Hinokiflavone Attenuates the Virulence of Methicillin-Resistant Staphylococcus aureus by Targeting Caseinolytic Protease P

ABSTRACT

Drug-resistant bacteria was the third leading cause of death worldwide in 2019, which sounds like a cautionary note for global public health. Therefore, developing novel strategies to combat Methicillin-resistant Staphylococcus aureus (MRSA) infections is the need of the hour. Caseinolytic protease P (ClpP) represents pivotal microbial degradation machinery in MRSA involved in bacterial homeostasis and pathogenicity, considered an ideal target for combating S. aureus infections. Herein, we identified a natural compound, hinokiflavone, that inhibited the activity of ClpP of MRSA strain USA300 with an IC₅₀ of 34.36 μg/mL. Further assays showed that hinokiflavone reduced the virulence of S. aureus by inhibiting multiple virulence factors expression. Results obtained from cellular thermal transfer assay (CETSA), thermal shift assay (TSA), local surface plasmon resonance (LSPR) and molecular docking (MD) assay enunciated that hinokiflavone directly bonded to ClpP with confirmed docking sites, including SER-22, LYS-26 and ARG-28. In vivo, the evaluation of anti-infective activity showed that hinokiflavone in combination with vancomycin effectively protected mice from MRSA-induced fatal pneumonia, which was more potent than vancomycin alone. As mentioned above, hinokiflavone, as an inhibitor of ClpP, could be further developed into a promising adjuvant against S. aureus infections.

INTRODUCTION

Staphylococcus aureus (S. aureus) infection is considered one of the most widespread diseases, and the emergence and widespread prevalence of multidrug-resistant bacteria represented by methicillin-resistant S. aureus (MRSA) have caused clinical treatment more difficult. The pathogenicity of S. aureus could not be underestimated, as there has been a dramatic increase in MRSA presence in hospitals and more recently, an epidemic of community-associated MRSA infections. The main reason for this formidable issue is the abuse of traditional antibiotics. In response to this tricky situation, an approach has been developed that effectively handles the infection and avoids the development of bacterial resistance, known as an “ anti-virulence strategy ” (1). This strategy inhibits the critical virulence factors of the bacteria and leverages the immune system to clear the germs. The pathogenicity of S. aureus is mainly due to its production of various pathogenic virulence factors at different stages of growth, including alpha-hemolysin (Hla), MgrA, Caseinolytic Protease P (ClpP), sortase A (SrtA) and coagulase (2). Bacterial virulence expression is a complex, interconnected, and collaborative process regulated by known and unknown regulators (3). S. aureus infection is rarely achieved by inhibiting a single virulence factor, and inhibiting some virulence factors may result in upregulation of others (4). Consequently, the development of anti-infective agents impacting multiple virulence factors has become a particularly promising strategy for preventing and treating S. aureus infections.

ClpP is a protein hydrolase that maintains physiological functions by degrading some erroneous and misfolded proteins (5). ClpP has received particular attention as it plays a vital role in regulating multiple virulence proteins in S. aureus, including Hla, Panton-valentine leukocidin (PVL), surface protein A (SpA) and phenol-soluble module α (3). Studies have indicated that ClpP is initiating in the initiation of adhesion to hosts and biofilm formation (6, 7). Intriguingly, activation or inhibition of ClpP both counteract the infection by S. aureus. Activation of ClpP directly causes uncontrollable and massive peptide hydrolysis, leading to bacterial death. Inhibition of ClpP activity to weaken bacterial pathogenicity mitigates damage to the host and promotes immune clearance by the host (3). Compared to the violent effects of ClpP activators, the inhibition of ClpP is relatively mild, which effectively reduces the development of resistance. Our objective was to identify natural compounds with ClpP inhibitory activity that could be used to attenuate the pathogenicity of S. aureus. Since flavonoids currently account for the vast majority of the few discovered natural ClpP inhibitors (8–10), as follow-up research, we set out to identify novel ClpP inhibitors from the compound library of flavonoids.

Hinokiflavone (4′,5,5′',7,7′'-Pentahydroxy-4′'' 6-oxydiflavone) is a natural compound derived from various plant species, mainly, including leaves of Chamaecyparis obtuse (11–13). The biological consequences and applications of hinokiflavone have been studied, including anti-inflammatory, anti-tumor, antiviral and it has been approved to fight a diverse range of diseases, such as influenza, dengue fever, HIV-AIDS, hepatitis, leishmaniasis and malaria (14–16). In this study, we demonstrated for the first time that hinokiflavone was a direct inhibitor of ClpP of S. aureus, which reduced the virulence of MRSA strain USA300. Mice suffering from MRSA-associated lethal pneumonia had higher survival rates when administered with hinokiflavone or in combination with vancomycin.

RESULTS

Hinokiflavone inhibits ClpP activity in vitro.

Natural compounds are an essential source of novel anti-infective agents (17). Since ClpP is capable of recognizing and hydrolyzing the fluorescent substrate peptide Suc-LY-AMC, leading to changes in fluorescence intensity, the FRET technique was employed to screen for ClpP inhibitors. One compound was identified in the primary screening, with a hit rate of 2.38% (Fig. 1a). The details and relative inhibitory activities of 42 flavonoids were listed in Table S1.

FIG 1.

Hinokiflavone was identified as an inhibitor of S. aureus ClpP. (a) Diagram of FRET assay for screening ClpP inhibitors. When inhibition activity was greater than 60%, the compound was considered a potential ClpP inhibitor. (b) Chemical structure of hinokiflavone. (c) Hinokiflavone inhibited the cleavage of the fluorescent substrate Suc-LY-AMC by ClpP and the IC₅₀ of hinokiflavone was 34.36 μg/mL. (d) The growth of USA300 was not affected by 64 μg/mL of hinokiflavone. Wild-type USA300 was supplemented with 0.5% DMSO as a control group. (e) Effect of different concentrations of hinokiflavone on the viability of HEK-293T cells.

ClpP protein characteristics were consulted on the website (https://beta.uniprot.org/uniprotkb/P63786/entry) with a molecular mass which is about 21,514 Da. The molecular mass of the purified protein by SDS-page assay was confirmed to be consistent (Fig. S2). The critical aspect of antiviral strategy is to inhibit the expression of virulence factors without affecting the growth of pathogenic bacteria. To this end, we screened the anti-virulence agents that did not threaten the viability of bacterium, and further determined the effect of hinokiflavone on the growth of USA300. The inhibitory effect of hinokiflavone (Fig. 1b) on ClpP activity was measured with an IC₅₀ value of 34.36 μg/mL (63.81 μM) (Fig. 1c). As illustrated in Fig. 1d, hinokiflavone did not suppress USA300 growth at a concentration below 64 μg/mL. The MIC of hinokiflavone against S. aureus USA300 was greater than 512 μg/mL, demonstrating that hinokiflavone has no significant antibacterial ability. Cytotoxicity was further tested and showed that hinokiflavone at 256 μg/mL did not affect HEK293T cells viability (Fig. 1e). These results elucidated that hinokiflavone was a promising ClpP inhibitor with noncytotoxicity.

Hinokiflavone reduces the expression of virulence factors in S. aureus.

ClpP plays a vital role in regulating the expression of virulence factors in S. aureus (3). We investigated the effect of hinokiflavone on the expression of several virulence factors through RT-qPCR. Results indicated that RNAIII and agrA, which are mainly responsible for the production of virulence factors were downregulated by 6.7-fold and 4.5-fold, respectively. Furthermore, the transcription levels of hla, lukes, psmα, and spa were also significantly reduced (Fig. 2a). However, the clpP gene showed no significant effect. PVL and alpha-toxin secreted by S. aureus are involved in the pathogenic process, leading to cell death (18–20). Western blot assay revealed that there is no difference in ClpP protein expression with or without the hinokiflavone (Fig. S4). However, hinokiflavone inhibited alpha-toxin and PVL protein expression in a dose-dependent manner in S. aureus USA300 (Fig. 2b and c) and Newman strain (Fig. 2d and e). Of these, hinokiflavone inhibited Hla and PVL protein expression by 17.32 ± 11.42% and 22.08 ± 7.17%, respectively, at 128 μg/mL, which was close to that in Newman-ΔclpP.

FIG 2.

Hinokiflavone depressed the expression of various virulence factors regulated by ClpP in S. aureus. (a) Transcript levels of the related genes agr, RNAIII, hla, luks, psm-α, spa, and clpP were determined by RT-qPCR exposed to the concentration of hinokiflavone at 64 μg/mL. The reference gene used in Quantitative Real-time PCR is 16sRNA. (b) Quantification of the effect of different concentrations of hinokiflavone on the expression of alpha-toxin and (c) PVL in S. aureus USA300 by Western blotting and statistical analysis of their corresponding gray-scale values. (d) Quantification of the effect of different concentrations of hinokiflavone on the expression of alpha-toxin and (e) PVL in S. aureus Newman by Western blotting and statistical analysis of their corresponding gray-scale values. (f) The effect of hinokiflavone on the urease production in USA300 and ΔclpP strains. (g) Effect of different concentrations of hinokiflavone on the hemolysis activity of S. aureus USA300. (h) Determination of the neutralizing activity of hinokiflavone on Hla in supernatant of S. aureus USA300. Data were expressed as mean ± SD for three independently experiments. *, P < 0.05, **, P < 0.01 and ***, P < 0.001 compared to the control group.

The results of hemolysis assays also demonstrated that the hemolytic activity of hinokiflavon-treated S. aureus USA300 supernatant was significantly reduced, which was consistent with the Western blot results of Hla expression (Fig. 2g). Furthermore, we observed that hinokiflavone was unable to neutralize alpha-toxin (Fig. 2h), suggesting that the ability of hinokiflavone inhibiting hemolysis is achieved by affecting alpha-toxin expression rather than directly inhibiting activity.

Urease is a critical factor in the pathogenesis of S. aureus and also an essential component of the system of acidic strain response of S. aureus (21, 22). It has been documented that deletion of ClpP or inhibition of ClpP activity in S. aureus notably induced urease production (23). Urease catalyzed the hydrolysis of urea to form ammonia, which turned the red phenol indicator in the medium red. We further detected the inhibitory effect of hinokiflavone on ClpP by a urease activity assay, which showed that deletion of ClpP in S.aureus as well as hinokiflavone-treated S.aureus USA300 markedly increased the production of urease, and the absorption spectra clearly showed this result (Fig. 2f), providing further evidence of its inhibitory effect on ClpP activity. Based on the above analysis, hinokiflavone regulated the expression of multiple virulence factors by inhibiting ClpP activity.

Hinokiflavone directly interacts with ClpP.

A prerequisite for determining the interaction between a natural compound and a protein is to detect the presence of binding between the components. Thermal shift assay (TSA) has become an essential technique for biophysical ligand screening and protein engineering as it measures the effect of ligand molecules on the thermal stability (24). To confirm whether there is a direct interaction between hinokiflavone and ClpP, the effect of hinokiflavone on the thermal stability of ClpP was also analyzed by the TSA assay. The 64 μg/mL of hinokiflavone shifted the T_m value to the left by 2°C (Fig. 3a). Subsequently, CETSA data further confirmed the ability of hinokiflavone to bind ClpP protein intracellularly (Fig. 3b). These results indicated that hinokiflavone affected the thermal stability of ClpP protein, pointing to a direct binding. Molecular docking is an instrumental vehicle for studying the interaction between biomolecules and inhibitors by simulating the amino acid sites at the molecular level (25, 26). In detail, the results showed that residues SER-46, LYS-26 and ARG-28 had strong hydrogen bonding interactions with hinokiflavone (Fig. 3c). Following the molecular docking results, we constructed three mutants of ClpP, including K26A-ClpP, R28A-ClpP and S46A-ClpP, and then purified the corresponding mutant proteins. The FRET method was applied to detect that these mutants were resistant to the inhibitory effect of hinokiflavone to a certain extent (Fig. 3d), further providing evidence of these active binding sites. Local Surface Plasmon Resonance (LSPR) effects have been widely used to detect molecular dynamics processes between drug molecules and target proteins (27). The dissociation constant (Kd), the binding constant (Ka) and the affinity constant (K_D) are the leading biochemical indicators that quantitatively describe the binding of proteins to small molecules (28). To further validate the binding mode of hinokiflavone with ClpP, the K_D value calculated by LSPR analysis was 1.35 × 10⁻_{6 M} (Fig. 3e). The association rate constant k_a and the dissociation rate constant k_d value of 6.47 × 10³ M⁻¹s⁻¹ and 8.72 × 10⁻³ s⁻¹, respectively, indicate a strong interaction between hinokiflavone and ClpP.

FIG 3.

Hinokiflavone binded to ClpP and inhibited its activity. (a) The binding of different concentrations of hinokiflavone to ClpP was detected by TSA technique. Hinokiflavone inhibited the thermal stability of ClpP in a concentration-dependent manner. (b) Grayscale images by SDS-PAGE and the quantification of CETSA of ClpP with and without hinokiflavone (128 μg/mL) incubation. (c) Molecular docking simulates the key amino acid sites for the binding of hinokiflavone and ClpP. (d) Three potential binding sites mutants of K26Α-ClpP, R28Α-ClpP and S46Α-ClpP showed increased resistance to ClpP inhibition by hinokiflavone. (e) LSPR assay confirmed that a direct interaction between hinokiflavone and ClpP. (f) The main mechanism of hinokiflavone inhibiting S. aureus ClpP and the biological consequence. Significance of data are expressed as mean ± SD. *, P < 0.05, **, P < 0.01 and ***, P < 0.001.

Protective effect of the combination of Hinokiflavone and vancomycin on pneumonia infection in mice.

To initially evaluate the clinical utility value of hinokiflavone, combination assays of hinokiflavone with traditional antibiotics were performed in vitro. The results of the checkerboard method enunciated that the FICI indices were all greater than 0.5, indicating that hinokiflavone did not have synergistic effects with these antibiotics in vitro (Table S3). We then evaluated the synergistic efficacy of vancomycin, the preferred drug for the clinical treatment of MRSA infection, in combination with hinokiflavone in a pneumonia-infected mice model. Mice in each group were treated with hinokiflavone (100 mg/kg·d⁻¹) or hinokiflavone (100 mg/kg·d⁻¹) + vancomycin (100 mg/kg·d⁻¹) after intranasal inoculation with S. aureus USA300 or USA300-ΔclpP, and the survival rate of mice were recorded for 96 h (Fig. 4a left panel). As shown above (Fig. 4b), the 96-h survival rates of mice in USA300-infected group was 20%, while that in the USA300-ΔclpP-infected group reached 90%, demonstrating the crucial role of ClpP in the pathogenesis of pneumonia infection in mice (3, 8). In addition, the survival of mice infected with USA300 ΔclpP strains had no significant difference with or without hinokiflavone treatment, suggesting that hinokiflavone acted directly on the ClpP of S. aureus that causes pneumonia. The results showed that hinokiflavone in combination with vancomycin increased survival from 60% to 70% compared to vancomycin alone (P = 0.57).

FIG 4.

Synergistic protective effect of hinokiflavone and vancomycin against pneumonia induced by S. aureus in mice. (a) Experimental model of pneumonia induced by MRSA in C57BL/6J mice. (b) Effect of hinokiflavone treatment (100 mg/kg·d⁻¹), vancomycin (100 mg/kg·d⁻¹) or hinokiflavone combination with vancomycin treatment on the survival of C57BL/6J mice (n = 10) exposed to lethal doses of S. aureus USA300. (c) After 48 h of treatment, lung tissue from each group of mice (n = 5) was homogenized and bacterial load quantified. (d) Gross and histopathological examination of lung tissue from S. aureus WT and WT-ΔclpP infected mice treated with hinokiflavone (100 mg/kg·d⁻¹), vancomycin (100 mg/kg·d⁻¹) or a combination of hinokiflavone and vancomycin by subcutaneous injection. Scale bar, 100 μm. (e-g) Levels of IFN-γ, IL-6 and TNF-α, inflammatory cytokines, in lung perfusion fluid of mice in each group (n = 3). Data are means ± SD of results for compared with USA300 + PBS, *, P < 0.05, **, P < 0.01 and ***, P < 0.001 (by one-way ANOVA); compared with USA300 + Vancomycin, #, P < 0.05 (by unpaired t test). Each experiment was independently duplicated three times.

To further illustrate the protective function of hinokiflavone in mice, each mouse was euthanized after 48 h and lung tissue was collected (Fig. 4a right panel). Gross pathological evaluation of the contracted lung tissue showed that the lungs of the uninfected mice were pale pink and spongy. In contrast, the lungs of the USA300-infected group showed diffuse solid lung tissue that was dark red, firm in texture, with apparent congestion and significantly damaged alveoli. Compared to the infected group, the hinokiflavone-treated group showed an improved congested state of focal solid lung tissue and a significant reduction in inflammatory cell infiltration in the alveolar lumen and improving lung inflammation (Fig. 4d). The number of viable bacterium in the lungs of hinokiflavone-treated mice was consistent with these pathological findings with a significant reduction compared to the control group. The lung bacterial loads of mice were superiorly decreased in treatment group with hinokiflavone in combination with vancomycin (Fig. 4c). Furthermore, there was no significant difference in the lung load of mice infected with the USA300 ΔclpP strain after treatment with or without hinokiflavone, indicating that hinokiflavone may directly target the ClpP of S. aureus. Subsequently, the ELISA further confirmed that hinokiflavone treatment significantly reduced the secretion of major inflammatory factors IFN-γ, IL-6 and TNF-α in alveolar lavage fluid. The combination of hinokiflavone with vancomycin more significantly reduced the expression levels of these above inflammatory factors than vancomycin treatment alone (Fig. 4e to g).

In simple terms, the combination treatment of hinokiflavone and vancomycin revealed a more potent therapeutic effect on reducing the number of viable bacterium in the lungs of mice, alleviating the pathological changes in the lungs reducing the levels of inflammatory factors compared to the hinokiflavone alone. Hinokiflavone had an adjuvant effect on vancomycin treatment, thereby protecting mice against lethal lung infection caused by S. aureus.

DISCUSSION

The abuse of broad-spectrum antibiotics has led to the evolution of resistance in S. aureus (29). With the threat posed by antibiotic-resistant bacteria, “from antibiotics to anti-virulence” has developed into a prospective therapeutic strategy to combat S. aureus infection. On the one hand, as a highly conserved serine protease, ClpP laid the foundation for the research and development of innovative antimicrobial agents. On the other hand, as a virulence-regulating hub, ClpP makes significant contributions to the virulence, stress response and maintenance of internal environment stability of S. aureus (30, 31). Due to the virulence factors of S. aureus interact with each other, it is difficult to eliminate virulence by inhibiting the expression of any single virulence. Therefore, anti-virulence approaches in S. aureus either target a determinant regulator with established widespread impact on pathogenicity or by eliminating several virulence factors simultaneously. As a multiplex bacterial protein regulator, ClpP is capable of modulating multiple virulence factors simultaneously, regulating the expression of downstream virulence factors by initiating RNAIII, which has a broader virulence influence. In contrast, sortase A, a widely studied virulence target, helps to anchor surface proteins involved in adhesion, invasion and evasion of the S. aureus cell wall. In theory, in comparison to ClpP, the anti-infection effect of its inhibition is limited. The critical roles of ClpP make it a promising target that has been involved in the virulence controls of pathogenic bacteria during host infections.

Currently, natural products are an essential source for developing anti-ClpP agents. In the present work, we indentified hinokiflavone, one of the natural flavonoids, which effectively inhibited the ClpP activity of S. aureus at low concentrations without cytotoxicity. In addition to hinokiflavone, our previous studies had identified two ClpP inhibitors of MARS, myricetin (8) and nepetin (9). In contrast to myricetin (8), hinokiflavone does not show a pan-detection interference structure (PAINS) for the molecule catechol, suggesting that hinokiflavone is highly specialized without the possibility of false-positive compounds. Notably, their chemical structures are similar and both are flavonoids (Fig. S5). Although our screening data showed a high hit rate of flavonoids targeting ClpP, a vast majority of flavonoids do not possess this profile. As more evidence accumulates in the future, the structure-activity relationships of flavonoids to ClpP targeting will become clear.

Moreover, hinokiflavone is an anti-inflammatory substance, showing significant anti-inflammatory potential. It also plays a vital role in antitumor effects and antioxidant efficacy (14–16), making it a promising agent with multiple effects on severe infections in the future. As such, the protective effect of hinokiflavone against S. aureus pneumonia in mice does not exclude its anti-inflammatory effect in addition to its ClpP inhibition activity. Furthermore, it cannot be left out that hinokiflavone may also affect the formation of persistent cells in vivo, which may be involved in its anti-infective activity (32, 33).

In our findings, TSA and CETSA data showed that hinokiflavone bonded to ClpP and reduced the thermal stability of ClpP. Afterward, we determined the affinity constant (K_D) of hinokiflavone to ClpP by LSPR at 1.35 nM, indicating the binding between hinokiflavone and ClpP was robust and molecular docking revealed that SER-46, LYS-26 and ARG-28 were the key binding residues. The transcription levels of genes such as agrA, RNAIII, hla, pvl, psmα, and spa were significantly decreased by hinokiflavone treatment. Moreover, PVL and HLA have potent abilities to lyse host cells (e.g., neutrophils), resulting in tissue necrosis (34). ClpP mutant as well as hinokiflavone-treated significantly reduced the levels of PVL and HLA and decreased the pathogenicity of S. aureus. Meanwhile, we observed hinokiflavone consistent with S. aureus USA300-ΔclpP, both spiked urease expression. Therefore, we believe that hinokiflavon is a potent anti-S. aureus toxicity compound.

Interestingly, a considerable number of bacterium were observed in the lungs of mice infected with the ΔclpP strain, yet these mice had a higher survival rate and mitigated pathological lung injury. This suggested that knockout of the clpP gene had little effect on the viability of S. aureus USA300 strain (35). Inhibition of ClpP attenuates the pathogenicity of S. aureus primarily by reducing the expression of virulence factors rather than directly killing the bacteria, which contributes to the reduction of bacterial survival pressure thereby declining the development of drug resistance, which is corroborated in the study of Gao (3).

Based on the above findings, hinokiflavone simultaneously affected the expression of multiple virulence factors associated with pathogenicity in S. aureus, which was undoubtedly an efficient inhibitor of ClpP (Fig. 3f). Meanwhile, its excellent anti-virulence efficacy and safety indicated that it could be an anti-infection agent worthy of further development.

MATERIALS AND METHODS

Bacterial strains, plasmids, culture conditions, bacterial strains, and reagents.

S. aureus USA300, Newman, ClpP deficient S. aureus USA300 (ΔclpP), and plasmid pET28a-clpP used in this study were obtained from laboratory collections. S. aureus and Escherichia coli (E. coli) were cultured in Trypticase soy broth (TSB, Hopebio, Qingdao, China) or Luria Bertani (LB, Hopebio, Qingdao, China) broth, respectively. If necessary and appropriate, kanamycin (50 μg/mL) was added to E. coli media. Unless stated otherwise, all bacterial cultures were aerobically incubated at 37°C with shaking at 220 rpm. Hinokiflavone (Purity > 98%, Vichy, Chengdu, China) was dissolved in DMSO to form a 10 mg/mL solution and the HPLC is presented in Figure S1. Conventional antibiotics include ceftaroline fosamil, cefoxitin, vancomycin, cefotaxime, oxacillin sodium, penicillin G, doxycycline, ceftiofur sodium, ceftriaxone sodium, cefepime, and clavulanate potassium were purchased from Shanghai Biotech Company (Shanghai, China).

Expression and purification of ClpP protein.

The pET28a-clpP plasmid was transformed into E. coli BL21(DE3) by heat shock and incubated in LB medium containing kanamycin (50 μg/mL) at 37°C with 220 rpm shaking until OD₆₀₀ reached 0.8. The 0.5 mM IPTG was added and induced overnight at 16°C and 180 rpm. Most of the ClpP protein was expressed in bacteria, so bacterial precipitates were collected by high-speed centrifugation. The bacterial precipitate was fragmented by sonication and ClpP protein was released into the fragmentation buffer. After high-speed centrifugation to remove the undissolved fraction, the Ni-NTA system was used to purify the ClpP protein. Throughout the process, samples from each stage are retained for SDS-PAGE identification (Figure S2).

Effect of Hinokiflavone on the protease activity of ClpP.

A fluorescence resonance energy transfer (FRET)-based assay was used to screen for ClpP inhibitors based on the ability of ClpP to specifically cleave the fluorescent peptide substrate Suc-LY-AMC (catalog number S1153, Sigma-Aldrich, St. Louis, MO, USA). The difference in RFU was used to define the enzymatic activity of ClpP and calculated the relative inhibitory activity. The compound was considered a potential inhibitor when the inhibitory activity was greater than 60%. The 100 μL reaction system consisting of 1 μΜ ClpP protein, various concentrations of hinokiflavone (0.5 to 256 μg/mL) and ClpP buffer (100 mM HEPES, 100 mM NaCl, pH = 7.0) were added to 96-well plates and incubated for 30 min, followed by the addition of the fluorescent substrate Suc-LY-AMC. After continuing incubation for 20 min, the activity of ClpP was determined by measuring excitation at 360 nm and emission at 465 nm using a microplate reader (Spectramax M3; Molecular Devices). Then the IC₅₀ value of hinokiflavone was calculated. The reaction system containing an equal volume of DMSO was used as a negative control.

Mutant proteins were expressed by low-temperature induction. The Ni-NTA system was applied to purify the proteins. The FRET system was used to assess the activity of the mutant protein based on the cleaving activity of ClpP. The details were similar to the FRET method for screening ClpP inhibitors. The inhibitory activity of the same concentration of hinokiflavone (64 μg/mL) versus different mutant proteins. The results are expressed as relative inhibitory activity.

Determination of growth curves.

The strain is kept in −20°C environment. Before experimenting, a single colony of USA300 and USA300 ΔclpP was picked for recovery after overnight incubation. The experiments are then carried out to ensure the activity of strains. Then the cultures were added to 2 mL of TSB medium at a ratio of 1:100, respectively. Hinokiflavone (64 μg/mL) or DMSO was also added. One hundred microliters of culture were used to measure absorbance values (OD₆₀₀) at different time points.

Determination of MIC.

The MIC of hinokiflavone against S. aureus USA300 was determined by following the Clinical and Laboratory Standard Institute guidelines (CLSI). Briefly, 96-well plates containing hinokiflavone (0 to 512 μg/mL) and S. aureus USA300 (1 × 10⁵ CFU) in 100 μL fresh CAMHB medium were incubated at 37°C for 16 to 20 h. The MIC was the lowest concentration at which no bacterium were found on visual observation.

Cytotoxicity experiments.

HEK293T cells were seeded into 96-well plates at 100 μL per well at a density of 5 × 10⁴ cells/mL and cultured for 16 h. After adhering, various concentrations of hinokiflavone (0 to 256 μg/mL) were added to the 96-well plate. Cells were continued to grow in a 5% CO₂ incubator at 37°C for 24 h. Subsequently, MTT solution was mixed to the medium at 10 μL/well and incubated for 4 h at 37°C. The culture medium was discarded, followed by the addition of 100 μL of DMSO. The absorbance at 490 nm was calculated using a microplate reader.

Hemolysis assay.

S. aureus USA300 and USA300-ΔclpP were incubated overnight in TSB and grown until OD₆₀₀ = 0.3. Hinokiflavone (0 to 128 μg/mL) was added for further continuing until OD₆₀₀ reached 2.5. Subsequently, the supernatant of the S. aureus culture (100 μL) was collected by high-speed centrifugation (10,000 rpm, 10 min) and added to sterile PBS with 25 μL of defibrinated rabbit blood for a total system of 1 mL. After incubation at 37°C for 30 min, the mixing system was then centrifuged (7,500 rpm, 4°C, 3 min) and the hemolytic activity was determined by measuring the absorbance of the supernatant at an OD₅₄₃. In addition, DMSO or 1% Triton X-100 treatment instead of S. aureus supernatant were used, respectively.

To detect the neutralizing activity of hinokiflavone against Hla, S. aureus USA300 supernatant was incubated with different concentrations of hinokiflavone (8 to 128 μg/mL) at 37°C for 30 min. PBS and rabbit RBCs were added for another 30 min, and the absorbance values were also measured above.

Quantitative real-time PCR.

S. aureus USA300 was cultured to OD₆₀₀ = 0.3, added hinokiflavone to a final concentration of 32 μg/mL and the total solution was grown at 37°C to OD₆₀₀ = 2.5. The bacterium was collected by centrifugation (5000 rpm, 4°C, 5 min) and total RNA was extracted using TRIzol, according to the manufacturer's instructions. The cDNA was reverse transcribed from total RNA with Easy Script All-in-One First-Strand cDNA Synthesis Super Mix for qPCR (catalog number AE341, TransGen Biotech, Beijing, China) according to the manufacturer's instructions. Quantitative Real-time PCR analysis was performed using a Fast Start Universal SYBR green Master (catalog number 4913850001, Roche Molecular Biochemicals, Mannheim, BW, Germany) and a CFX96 Touch Real-Time PCR Detection System (Bio-Rad). The reference gene used in Quantitative Real-time PCR is 16sRNA. Each experiment was performed independently and repeated three times. Primers used for qPCR are listed in Table S2.

Western blot analysis.

S. aureus USA300 and Newman were incubated overnight with different concentrations of hinokiflavone (0 to 128 μg/mL), and the supernatant was collected by centrifugation (5000 rpm, 4°C, 5 min). After electrophoresis, the gel was transferred to a PVDF membrane (GE Healthcare, Amersham, UK) by the wet transfer method, which was blocked for 2 h at room temperature. The following antibodies were applied and incubated with the membrane: the rabbit anti-staphylococcal alpha-toxin antibody (1:10,000, catalog number S7531, Sigma-Aldrich, St. Louis, MO, USA) and the rabbit polyclonal anti-PVL LukS subunit (0.5 μg/mL, catalog number ab190473, Abcam, Cambridge, United Kingdom) rotated the culture overnight at 4°C. Subsequently, the membranes were incubated again for 1 h with the addition of HRP-labeled goat anti-rabbit IgG (1:5,000, catalog number A0208, Beyotime, Shanghai, China). After cleaning the PVDF membrane, ECL solution (Cat#P0018. Beyotime, Shanghai, China) was added to observe the Western blot bands under a chemiluminescence imager (FUSION FX Spectra system, VILBER, China).

Urease activity.

Hinokiflavone (32 μg/mL) was added to mediums of S. aureus USA300 and USA300-ΔclpP with an initial OD₆₀₀ of 0.3 and further incubated until OD₆₀₀ of 2.5. Then, hinokiflavone-treated S. aureus or DMSO-treated S. aureus were inoculated into a urease agar medium containing the indicator phenol red at 37°C for 2 d, respectively. ClpP mutant strain was set up as a control group. The urease activity was assessed by measuring the absorption spectrum(450 nm to 650 nm) with a microplate reader (Multiskan Go, ThermoFisher).

Thermal shift assay (TSA).

To determine the interaction between hinokiflavone and ClpP, a TSA assay was performed. ClpP protein (final concentration, 2 μM), SYPRO Orange, hinokiflavone and TSA buffer (150 mM NaCl, 10 mM HEPES, pH = 7.5) were added to a 96-well optical PCR plate. The mixture was heated at the rate of 1°C/min from 25 to 90°C by a real-time PCR instrument. This interaction was verified by observing the T_m shift after drug treatment. T_m was estimated as the temperature corresponding to 50% denaturation of the protein.

Cellular thermal shift assay (CETSA).

E. coli BL21(DE3) containing pET28a-ClpP plasmid was incubated until OD₆₀₀ of 0.8, and 0.5 mM IPTG was added to induce ClpP protein expression. The supernatant was incubated with hinokiflavone and an equal volume of DMSO at 37°C for 1 h through centrifugation at 18,000 g for 20 min at 4°C. The reaction mixture was heated for 5 min at a temperature gradient of 25.0, 45.0, 51.2, 56.2, 61.4, and 65.0°C, respectively. It is worth noting that the mixture should be placed into ice water for 3 min immediately. Then gently centrifuged at 18,000g for 20 min to yield a supernatant and analyzed by SDS-PAGE. The samples were incubated with Coomassie brilliant blue G-250 stain, and the relative intensities of the indicated proteins were visualized using ImageJ software.

Localized surface plasmon resonance (LSPR).

The interaction between hinokiflavone and ClpP was performed by OpenSPR instruments (Nicoya, Canada). After a stable baseline is reached, PBS containing 1% DMSO (pH = 7.4) is first applied to enable activation of the chip surface. Following this, the ClpP protein was used as an analyte on the sensor chip. After stabilizing the reaction signal, and infusion of each concentration of hinokiflavone, the time for association and dissociation was 240 s, respectively. The results were finally analyzed by Trace Drawer (Ridgeview Instruments ab, Sweden) and One To One.

Molecular docking (MD) and dynamic simulation.

The 3D structure (PDB ID: 3V5E) of ClpP was downloaded from the Protein Data Bank (www.rcsb.org.). The three-dimensional chemical structure of hinokiflavone (CAS-number19202 to 36-9) has been downloaded from PubChem Compound. Subsequently, molecular docking was performed using Autodock vina 1.1.2 (36, 37) to investigate the binding mode of hinokiflavone to ClpP according to previous studies in the literature (15).

Checkerboard susceptibility assay.

The MICs of hinokiflavone or antibiotics against S. aureus USA300 were determined as previously described. Serial dilution antibiotics of hinokiflavone with were mixed in CAMHB and followed by the addition of S. aureus (10⁵ CFU/mL) to the reaction system. The absorbance at 600 nm was measured by incubation in an incubator at 37°C for 16 h. The interaction between the drugs was quantified by determining the fractional inhibitory concentration index (FICI) in vitro. The calculation method of FICI was as described in the previous literature (38). The fractional inhibitory concentration index (FICI) was calculated according to the formula:

$$\text{FICI\hspace{0.17em}}=\frac{{\text{FIC}}_{\text{antibiotic}}+{\text{\hspace{0.17em}FIC}}_{\text{hinokoflavone}}={\text{\hspace{0.17em}MIC}}_{\text{combine}-\text{antibiotic}}}{{\text{MIC}}_{\text{antibiotic}}+{\text{\hspace{0.17em}MIC}}_{\text{combine}-}{}_{\text{hinokiflavone}}/{\text{MIC}}_{\text{hinokiflavone}}}$$

FICI ≤ 0.5 indicates synergism; 0.5 < FICI ≤ 1 indicates additivity; 1< FICI ≤ 2 indicates irrelevance; FICI > 2 indicates antagonism.

Pneumonia model experiment.

Hinokiflavone was evaluated in vivo using a mouse model of pneumonia infection. Liaoning Changsheng Biologicals provided the mice as 6 weeks old male SPF grade C57BL/6J mice, weight 18 to 22 g. The mouse model of S. aureus pneumonia was constructed as previously described (39).

To assess survival rates, mice were randomly divided into 7 groups of 10 mice. Each group was infected with 2 × 10⁸ CFU/30 μL S. aureus. One hour after inoculating, hinokiflavone (100 mg/kg/d) was injected subcutaneously every 12 h to assess the survival rate at 96 h. After 2 h of infection, a specific dosage of hinokiflavone (100 mg/kg·d⁻¹), vancomycin (100 mg/kg·d⁻¹), or the combination of hinokiflavone (100 mg/kg·d⁻¹) and vancomycin (100 mg/kg·d⁻¹) was administered through the subcutaneous injection route for hinokiflavone and the intraperitoneal route for the antibiotic.

Each group of infected mice was assessed further for histological changes in lung pathology, bacterial load, and inflammatory factor levels in alveolar lavage fluid. Each mouse was given S. aureus resuspension (1 × 10⁸ CFU), and the route of administration and dose administered was the same as the survival rate assay. Mice were euthanized 48 h after administration, and then all mice were removed from lung tissue under aseptic conditions to observe the changes in their appearance. The collected left lung tissues were weighed, homogenized, and the CFU of bacteria containing each gram of tissue was determined. The right lung was weighed and later fixed in 10% of neutral buffered formalin. The tissues were sectioned, stained with H&E and subjected to analyze histopathological changes.

To determine the inflammatory factor levels, alveolar lavage fluid was collected aseptically from each group. Meanwhile, the cytokine levels (INF-γ, IL-6, TNF-α) were measured by ELISA (catalog number M6140, catalog number M6149, catalog number M6152, US EVERBRIGHT, China).

Statistical analysis.

Each study had three independent replicates. Comparisons between groups were made statistically using the Student's t test or analysis of variance on a one-way basis (ANOVA), and survival profiles were analyzed with Log-rank (Mantel-Cox) tests. P < 0.05 was considered statistically meaningful.

Ethical statement.

All animal experiments in this study were strictly conducted in accordance with the guidelines of the Experimental Animal Ethics Committee of Changchun University of Chinese Medicine.

Data availability.

The data that support the findings of this study are available from the corresponding author.