Corilagin alleviates Staphylococcus aureus pathogenicity by interacting with amidase and α-hemolysin.

Introduction
The notorious Staphylococcus aureus (S. aureus) has posed a challenge to food safety, public health and livestock farming, Diseases such as pneumonia caused by this pathogen have significantly increased medical costs and reduced the economic benefits of livestock and poultry farming1–5. The development of bacterial resistance especially the widespread of methicillin-resistant S. aureus (MRSA) has led to a bottleneck in the sustainability of antibiotics against S. aureus infections6–8. Thus, the development of novel adjuvants with sustainable application prospects is of great importance and urgency.
S. aureus amidase (amiA) is a cell wall hydrolase that involves in the separation of bacterial parents and offspring, bacterial adherence to host cells and biofilm formation. AmiA can split the amide bond between N-acetylamino and L-alanine residues of stem peptides, this cause break of the peptidoglycan that connects parents and offspring and achieve cell division and proliferation9–11. When aimA is inhibited or inactivated, the peptidoglycan (PGN) cannot be splited sufficient, the division between parents and offspring is defective, then bacterial cells grow too large and bacterial proliferation is slowed down12–14. In addition, amiA plays an important role in early bacterial surface adhesion, biofilm formation and maintenance, as it is an important component required for biofilm formation and development15–17. Bacteria that formed biofilm and adhere to smooth surfaces are protected by an extracellular polymer matrix, as a result, the antibiotic could not enter the bacteria sufficient or inhibited by the specific physiological functions of the bacteria in the biofilm to lose its bactericidal ability. S. aureus biofilm is one of the most common causes of implant-associated infections18,19, which mainly occurred in mechanical heart valves, prostheses and other implanted medical devices, this brings a heavy burden to patients. It was found that amiA was overexpressed under biofilm conditions, then it promotes the release of genomic DNA by cleaving bacterial cell subsets. Inhibition or inactivation of amiA will result in significantly reduction on cleavage, genomic DNA release, and biofilm development, amiA-deficient mutants of S. aureus have been shown to have a reduced ability to form biofilms, suggesting that amiA is involved in this process15,20,21. Therefore, inhibiting the function of amiA can disrupt the biofilm formation of S. aureus. Overall, amiA plays an important role in bacterial parental and progeny separation, biofilm formation and development, promising its potential as a target for the development of new therapies to combat S. aureus infection. S. aureus α-hemolysin (Hla), one of the important exotoxins secreted by this pathogen, has been well known for us. Its function and mechanism on promoting S. aureus infections have also been elucidated, such as the formation of polymers that consisted of seven monomers, which can punch cells and trigger cytotoxic and inflammation, there are some compounds have been identified as Hla inhibitors22–25. However, compound targeting amiA and Hla simultaneously to suppress S. aureus has not been reported.
Corilagin is a polyphenolic compound that exist in a variety of medicinal plants and have diverse biological and pharmacological activities such as antioxidant, anti-inflammation, liver-protective, antiviral, antihypertensive26–30, etc. In recent years, corilagin attract more and more attention31,32, it has been reported possessing anti-S. aureus characters but the exact mechanism was still unclear33,34. considering the critical roles of amiA and the inhibition phenotypes of corilagin, we performed docking calculation assay and found that they have excellent binding potential. Then the inhibition of the bacterial growth, biofilm formation and the adhesion of S. aureus to host cells provided evidence for the binding. The binding mechanism analysis showed that corilagin interacted with the critical residues that comprise the active pocket. As a important virulence factor of S. aureus, Hla has been defined as ideal target for developing inhibitors of S. aureus. Here, based on the multi-target theory of natural compounds, we investigate the inhibitory effect of corilagin against Hla and found that it inhibited the hemolytic activity of bacterial culture supernatant and alleviated the cytotoxicity from this pathogen. In vivo, corilagin reduced bacterial colonization in mice lung tissue, alleviated inflammation and edema, and improved the survival of the infected mice. And it showed superiority on protecting Galleria mellonella from S. aureus infection than ampicillin (Amp), but did not have side-affect. These results provide a potential lead compound for the development of S. aureus infection inhibitors.

Materials and methods

Reagents, cell lines, bacteria and growth conditions
Corilagin (purity ≥ 98%) was purchased from Chengdu Herbpurify CO.,LTD. Sterile de-fibrous sheep blood and Luria-Bertani (LB) culture medium were purchased from Qingdao hopebio Co., Ltd. Lactate dehydrogenase (LDH) and enzyme linked immunosorbent assay (ELISA) kits were purchased from Shanghai Beyotime Biotechnology Co., Ltd. Mouse macrophages RAW264.7, human lung cancer epithelial cells A549, and S. aureus MRSA strain USA300 were purchased from American Type Culture Collection (ATCC). Dulbecco’s modified eagle medium (DMEM), 0.25% trypsin-EDTA, penicillin-streptomycin solution and fetal bovine serum (FBS) were purchased from Sangon Biotech (Shanghai, China) Co.,Ltd. RAW264.7 and A549 cell were cultured in DMEM with 10% FBS at 37 ℃ with 5% CO2. S. aureus USA300 was cultured in LB medium at 37 ℃ with shaking.

Molecular docking and molecular simulation
This assay was carried out with reference to the methods previously reported35,36. Briefly, the crystal structures of amiA and corilagin were downloaded from RCSB PDB (ID: 4knl) and PubChem (CAS number: 23094-69-1). Before docking, amiA and corilagin were treated with AutoDock Tools 1.5.6, and AutoDock Vina37 was used to perform the docking calculation. The potential binding model was used for molecular simulation based on GROMACS 2020.6 version38, the amber ff14SB force field39 and the TIP3P water model were used. Other details were same as the references. RMSD (Root Mean Square Deviation) and hydrogen bonds (Hbonds) were analyzed. The binding free energy was calculated based on Molecular Mechanics / Poisson Boltzmann Surface Area (MMPBSA) method that reported previously40. The interact mechanism between Hla (PDB ID:7ahl) and corilagin was analyzed based on the same procedure.

Susceptibility and growth curve
LB medium that includes serious concentrations of corilagin (0, 1, 2, 4, 8, 16, 32, 64, 128 µg/mL) in 96 well plate was prepared based on doubling dilution. Then USA300 was added to each well with a final concentration of 5 × 105 colony-forming units per milliliter (CFUs/mL). The sample was cultured at 37 ℃ static for 24 h. The minimum concentration without visible bacterial growth was defined as the minimum inhibitory concentration (MIC) of corilagin. The optical density at 600 nm (OD600) of the USA300 strain that cultured overnight was adjusted to approximately 0.173. Then different concentrations of corilagin (0, 4, 8, 16, 32 ug/mL) were added to the samples and cultured at 37 ℃ with shaking. The samples were harvested every hour to detect OD600 values, and the growth curves were plotted to explore the effect of corilagin against the growth of the bacteria.

Biofilm inhibition
The assay was carried out in a 96-well cell cultural plate. LB medium containing different concentrations of corilagin (0, 4, 8, 16, 32, 64 µg/mL) was obtained by the double dilution method, and then S. aureus USA300 strain that cultured overnight was added to each well, the final concentration is 3 × 107 CFUs/mL, samples were incubated at 37 ℃. Twenty four hours later, the culture medium was discarded, each well was cleaned with sterile phosphate-buffered saline (PBS) for three times. After drying, 0.1% crystal violet was used to treat each well for thirty minutes. The dyeing solution was discarded and the samples were cleaned. After drying, 33% glacial acetic acid was added to each well to dissolve the samples and OD570 was detected to analyze the influence of corilagin on the formation of S. aureus biofilm. To analyze the colonies in the samples, the plankton were removed and samples were washed, bacteria in biofilm were harvested and 10 µL were coated onto LB agar medium after dilution. After culturing at 37℃ overnight, the colonies were counted.

Hemolytic activity inhibition and Protein secretion assay
S. aureus USA300 (OD600 values approximately 0.2) was co-cultured with various concentrations of corilagin (0, 4, 8 µg/mL) for eight hours. Then the cultural supernatant from samples that containing an equal amount of bacteria were harvested (12000 rpm, 5 min) and treated with a 0.22 μm filter. Sterile de-fibrous sheep blood was added to the supernatant (final concentration 2.5%), and samples were incubated at 37℃. Ten minutes later, supernatant was obtained by centrifuge (12000 rpm, 1 min), and OD543 was detected to analyze the release of hemolysin from different samples. The inhibitory effect of corilagin against Hla protein hemolytic activity was based on the same method. To evaluate the hemolytic activity of corilagin, sterile de-fibrinated red blood cells of sheep (2.5%) in sterile PBS were treated with various concentrations of corilagin (0, 16, 32, 64, 128 µg/mL) for 16 h. Then the image was obtained after centrifugation. To evaluate the effect of corilagin against the secretion of the Hla, proteins in the supernatant were precipitated by using acetone and harvested after centrifuging (12000 rpm, 5 min) at 4 ℃. Then samples were boiled at 100 ℃ for five minutes after sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer was added. Proteins were separated by 10% SDS-PAGE gel, the image was obtained after coomassie brilliant blue dyeing.

LDH detection and cell staining
RAW264.7 cells were seeded in 96 well plate (1 × 104 cells/well) and cultured overnight. The next day, S. aureus USA300 in DMEM with different concentrations of corilagin (0, 4, 16 µg/mL) was used to replace the medium and cultured for 6 h, the multiplicity of infection (MOI) was 50. After centrifugal (1000 rpm, 10 min), the supernatant was used to detect LDH activities, and cells were staining with 50 µg/mL ethidium bromide (EB), images were obtained by using a fluorescence microscopy (IX83, OLYMPUS). The potential cytotoxicity of corilagin (0, 4, 8, 16, 32, 64 µg/mL) was explored by measuring LDH level. Cells treated with 0.1% Triton X-100 or DMEM alone were set as the positive or negative control.

Adhesion inhibition
A549 cells that cultured in 24 well plate (2 × 105 cells/well) were treated with S. aureus USA300 (MOI = 40) and different concentrations of corilagin (0, 8, 32 µg/mL) for one hour. Then, the culture medium was removed, and cells were washed with sterile PBS. Cells were harvested and 10 µL samples were used to count colonies after being coated onto LB agar medium and cultured overnight.

Mouse pneumonia model
The animal assays carried out here are in accordance with the requirements of the Animal Care and Use Committee of Jilin University and ARRIVE guidelines. All the experimental protocols were approved by the Animal Care and Use Committee of Jilin University. C57BL/6J mouse (male, 6–8 weeks, approximately 20 g) were obtained from Liaoning Changsheng Biotechnology Co., Ltd (Shenyang, China), water and food are freely accessible. A mouse pneumonia model (4 × 108 CFUs/mouse) was constructed by inoculation through the left nose. Two hours later, corilagin (80 mg/kg) was injected into the mice by subcutaneous injection, the infected group and the control group (blank) were treated with equal volumes of solvent or PBS. The mortality of mice was counted at the indicated time points to analyze the protective effects of corilagin against mouse pneumonia induced by S. aureus. To analyze other indicators, mice were treated with a sub-lethal dose of S. aureus USA300 (2 × 108 CFUs/mouse), other treatments was accordance with the survival assay. After 48 h, pentobarbital sodium (50 mg/kg) was used to anesthetize mice by injection method, then the mice were euthanized by cervical dislocation. Alveolar lavage fluid was harvested and the levels of tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) were measured to evaluate the alleviation effect of corilagin on inflammation. The lung tissue was obtained and homogenized with sterile 0.9% sodium chloride, samples were coated onto LB agar medium and cultured overnight. The effect of corilagin on the bacterial load of lung tissue was analyzed. The lung tissue lesions were observed after hematoxylin-eosin staining.

G. mellonella infection and corilagin biosafety evaluation.
G. mellonella was purchased from Henan Jiyuan Baiyun Industrial Co., Ltd. (Jiyuan, China), the weight of each individual was approximately 400 mg. S. aureus USA300 was injected to G. mellonella (5 × 105 CFUs/sample), then corilagin or Amp was injected to the samples with a final concentration of 12.5 mg/kg. Samples treated with S. aureus USA300 but did not receive corilagin treatment were defined as infection group, samples injected with equal volume sterile PBS were defined as blank control group. The survival of the individual was observed for a 12 h interval, ten individuals were arranged to each group. To evaluate the biosecurity, various concentrations of corilagin (0, 25, 50, 100 mg/kg) were injected into G. mellonella and the survival was monitored.

Statistical analysis
The experimental data were presented in the form of mean with standard deviation (SD), an unpaired t-test method that merged in GraphPad Prism 9.5.0 software was used for statistical analysis. P ≤ 0.05 was defined as significant difference. Three independent experiments were carried out.

Results

Corilagin binds with amiA to inhibit S. aureus USA300 growth and biofilm
Molecular docking results showed that corilagin (Fig. 1a) was directly bound to the active pocket of amiA with an affinity of -10.4 kcal/mol (Fig. 1b). Then the effect of corilagin against S. aureus USA300 growth was evaluated, it was found that the bacteria showed normal growth trend when there was no corilagin existed, but the grown of the bacterial showed different degrees delay when the samples received various concentrations of corilagin treatment (Fig. 1c and d). The MIC values of corilagin against S. aureus USA300 was 128 µg/mL, indicating that the delay of the bacterial grown was not originate from the antibacterial property. A large number of biofilm formation was detected in the untreated group, while the bacterial biofilm formation gradually decreased after samples were treated with various concentrations of corilagin. When the concentration of corilagin was 4 and 16 µg/mL, the biofilm formation was reduced to 47.77% and 18.54% of the untreated group (Fig. 1e), and the log10 values of bacterial density were also decreased by approximately 2 and 4 folds (Fig. 1f), respectively. AmiA can cleave the bond between the carbohydrate and the peptide moieties of PGN that connecting parents and offspring, and amiA is also involves in the formation of biofilm, our results showed that the binding of corilagin on the active pocket of amiA affected the function of amiA to clave the connecting between parents and daughter, which results in a delay in the growth, and reduced the formation of bacterial biofilm.

Residues involves in the binding between corilagin and amiA
Preliminary analysis showed that His370, Asn284 and Asn287 formed Hbonds with corilagin, respectively, and Val312, Ala288, His382, Gly379, Thr380, and Gly313 also generated interaction with corilagin (Fig. 2a). For confirmation, a 100 ns kinetic simulation experiment was carried out, the RMSD fluctuation and the distance between corilagin and amiA confirmed that the binding was stable and reliable (Fig. 2b and c). In the last 10 ns of equilibrium trajectory, more than twenty pairs Hbonds was detected, though most of them have low occupancy (Fig. 2d and e), the three Hbonds that with high occupancy were existed between Asn284, Asn287, His382 and corilagin (Fig. 2f).

The binding free energy analysis showed that the van der waals (vdw) interaction (-114.31 kJ/mol) was important on promoting the binding, and the electrostatic interaction (ele) also promoted (-43.28 kJ/mol) the binding (Fig. 3a). The results of residual energy decomposition showed that ten residues were involving in the interaction, among which Glu324, Ala288, Asp384, Asp266, Glu277, Asn287 and His370 contributed more energies with the bind free energy values of -6.67 kJ/mol, -5.78 kJ/mol, -3.30 kJ/mol, -3.21 kJ/mol, -3.03 kJ/mol, -4.04 kJ/mol, -1.8 1 kJ/mol (Fig. 3b and c). More specifically, the main energies between Glu324 and corilagin was ele (-3.61 kJ/mol) and solvation energy (sol − 2.79 kJ/mol) (Fig. 3d and e), and Ala288 mainly contributes vdw forces (-5.23 kJ/mol) (Fig. 3f), both ele (-3.74 kJ/mol) and vdw forces (-11.98 kJ/mol) were generated between Asn287 and corilagin (Fig. 3e and f), Asp384, Asp266 and Glu277 also provided sol and ele (Fig. 3d and e) to promote the binding. As a critical enzyme domain of S. aureus autolysin, the amino acid sequences of amiA from different staphylococcal strains are highly conserved41(Figure S3), indicating that corilagin maybe have anti-staphylococcal infection potential by targeting amiA.

Corilagin surpresses Hla function by dircet binding and reducing production
Hla is a well-known key virulence factor of S. aureus, it plays a crucial role in the infection process of S. aureus. Based on the multi-target theory hypothesis, we explore the effect of corilagin against Hla. It was found that large amounts of hemolysin were detected in the free-growing group, indicating that Hla was secreted into the supernatant and led to lysis of sheep red blood cells. However, the release of hemolysin reduced to 10.96% and 5.58% in the samples that were cultured with 4 µg/mL and 8 µg/mL corilagin (Fig. 4a). Consistently, a large amount of Hla was detected in the supernatant of the free group, but the Hla secretion in different concentrations of corilagin treatment groups decreased gradually (Fig. 4b), indicating that corilagin inhibits the secretion of Hla to the supernatant. For further confirmation, hemolytic test was carried out with purified Hla protein, and it was found that the levels of hemoglobin that detected in the corilagin treatment groups were lower than those in the control group (Fig. 4c and d). The inhibition of corilagin on purified Hla protein suggests a direct interaction was existed between them, which was confirmed by the docking result, as corilagin was located on the binding pocket of Hla with an affinity of -9.5 kcal/mol (Fig. 4e).

Critical residues are identified for the binding
To explore the interactive mechanism between corilagin and Hla, a 100 ns molecular simulation assay was performed. The stability and persistence of the binding between them was confirmed by the structural superposition, the RMSD fluctuation and the distance between corilagin and Hla (Fig. 5a, b and c). The predictive analysis showed that Gln194, Asn176 and Lys198 formed Hbonds with corilagin, and Gln177, Trp179, Arg200, Tyr182 and Met197 also interacted with corilagin (Fig. 5d). More exactly, 25 pairs of Hbonds appeared in the last 10 ns of equilibrium trajectory (Fig. 5e), six of them occupied more than 39.0% (Fig. 5f). The residues involved in the Hbonds interaction included Trp179, Gln194, Lys198, and Asn176.

To confirm the binding free energy generated between corilagin and Hla, MMPBSA method was used to perform the calculation. It was found that the total binding free energy between corilagin and Hla was − 73.77 kJ/mol, which includes vdw (-172.02 kJ/mol), ele (-90.66 kJ/mol) and sol (167.31 kJ/mol) (Fig. 6a). Asn176, Gln177, Trp179, Tyr182, Gln194 and Arg200 contributed more energies to the binding, the values were − 5.097 kJ/mol, -5.901 kJ/mol, -10.933 kJ/mol, -5.095 kJ/mol, -4.55 kJ/mol and − 9.438 kJ/mol (Fig. 6b). More vdw interaction were generated between Asn176 (-3.81 kJ/mol), Gln177 (-7.01 kJ/mol), Trp179 (-15.58 kJ/mol), Tyr182 (-10.84 kJ/mol), Gln194 (-8.34 kJ/mol) and Arg200 (-16.51 kJ/mol) (Fig. 6c), the sol for most of these residues were positive number, indicating these residues may have stronger hydrophobic effect (Fig. 6d), Asn176 and Arg200 contributed more ele (Fig. 6e). For further confirmation, the binding free energies between these residues mutants and corilagin were calculated, it was found that the binding free energies between these mutants and corilagin reduced significantly except for Q177A (Fig. 6f), indicating Arg200, Tyr182, Asn176, Trp179, Tyr182 and Gln194 were important for the binding.

Corilagin alleviates S. aureus-mediated cytotoxicity and cell adhesion
The cells treated with 0.1% Triton X-100 definitely died because a large amount of LDH release was detected, while the LDH levels detected in RAW264.7 cells that treated with different concentrations of corilagin did not different significantly from those treated with DMEM alone (Fig. 7a), indicating that corilagin did not show cytotoxicity under the test concentrations. RAW264.7 cells treated with S. aureus USA300 released lots of LDH, but the LDH levels decreased to 65.46% and 33.82% under 4–16 µg/mL corilagin treatments (Fig. 7b). EB staining showed that a large number of dead cells were detected in the S. aureus USA300 treatment group, while the number of cells stained by EB was significantly reduced in the corilagin treatment group (Fig. 7c). S. aureus USA300 adhered to human lung epithelial cells abilities reduced to 62.13% and 31.06% under 8–32 µg/mL corilagin treatments when compared with the untreated group (Fig. 7d). These results suggest that corilagin can significantly mitigate the cytotoxicity and the adhesion of S. aureus to cells.

Corilagin protects mice from S. aureus USA300 pneumonia
To evaluate the application potential of coriagin in the future. We established mice infection model. The dead mice were found after 24 h of infection in the infection group, when the time reached 72 h, the probability of death in this group was 96.67%; while, when the infected mice received coriagin treatment, the death time was delayed to 36 h, and the final death probability of this group was 53.33% (Fig. 8a). The lung tissue of the mice that from the infection group was destroyed and was infiltrated by a large number of inflammatory factors, while the inflammatory infiltration and the damage degree of the samples from the corilagin treatment group was weakened (Fig. 8b). Besides, other indicators from the corilagin treatment group were lower than those from the infection group, exactly, the logarithm value of the bacterial load reduced approximately 1.2 fold (Fig. 8c), the ratio of lung weight (wet/dry) was close to the control group (Fig. 8d), and the level of IL-1β and TNF-α reduced approximately half when compared with the infection group (Fig. 8e). These results indicate that corilagin reduced the bacterial colonization in the lung tissue of the infected mice, and alleviated the pathological damage by inhibiting edema and inflammation, thus it delayed the death time and decreased the mortality of the infected mice. These results provide valuable opportunities for clinical infection treatment.

Corilagin is safe under the tested concentrations and is superior to Amp on protecting G. mellonella from S. aureus infection
To evaluate the bio-safety of corilagin, sterile de-fibrinated red blood cells of sheep were treated with various concentrations of corilagin for 16 h, but we did not observe hemolytic activity (Figure S1a), suggesting corilagin did not have cytotoxicity to mammalian cell. This was confirmed by the G. mellonella assay, because all samples survive when they received various concentrations of corilagin treatment (Figure S1b), which evident the bio-safety of corilagin in vivo. The probability of survival of the infected G. mellonella in the infection or Amp treatment groups drops sharply within 72 h, and the final survival rate for these two groups were 6.67% and 10.0%, but the dead individuals were found in the corilagin treatment group at 36 h after infection, and the survival probability of this group was 50.0% (Figure S2). These results provide evidence for the bio-safety of corilagin and the superiority of this compound on combating S. aureus infection than Amp.

Discussion
The crystal structure of amiA domain has been resolved, a large number of conserved residues in this protein domain form a spacious substrate binding pocket, among them, Thr267, Glu277, Met281 and P293 form the bottom of the binding pocket. His370, Asp384 and His265 is the key active site of amiA protein. AmiA completely loses the ability to hydrolyze PGN after His370 was mutated. In addition, Glu324, His382, Asp266, Glu324 and Ala288 also play an important role in hydrolyzing PGN42. In this study, it was found that corilagin directly binds to the active pocket of amiA and has a direct and strong interaction with the active sites His370 and Asp384, as the energy decomposition of residues showed that these residues contributed more energy. Besides, Asn284, Asn287, His382 and His370 formed Hbonds with corilagin, indicating that corilagin hinders the binding of amiA to its substrate PGN through direct steric hindrance. This binding leads to the PGN that connecting parent and offspring could not be split sufficient, the direct results for this was bacterial proliferation slowing down. In addition, amiA was also shown to be directly related to the formation of bacterial biofilms15. Here, the formation of S. aureus USA300 biofilm decreased significantly when the pathogen was treated with different concentrations of corilagin, indicating corilagin successfully inhibits the biofilm formation of S. aureus USA300 by targeting amiA.
The rim region of Hla plays an important role in its binding to cell membrane phospholipid molecules, among which Trp179, Tyr182, Trp187 and Arg200 form the rim side of the crevice, and they are potential binding sites of phospholipid molecules43. Some compounds have been identified as inhibitors of Hla22,44–46, one of them have been reported to be located on the triangle region of the protein to perform an inhibitory effect47. Different from this action mode, here, we found that corilagin binds to the rim region of Hla and interacts with Asn176, Trp179, Tyr182, Gln194 and Arg200. This binding may hinder the binding of Hla to cell membrane phospholipid molecules, thus affecting its biological function. However, the inhibitory effect of corilagin against Hla does not stop here, it also reduces the hemolytic activity of bacterial cultural supernatant by reducing Hla secretion. These results also indicate that corilagin may have other potential targets that have not been disclosed, which requires in-depth research in the future.
Corilagin, as a polyphenolic compound, has been reported to restore oxacillin susceptibility to S. aureus48,49. Noteably, researcher developed a safe micro-particulate system for corilagin to improve its bioavailability in vivo33. In this study, we found that corilagin reduces S. aureus-mediated cytotoxicity and adhesion to host cells, and it improves the survival of S. aureus USA300-infected mouse pneumonia model by targeting amiA and Hla. Corilagin relieves the lung edema symptoms, inflammation and reduces bacterial colonization in the lungs. These results enrich the function of this compound and expand its potential to be developed as an antibacterial drug.
Bio-saftey, bioavailability and toxicity are important parameters of natural active compounds, which directly affect their application in the future. Here, we evaluate the bio-saftey of corilagin based on hemolysis assay, LDH activity assay and G. mellonella assay, corilagin did not show cytotoxicity to mammalian cell and did not show any toxic effects to G. mellonella, which provides evidence for the bio-safety of corilagin. The drug metabolism and bioavailability of corilagin in rats have been disclosed, the authors investigated the pharmacokinetics and bio-availability of corilagin in rats via oral and intravenous administration, they found that the blood concentration of corilagin reaches peak in approximately 2 h after oral administration, but the bioavailability of oral is relatively low. After intravenous injection, corilagin is mainly concentrated in the plasma, the time to reach the peak blood concentration and the maximum blood concentration are superior to those of oral administration, the authors suggest that intravenous injection may have a better therapeutic effect. The highly conserved of amiA in staphylococcal41, indicating corilagin possesses the potential for broad-spectrum anti-staphylococcal infection.

Conclusion
Corilagin interacts with active sites residues of amiA, occupies its active center and inhibits its ability to hydrolyze PGN, resulting in slower bacterial growth and fewer biofilm formation. The dual function of corilagin reduces the hemolytic capacity of Hla and bacterial culture supernatants by a direct binding and reducing Hla secretion. In vivo, corilagin reduces bacterial-mediated cytotoxicity and adhesion to host cells, and showed a significant protective effect against a mouse pneumonia model infected by S. aureus USA300. Corilagin shows more excellent ability than Amp on protecting G. mellonella from S. aureus infection. These results promise the potential of corilagin to be developed as a drug to treat S. aureus infections.

Supplementary Information
Below is the link to the electronic supplementary material.