a high-resolution field emission scanning electron microscope. The same volume of sterile water was added as a blank control group.

A monoclonal colony of MRSA was transferred under sterile conditions into sterile MHB medium and cultured for 4 h. After centrifugation, the bacterial pellet was washed 2 to 3 times with PBS and resuspended in an equal volume of PBS containing the fluorescent dye dichlorodihydrofluorescein diacetate (5 μM). The suspension was then incubated in the dark at 37 °C for 30 min. Subsequently, 160 μl of the bacterial suspension was added into a 96-well plate, followed by 40 μl of prediluted compound at different concentrations. After incubation at 37 °C for 45 min, fluorescence intensity was measured at excitation/emission wavelengths of 488/530 nm.

After incubating MRSA in MHB medium (2.0 ml) at 37 °C for 12 h, 9b (8× MIC) was added and cultivated under light-avoidance conditions for 4 h. After centrifuging (5,000 rpm, 4 °C, 5 min) and washing with PBS 3 times, the bacteria were resuspended in PBS (180 μl) buffer, PI (10 μl, 20 μg/ml) dye was added, and the mixture was incubated for another 20 min in an ice bath. No compound was introduced as a blank control for laser scanning confocal microscope images.

The checkerboard dilution method [61,62] was used to measure the interactions of 9b with bacterial cell wall PGN and cell membrane phospholipids. After the mixing of compound 9b (50 μl, 0 to 128 μg/ml) with different concentrations of PG, PE, CL, and PNG (50 μl, 0 to 128 μg/ml) in equal amounts, it was followed by the addition of MRSA-16 bacterial suspension (100 μl, 1 × 10⁵ CFU/ml). The mixture was placed in a 96-well plate and incubated at 37 °C for 16 to 18 h before reading the results.

A monoclonal colony of MRSA-16 was cultured in 2 ml of MHB at 200 rpm for 5 to 7 h, harvested by centrifugation (5,000 rpm, 4 °C, 5 min), washed 3 times with PBS, and resuspended in an equal volume of dichlorodihydrofluorescein diacetate solution. After incubation at 37 °C for 40 min, the cells were centrifuged again and resuspended in PBS to remove excess dye. Then, 180 μl of bacterial suspension was mixed with 20 μl of 9b at different concentrations and incubated at 37 °C for 1 h. Fluorescence intensity was recorded on a microplate reader (λex/λem = 488/530 nm). Each assay was performed in triplicate.

Compound 9b (50 μl, 0.5 μg/ml), genomic DNA (50 μl, 0 to 0.2 μg/ml), and vancomycin (50 μl, 1.0 μg/ml) were added into 96-well plates, and 100 μl of MRSA-16 suspension (6 × 10⁵ CFU/ml) was added. The final volume of each well was 200 ml, then incubated at 37 °C for another 14 to 18 h. OD₆₀₀ (optical density at 600 nm) value of each well was determined to evaluate the growth of bacteria.

To Luria–Bertani medium containing 100 μg/ml ampicillin, E. coli DH5α containing pUC19 was added, and then the mixture was cultured (37 °C, 200 rpm) for 12 h. Plasmid extraction was carried out according to the plasmid extraction kit, and 6 μl of different concentrations of 9b (100, 75, 50, 25, 10, 5, 2.5, 1 μg/ml) were mixed with plasmid (1 μl, 150 ng/μl) and incubated at 37 °C away from light for 6 h. The plasmid (4.5 μl) was mixed with 10 × loading buffer (0.5 μl) using a 0.8% gel run at 150 V for 15 min. The gel bands were visualized and photographed for observation by a UV gel acquisition system.

The compound 9b (4.5 μl, 50 μg/ml) was mixed with 1 μl of plasmid (150 ng/μl), and then 4.5 μl of plasmid was mixed with 10 × loading buffer (0.5 μl) in each group at each time point of 0.5, 1, 1.5, 2, 3, 4, 5, and 6 h. The plasmid was run for 15 min at 150 V using a 0.8% gel. The gel bands were photographed and analyzed by a UV gel acquisition system.

Different concentrations of 9b (6 μl, 100, 75, 50 μg/ml) were mixed with 1 μl of plasmid (150 ng/μl), respectively, and then the mixture was incubated at 37 °C for 6 h without light. The product was purified by the product purification kit and then run for 15 min using a 0.8% gel at 150 V. The gel bands were observed and photographed for analysis by a UV gel acquisition system.

Cultured MRSA-16 was centrifuged (5,000 rpm, 4 °C, 5 min) and washed 3 times with PBS buffer. The MRSA-16 was resuspended in MHB containing different concentrations of 9b (1 to 8 μg/ml) at 37 °C for 3 h. Genomic DNA was extracted using a genome kit and then analyzed by 0.8% gel running at 150 V for 15 min and visualized by a UV gel acquisition system.

The monoclonal colony of MRSA-16 was picked up in MHB medium (2 ml) and incubated (37 °C, 200 rpm) for 12 h. On the next day, the MRSA bacterial solution was diluted 100-fold and cultured for another 3 to 6 h, and then 9b was added to achieve a final concentration of 8 μg/ml, and the mixture was further incubated for 8 h. After centrifugation, the supernatant was collected and frozen in liquid nitrogen for 10 to 15 min and then transferred to −80 °C for refrigeration. Finally, the samples were measured by Shanghai Vina Biotechnology Co.

The detailed procedure for the determination of hemolytic activity (HC₅₀) was according to our previously reported methods [33,62,63], and it can be found in the Supplementary Materials.

To determine the stability of compound 9b in plasma, we incubated different concentrations of 9b in 50% plasma for 3 and 6 h. After incubation, we added bacterial suspension (150 μl) in 96-well plates and incubated at 37 °C for 16 to 24 h. After reading the MIC, the MBC was determined. The plasma stability of compound 9b’s antibacterial effect was assessed through MBC variation over time.

The cytotoxicity and membrane selectivity of compound 9b on LO2 and NIH3T3 cells were determined by the cell counting kit-8 method according to our previous reports [59,62].

All animal procedures complied with institutional ethical regulations (SYXK (X) 2022-0008) approved by the Animal Protection and Utilization Committees of Zhengzhou University and the University of South China.

An in vivo mouse skin abscess infection model was used to test 9b’s anti-MRSA effectiveness.

An established murine skin abscess model was employed to evaluate the anti-MRSA efficacy of compound 9b. A total of 30 healthy female KM mice were randomly allocated into 4 groups: the blank group (0.9% NaCl), the infected control group (MRSA-16 only), the 9b-treated groups (2.5 and 5 mg/kg), and the positive control group receiving vancomycin (5 mg/kg). After depilation of the dorsal area, mice were anesthetized and subcutaneously injected with MRSA-16 (60 μl, 6 × 10⁸ CFU/ml). Two hours later, each group received its designated treatment at the infected site. For the thigh infection model [64,65], another cohort of 30 female KM mice of similar body weight was randomly assigned into 5 treatment groups: the blank control, the infected control, the vancomycin group (10 mg/kg), the high-dose 9b group (10 mg/kg), and the low-dose 9b group (5 mg/kg). After removing hair from the left thigh and keeping the animals for 24 h, all mice except the blank controls were intramuscularly inoculated with MRSA-16 suspension (60 μl, 1 × 10⁸ CFU/ml). Two hours following infection, compound 9b at different doses was injected locally at the infection site for two consecutive days with a 24-h interval. Equal volumes of saline were administered to the blank and infected control groups. Twenty-hour hours after the last injection, the animals were euthanized. The infected skin or thigh tissues, together with major organs (heart, liver, spleen, lungs, and kidneys), were collected, fixed in 4% paraformaldehyde, and embedded in paraffin for hematoxylin and eosin staining. Bacterial loads at the infection sites of each group were subsequently quantified using the drop-plate counting method to assess the in vivo antibacterial activity of compound 9b.

Results are expressed as means ± SD or means ± standard error of the mean (SEM), as indicated. All statistical evaluations were carried out using SPSS software (version 21.0; IBM Corp.). Differences between 2 groups were analyzed by 1-way analysis of variance (ANOVA), depending on the experimental design. A P value ≤ 0.05 was considered statistically significant. Levels of significance are denoted as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Each experiment was independently performed at least 3 times.