Insight into the Dual Inhibition Mechanism of Corilagin against MRSA Serine/Threonine Phosphatase (Stp1) by Molecular Modeling.

Serine/threonine phosphatase (Stp1) is known to be involved in the regulation of cysteine phosphorylation levels in many different pathways, such as virulence factor regulation in methicillin-resistant Staphylococcus aureus (MRSA). Therefore, Stp1 can be used as a potential target for inhibiting MRSA infection. In this study, using virtual screening, we found that corilagin, a natural compound, was screened as a potential Stp1 inhibitor. Then, the phosphatase assay exhibited high inhibitory activity against Stp1. On the basis of the enzyme kinetics experiment, we found that corilagin exhibited a dual inhibitory mechanism of competitive and allosteric inhibition. To further elucidate the mechanism of interaction between corilagin and Stp1, molecular dynamics (MD) simulations were performed on the Stp1-corilagin complex. Consistent with the mutagenesis assays and fluorescence quenching assays results, the competitive and allosteric binding sites of corilagin with Stp1 were identified. In the competitive binding site of Stp1, Asn162, Ile164, Tyr199, and Lys232 were found to play a key role in this complex. In the allosteric binding site, hydrophobic interaction was the main binding force. The Asn142, Val145, Leu146, Pro152, and Phe179 residues of Stp1 were found to play a critical role in the binding of corilagin with Stp1. In this study, we used MD simulation to reveal the ligand-protein interactions, providing a theoretical basis. This research work, thus, lays down the foundation for the development of new Stp1 inhibitors to be utilized in the future.

Introduction

Staphylococcus aureus (S. aureus) is a common pathogen that attaches to the human skin with a potentially fatal infection through wounds and mucous membranes.[1] Genome plasticity and the widespread use of antibiotics has led to the increased development of antibiotic resistance.[2] The emergence of methicillin-resistant S. aureus (MRSA), first isolated in the 1960s, has gradually become an important hospital pathogen that considerably impacts vulnerable patients and staff.[3,4] Since 2002, community-acquired and intrafamily MRSA infections have increased rapidly.[5] This indicates that MRSA spreads gradually. Therefore, it is urgent to find effective inhibitors to reduce MRSA toxicity.

Currently, the discovery of effective regulators of bacterial virulence factors is one of the strategies for treating such bacterial infections. Serine/threonine phosphatases (Stp1) and kinases (Stk1) are involved in the regulation of the cysteine phosphorylation level of virulence factors, such as the SarA/MgrA family proteins.[6] Stp1 positively regulates the expression of virulence factors, and Stk1 plays the opposite role.[7] Stp1 is a metal-dependent protein and belongs to the phosphatase/protein phosphatase 2C (PPM/PP2C) family. The four metal ions in the Stp1 structure are localized to the catalytically active center, of which M4 is a unique structural domain of Stp1. Metal ions play a crucial role in substrate binding and catalytic activity.[8]

Stp1 is involved in the regulation of toxins, such as hemolysin,[9] superantigen-like protein, and phenol-soluble modulin.[10] In the murine bacteremia model, the Stp1 deletion mutant strain not only exhibits an improved survival rate but also does not form liver abscess compared to that exhibited by the normal strain.[10] Additionally, Stp1 plays a crucial role in cell wall biosynthesis.[11] At present, there are few studies on Stp1 inhibitors, mainly focusing on 5,5′-methylenedisalicylic acid (MDSA)[12] and aurintricarboxylic acid (ATA).[11] These studies show that the inhibition of Stp1 can reduce the toxicity of S. aureus and improve the survival rate of the infected mice. Therefore, Stp1 is considered an effective target for reducing S. aureus toxicity.

Corilagin is a gallotannin derived from a Chinese herbal plant.[13] Since Kakiuchi et al. discovered that it can inhibit reverse transcriptase activity in 1985,[14] many researchers have subsequently reported on several additional pharmacological activities, such as antibacterial,[15] antihypertensive,[16] anti-inflammatory,[17] anticancer,[18] and antiviral.[19] Simultaneously, corilagin derived from herbal medicine has low toxicity[13] and is considered a highly active substance.

At present, there are few studies related to effective Stp1 inhibitors.[12,8] In this study, corilagin (Figure A) was screened as a potential Stp1 inhibitor by virtual screening. The phosphatase assay showed that corilagin exhibited a significant inhibitory effect against Stp1. However, corilagin inhibitory mechanisms at the atomic level are unclear. Research at the atomic level is conducive to elucidating the mechanism of drugs at the molecular level and providing useful information to assist drug design. This makes the study of dynamic characteristics more meaningful. Therefore, to further study the inhibition mechanism, molecular dynamics (MD) simulation, binding free energy calculations, and mutagenesis assay were used to explore and verify the binding mode and key amino acid residues. The results of this study will provide new insights into the strategies for the development of new antitoxins for reducing S. aureus toxicity.

Figure 1

Inhibition of Stp1 by corilagin. (A) Structure of corilagin. (B) Inhibitory effect of corilagin on Stp1. **Compared with the control group, P < 0.001.

Results and Discussion

Corilagin Inhibits Stp1 Activity

To screen a large number of Stp1 inhibitors, the ZINC database was used as a ligand to dock with Stp1. It can obtain the binding affinity (kcal/mol) of the ligand–Stp1 complexes. The top 50 ligands (binding affinity < −8.0 kcal/mol) were screened using binding affinity. Subsequently, 12 compounds (binding affinity < −8.4 kcal/mol) (Table S1) were screened by PyMOL[20] and Ligplus[21] software. The software were used to analyze the binding sites of the ligand and the catalytically active region of Stp1. Therefore, through virtual screening, corilagin was found to potentially inhibit Stp1. The inhibitory activities of 12 compounds against Stp1 were tested by phosphatase assay. Corilagin was found to inhibit Stp1. Then, Stp1 was tested using different concentrations of corilagin through phosphatase assay. The experimental results revealed that corilagin exhibited a high inhibitory effect on Stp1. As the concentration of corilagin was increased, the activity of Stp1 decreased and the inhibitory effect was found to be increased (Figure B). When the concentration of corilagin was 4 μg/mL, the activity of Stp1 declined. At a concentration of 32 μg/mL of corilagin, the inhibition of Stp1 was found to be more than 90%.

To further investigate the mechanism of the inhibition of Stp1 by corilagin, the enzymatic–kinetic assays were performed. As shown in Figure A, both KM and Vmax of the reaction were changed. Therefore, corilagin exhibited a dual inhibitory effect, which means that it acted as a competitive as well as a noncompetitive inhibitor of Stp1.

Figure 2

Determination of the inhibitory effect of corilagin on Stp1. (A) Lineweaver–Burk plot of Stp1 activity against different concentrations of corilagin indicates that corilagin is a mixed inhibitor (competitive as well as noncompetitive inhibitor). (B, C) Two low-energy models through molecular docking. These two binding regions correspond to the catalytically active region of Stp1 (competitive binding site) (B) and the vicinity of Stp1 “flap region” (allosteric binding site) (C).

Determination of the Interaction Mechanism between Corilagin and Stp1

The initial structure of the corilagin and Stp1 complex was obtained by performing molecular docking. In molecular docking, the conformation exhibiting the lowest energy was selected for further analysis. The data revealed that the two selected binding sites were present near the Mn2+ of Stp1 active site (the binding affinity is −9.8 kcal/mol) (Figure B) and near the flap region of Stp1 (the binding affinity is −9.4 kcal/mol) (Figure C). These two binding modes were predicted to be responsible for competitive and allosteric inhibitions, respectively. The active center of Stp1 directly participates in binding with the substrate, while the flexible flap domain promotes substrate binding and participates in catalytic dephosphorylation.[8] To verify the above prediction and obtain the stable binding mode, a standard MD simulation analysis was performed using the Stp1–corilagin complex.

Identification of the Binding Sites in the Competitive Binding Conformation

To investigate the ligand–protein interactions of corilagin and Stp1, a molecular dynamics simulation of 200 ns was performed using the complex system. To determine whether the system reached an equilibrium, the root mean square deviation (RMSD) of the Cα atom of the protein backbone was calculated. In Figure A, the RMSD values of protein changed significantly before 140 ns. During the 150–200 ns simulation, the RMSD values of Stp1 and corilagin became stable (Figure A). This result showed that during the last 50 ns of the simulation, the system was in equilibrium. Therefore, the trajectory of this molecular dynamics simulation was used for further analysis. In Figure B,C, the stable structure was obtained using potential energy value in the last 50 ns MD simulation. In the Stp1–corilagin complex, the interaction between Stp1 and corilagin was mainly through hydrogen bond and hydrophobic interaction (Figure B,C). Corilagin could stably bind to the active center of Stp1. In detail, Asn162, Tyr199, and Lys232 residues of Stp1 were identified to be in close proximity to the two carbonyl groups of corilagin 12-membered oxygen-containing heterocycle. Ile164 residue of Stp1 interacted with the hydroxyl group on the corilagin benzene ring. The root-mean-square fluctuation (RMSF) describes the fluctuation of each atom or residue in a certain period of time to obtain the flexible change information of the system. As shown in Figure A, the complex RMSF value of some amino acid residues was less than that of free protein. In particular, compared with the free protein, the RMSF of the residue at the binding site of this complex (<0.15 nm) decreased. This indicated that these residues became more rigid due to their low flexibility when combined with corilagin.

Figure 3

Demonstration of the potently competitive binding mode of Stp1 and corilagin through molecular modeling. (A) RMSD values of Stp1 and corilagin in 200 ns. (B) Stable three-dimensional (3D) structure of Stp1 binding with corilagin based on MD simulation. (C) Detailed presentation of the binding sites of competitive binding mode.

Figure 4

Confirmation of the binding sites of the competitive binding mode. (A) RMSF of the residues in Stp1 and corilagin–Stp1 complex. (B) Decomposed binding free energies of the residues in the binding site. (C) Enzyme inhibition assays using corilagin and wild-type-, I164A-, and Y199A-Stp1.

To further explore the contribution of the residues around the binding site to the system, the molecular-mechanics-generalized born surface area (MM-GBSA) method was used to calculate the individual contribution of van der Waals force (ΔEvdw), electrostatic force (ΔEele), solvation (ΔEsol), and total binding energy (ΔEtotal). In Figure B, Asn162 exhibited high van der Waals (ΔEvdw = −3.80 kcal/mol) and electrostatic contribution (ΔEele = −5.61 kcal/mol), with ΔEtotal value of −4.16 kcal/mol, indicating that a hydrogen bond can be formed between Asn162 and corilagin. At the same time, the total binding energy contribution (ΔEtotal) of Lys232 was −1.36 kcal/mol, indicating that it had a hydrophobic interaction with the O12 of the corilagin carbonyl group. Additionally, the electrostatic and van der Waals forces of Tyr199 contributed significantly, leading to high total binding energy contribution (ΔEtotal = −1.86 kcal/mol). This indicated that there existed a hydrogen bond between Tyr199 and corilagin. Ile164 had a high total binding energy (ΔEtotal = −0.65 kcal/mol) due to van der Waals contribution (ΔEvdw = −0.58 kcal/mol). This was due to the hydrophobic interaction of the hydroxyl group on the corilagin benzene ring and the methylene group of Ile164.

In Figure A, the number of hydrogen bonds fluctuated between 1–3 in molecular simulations, which indicated that one to three hydrogen bonds were formed between corilagin and Stp1. Simultaneously, LigPlus software[21] was used to analyze the interaction between corilagin and Stp1 binding site residues. The last 50 ns average structure was the input file. The 12-membered heterocyclic ring of corilagin formed hydrogen bonds with Asn162 and Tyr199 residues, respectively (Figure B). The occupancy (%) of the two hydrogen bonds was higher than 80%, indicating that the hydrogen bonds were highly stable (Table ).

Figure 5

Interaction between corilagin and Stp1 demonstrated in the competitive binding mode. (A) Number of hydrogen bonds between corilagin and Stp1 during the last 50 ns simulation process. (B) Interaction between corilagin and the residues of binding sites of Stp1 identified using LigPlus software.

Table 1

Stp1–Corilagin H-Bonds in MD Simulations

acceptor	donor	presence (%)	distance (Å)
corilagin O8	Asn162 −NH2	80.39	2.88
corilagin O12	Tyr199 −OH	89.22	2.82

In Table , residues Asn162, Ile164, Tyr199, and Lys232 of Stp1 were found to be close to corilagin, and the distance was <3 nm. The above results indicated that the Asn162, Ile164, Tyr199, and Lys232 residues are the key amino acids that are responsible for the competitive binding of corilagin with Stp1.

Table 2

Radial Distance of the Protein Residues and Corilagin

residues	distance (nm)
Asn162	2.620
Ile164	2.168
Tyr199	2.112
Lys232	2.770

To confirm the above prediction, the residues of Ile164 and Tyr199 were mutated to alanine, and MD simulations were performed on I164A-Stp1–Corilagin and Y199A-Stp1–Corilagin complexes. Subsequently, the binding free energy of complexes was calculated by the MM-GBSA method. The ΔGbind values of wild-type (WT)-Stp1–, I164A-Stp1–, and Y199A-Stp1–corilagin complexes were −21.70, −17.95, and −18.36 kcal/mol, respectively (Table ). In addition, a fluorescence quenching experiment was performed using two different Stp1 mutants. It was found that the KA value of the interaction between corilagin and Stp1 decreased in the following order: WT > Y199A > I164A (Table ). This result reveals that the WT-Stp1 binds to corilagin with the highest affinity. The inhibitory effect of corilagin on the Stp1 mutants was also measured using the phosphatase assay (Figure C). I164A-Stp1 and Y199A-Stp1 mutants exhibited well phosphatase activity, but the drug inhibitory effect on the two mutants was found to be reduced. These experimental results indicate that the incorporation of mutations in binding site residues reduced the inhibitory effect of corilagin. These results therefore demonstrated the reliability and stability of the molecular simulation results.

Table 3

Binding Free Energy and Binding Constants of WT-Stp1–Corilagin, I164A-Stp1–Corilagin, and Y199A-Stp1–Corilagin Complexes in the Competitive Binding Conformation

energy components (kcal/mol)	Stp1–corilagin	I164A–corilagin	Y199A–corilagin
ΔEele	–29.62	–32.72	–26.95
ΔEvdw	–37.87	–35.63	–30.56
ΔEMM	–67.49	–68.35	–57.51
ΔGele,sol	47.68	51.89	40.05
ΔGnonpolar,sol	–5.62	–5.00	–4.11
ΔGsol	42.06	46.89	35.94
ΔGele,sol + ΔEele	18.07	19.17	13.11
ΔGnonpolar,sol + ΔEvdw	–43.49	–40.63	–34.68
ΔGtotal	–25.43	–21.46	–21.57
–TΔS	3.73	3.51	3.21
ΔGbind	–21.70	–17.95	–18.36
KA(1 × 105) L/mol	5.15	1.13	3.49

Identification of the Binding Sites in the Allosteric Binding Conformations

Based on the results of molecular docking, a 170 ns molecular dynamics simulation was conducted using the Stp1–corilagin complex to explore the binding mechanism of corilagin and Stp1. In Figure A, the RMSD values of the Stp1 and corilagin were found to be stable from 100 ns. Therefore, during the simulation process between 100 and 170 ns, the complex was found to be in a stable state. Also, the trajectory of this molecular dynamics simulation was used for further analysis. After the molecular dynamics simulation, the stable binding mode of the corilagin and Stp1 complex was obtained. The stable structure was obtained using the potential energy value in the last 50 ns MD simulation. In Figure B, corilagin is shown to mainly bind to the inactive domain of Stp1 through hydrophobic interaction. Specifically, the Phe179 residue of Stp1 was found to bind to the corilagin benzene ring (Figure C). The side chain of Asn142 residue of Stp1 interacted through hydrophobic interaction with the corilagin ester group. The side chains of Val145, Leu146, and Pro152 were found to be close to the corilagin six-membered heterocyclic ring, indicating that they could potentially interact with the six-membered ring. In Figure A, the RMSF value of the most complex residues was reduced compared with that of the free protein. Specifically, compared with that of the free protein, the RMSF value of the residue at the binding site of the complex was reduced (Figure A), depicting low flexibility. These residues became more rigid upon binding with corilagin.

Figure 6

Demonstration of the potently allosteric binding mode of Stp1 and corilagin through molecular modeling. (A) RMSD values of Stp1 and corilagin in 170 ns MD simulation. (B) Stable 3D structure of Stp1 with corilagin based on the MD simulation. (C) Detail binding sites of allosteric binding mode.

Figure 7

Confirmation of the binding sites of the allosteric binding mode. (A) RMSF of all residues in Stp1 and corilagin–Stp1 complex. (B) Decomposed binding free energies of residues in the binding site. (C) Enzyme inhibition assays of corilagin to wild-type-Stp1, P152A-Stp1, and F179A-Stp1.

According to the free binding energy decomposition (Figure B), the amino acid residue with the highest total binding energy was found to be Asn142 (ΔEtotal = −2.62 kcal/mol), which was mainly due to the van der Waals (ΔEvdw = −2.73 kcal/mol) and the electrostatic forces (ΔEele = −1.14 kcal/mol). This indicated that the carbonyl group on the Asn142 residue of Stp1 and O4 on the corilagin ester group exhibited a strong interaction. The van der Waals value of Phe179 was −1.76 kcal/mol, and thus the total binding energy (ΔEtotal = −1.96 kcal/mol) was strong. This showed that a hydrophobic interaction exists between Phe179 on the benzene ring of Stp1 and O9 present on the corilagin benzene ring. Pro152 and Val145 residues of Stp1 interacted with the O18 H group on the six-membered ring of corilagin, and the total binding energy contribution was high. Additionally, Leu146 of Stp1 and the O11H group of the corilagin six-membered ring exhibited hydrophobic interaction, with ΔEtotal = −1.68 kcal/mol.

The results based on hydrogen bond analysis (Figure A) indicate that the number of hydrogen bonds between Stp1 and corilagin fluctuated between 0 and 2. Upon performing further analysis using Ligplus software (Figure B), Asn142, Val145, Leu146, Pro152, and Phe179 residues of Stp1 were found to interact with corilagin through hydrophobic interaction. Therefore, there were no stable hydrogen bonds identified in this complex. Simultaneously, upon calculating the distance between the residues of Stp1 and corilagin, the results indicated that Asn142, Val145, Leu146, Pro152, and Phe179 residues of Stp1 were found to be close to corilagin (Table ). Therefore, Asn142, Val145, Leu146, Pro152, and Phe179 potentially play an important role in noncompetitive/allosteric inhibition of Stp1 by corilagin.

Figure 8

Interaction between corilagin and Stp1 in the allosteric binding mode. (A) Number of hydrogen bonds between corilagin and Stp1 during the last 50 ns simulation process. (B) Interaction between corilagin and the residues of binding sites in Stp1 using LigPlus software.

Table 4

Radial Distance of the Protein Residues and Corilagin

residues	distance (nm)
Asn142	0.510
Val145	0.448
Leu146	0.498
Pro152	1.004
Phe179	1.020

To verify the above predictions, Pro152 and Phe179 residues of Stp1 were mutated. Fluorescence quenching experiment results showed that the binding constant of the free protein was higher than that of the point mutant (Table ). It was consistent with the results of computational chemistry (Table ). Phosphatase experiment results showed that both P152A-Stp1 and F179A-Stp1 mutations still had the catalytic activity of Stp1, but corilagin almost lost the inhibitory effect against these mutants (Figure C). These results suggest that Asn142, Val145, Leu146, Pro152, and Phe179 residues of Stp1 played a significant role in noncompetitive/allosteric inhibition of Stp1 by corilagin.

Table 5

Binding Free Energy and Binding Constants of the WT-Stp1–Corilagin, P152A-Stp1–Corilagin, and F179A-Stp1–Corilagin Complexes in the Allosteric Binding Conformations

energy components (kcal/mol)	Stp1–corilagin	P152A–corilagin	F179A–corilagin
ΔEele	–17.02	–8.05	–8.88
ΔEvdw	–49.55	–40.32	–42.36
ΔEMM	–66.57	–48.37	–51.24
ΔGele,sol	36.18	27.72	29.08
ΔGnonpolar,sol	–6.06	–5.29	–5.39
ΔGsol	30.13	22.44	23.69
ΔGele,sol + ΔEele	19.16	19.68	20.20
ΔGnonpolar,sol + ΔEvdw	–55.60	–45.60	–47.75
ΔGtotal	–36.44	–25.93	–27.55
–TΔS	3.00	2.97	3.29
ΔGbind	–33.44	–22.96	–24.25
KA (1 × 105) L/mol	5.15	1.38	3.91

Free Binding Energy Analysis

The ΔGbind value shows that the competitive and allosteric interactions between corilagin and Stp1 are favorable, which are −21.70 and −33.44 kcal/mol, respectively (Table ). The free binding energy decomposition shows that the contributions of van der Waals interaction (ΔEvdw), electrostatic (ΔEele), and nonpolar solvation (ΔGNP) energies are beneficial for the interaction of the complexes. The contribution of nonpolar interaction (ΔGNP) from the solvent-accessible surface area in competitive binding conformation (ΔGNP = −5.62 kcal/mol) and allosteric binding conformation (ΔGNP = −6.06 kcal/mol) was small, but it was conducive for the formation of the complex.

Table 6

Calculated Energy Components of Two Stp1–Corilagin Complexes Binding Conformation Based on MM-GBSA

energy components (kcal/mol)	competitive binding conformation	allosteric binding conformation
ΔEele	–29.62	–17.02
ΔEvdw	–37.87	–49.55
ΔEMM	–67.49	–66.57
ΔGele,sol	47.68	36.18
ΔGnonpolar,sol	–5.62	–6.06
ΔGsol	42.06	30.13
ΔGele,sol + ΔEele	18.07	19.16
ΔGnonpolar,sol + ΔEvdw	–43.49	–55.60
ΔGtotal	–25.43	–36.44
–TΔS	3.73	3.00
ΔGbind	–21.70	–33.44

In the results of our analysis, the van der Waals force contribution of competitive binding (ΔEvdw = −37.87 kcal/mol) was much smaller than that of allosteric inhibition (ΔEvdw = −49.55 kcal/mol). Consistent with the previous results, the amino acid residues present at the binding site of the allosteric binding mode, especially Asn142, Val145, Leu146, Pro152, and Phe179 exhibited strong hydrophobic interaction. Interestingly, although the free binding energy of the allosteric sites was stronger, the contribution of the electrostatic forces was found to be reduced (Table ). This can be mainly attributed to the polar effects of residues in competitive binding conformation.

Conclusions

Corilagin is a naturally active substance with low toxicity, which is derived from longan and leaf beads.[22] The present study was focused mainly on its antitumor activity. In this work, a virtual screening method was used. Corilagin was found to exhibit good inhibition potential among different natural compounds, which was verified through experiments. Further enzymatic reaction kinetic experiments found that corilagin was a mixed inhibitor of Stp1.

To elucidate the inhibition mechanism at a molecular level, two methods of docking and simulation were conducted with the assistance of molecular simulation. According to the experimental results, corilagin was found to interact with Stp1 in two regions. One was found to be located at the active center of Mn2+. In this active center, the enzyme activity was reduced due to the competition with the substrate for binding to the protein. The residues Asn162, Ile164, Tyr199, and Lys232 were found to play important role in this binding of Stp1 with corilagin. The two amino acids with the strongest total binding energy identified were Asn162 and Tyr199. Concurrently, they formed stable hydrogen bonds with the two carbonyl groups, O8 and O12, in the 12-membered ring of corilagin, respectively. Therefore, the new inhibitor design can replace the two carbonyl groups of the 12-membered ring with the strong hydrophilic group to facilitate the formation of the hydrogen bond. Another type of binding mode identified was the binding corilagin to the inactive flap region of Stp1. Asn142, Val145, Leu146, Pro152, and Phe179 were the important amino acid residues identified in this complex. Binding of corilagin to the flap region affected the binding of the substrate and Stp1, thereby leading to a reduction in the Stp1 activity. Concurrently, these residues of Stp1 were found to mainly interact with corilagin through hydrophobic interaction. The hydroxyl group on the six-membered ring of corilagin is an important binding functional group in hydrophobic interaction. Therefore, the binding of the allosteric inhibitor can be enhanced by replacing the hydrophobic functional group on the six-membered ring. This further showed that Stp1 mainly interacted with corilagin through hydrogen bonding in the active center. Also, the key groups responsible for the interaction in the active center were the two carbonyl groups of the 12-membered heterocyclic ring of corilagin. In the inactive center, Stp1 mainly interacted through hydrophobic interaction.

Therefore, corilagin was identified and verified as a potential inhibitor of Stp1 with dual inhibitory action. The experiments conducted in this study combined with high-throughput screening provided an effective screening method for subsequent discovery of inhibitors.

Materials and Methods

Bacterial Strains and Materials

S. aureus strain USA300 (ATCC BAA-1717) was acquired from the College of Food Science and Engineering Laboratory, Jilin University. Escherichia coli DH5α and BL21 (DE3) were purchased from Tiangen Biotech (Beijing) Co., Ltd. Corilagin was purchased from Chengdu Herbpurify Co., Ltd. PNPP, isopropyl β-d-thiogalactoside (IPTG), and kanamycin were purchased from Dalian Meilun Biotechnology Co., Ltd. Taq DNA polymerase and DpnI restriction endonuclease were purchased from TransGen Biotech Co., Ltd. Other drugs used in this study were purchased from Sigma-Aldrich.

Plasmid Construction, Protein Expression, and Purification

Polymerase chain reaction (PCR) was performed using S. aureus USA300 genome as a template. After PCR, agarose gel electrophoresis was conducted for inspection and gel recovery of the PCR-amplified gene product by the TIANgel Midi Purification Kit. The amplified gene and PET28a vector were digested with BamHI and XhoI restriction enzymes and ligated with T4 ligase to prepare the final construct named PET28a-Stp1. The construction of the mutant plasmid was performed using the QuikChange site-directed mutagenesis kit. After incorporating the point mutation, it was digested with DpnI restriction endonuclease and transformed into E. coli DH5α competent cells for sequencing. The primers used in the experiments are listed in Table .

Table 7

Oligonucleotide Primers Used in This Study

primer name	oligonucleotide (5′–3′)
Stp1-F	GAAGGATCCATGCTAGAGGCACAATTTTTTAC
Stp1-R	TCTCTCGAGTCATACTTTATCACCTTCAATAG
P152A-F	GGGTCAAATTACGGCAGAAGAAGCATTTAC
P152A-R	GTAAATGCTTCTTCTGCCGTAATTTGACCC
I164A-F	CAACGTAATATTGCAACGAAGGTGATG
I164A-R	CATCACCTTCGTTGCAATATTACGTTG
F179A-F	AGTCCAGATTTGGCTATTAAGCGAT
F179A-R	ATCGCTTAATAGCCAAATCTGGACT
Y199A-F	GGATTAACTGATGCGGTTAAAGACAATG
Y199A-R	CATTGTCTTTAACCGCATCAGTTAATCC

The plasmid with a positive sequence was then introduced into BL21 (DE3) for expression. The bacteria were cultured in the Luria–Bertani medium at 37 °C to obtain OD600 = 0.5. IPTG was added to the medium and cultured at 16 °C overnight. Bacteria were then harvested and sonicated. The supernatant was purified through a Ni-NTA agarose column and concentrated.[23]

Phosphatase Assay

Based on the previously published method,[12] different concentrations of inhibitors were mixed with 0.1 μM Stp1 protein buffer (10 mM Tris–HCl, 5 mM MnCl2, 1 mM ethylenediaminetetraacetic acid (EDTA), 0.02% β-ME) and incubated for 10 min. Then, 0.75 mM pNPP was added and the suspension was incubated for 15 min at room temperature. Subsequently, 20 μL of 5 M NaOH was added to stop the reaction, and the absorbance was measured at 405 nm of wavelength. Positive and negative controls were used. The positive control was without a drug and the negative control was without the protein. This experiment was repeated thrice and the data was processed using GraphPad Prism software.

Types of Enzyme Inhibitors

Phosphatase assays were performed using different concentrations of corilagin and pNPP. The initial reaction rate was determined by assessing the changes in absorbance within the first 1 h. Using the inverse of the initial reaction rate [V] as the ordinate and the inverse of the substrate concentration [S] as the abscissa, a Lineweaver–Burk diagram was generated.

Virtual Screening

Virtual screening was performed using AutoDock Vina software.[24] The docking results were then sorted further to obtain the experimental results. Among these, 143 758 natural compounds were identified, which are available in the ZINC database.[25,26] The processes and methods are described in detail in the Supporting Information.

Molecular Dynamics Simulation

For this experiment, AutoDock Vina software was used to perform the docking of corilagin with Stp1. The docking results were used as the initial files for performing molecular dynamics (MD) simulations. The antechamber program[27] was used to estimate the parameters of corilagin and the HF/6-31G* RESP from the Amber suite[28] was used to calculate the charge (Supporting Information). Among them, the corilagin was optimized through Gaussian 09 program. The molecular dynamics simulations were performed using Gromacs 4.5.5 software package.[29] The Amberff99sb force field[30] and TIP3P water model[31] were applied in combination. Other parameters were also assessed on the basis of the previous literature.[32] The Stp1–corilagin system first relaxed the energy by 2000 steps of the steepest-descent energy minimization, and then conjugate-gradient energy minimization was performed. The complex was equilibrated by 500 ps molecular dynamics operation restricted by position on both the protein and the ligand. After the first equilibrium run, the 200 ns MD was performed, and there is no restriction on the solute position. The last 50 ns trajectory was used for subsequent analysis to minimize convergence artifacts. To check the equilibration of the trajectory, the equilibration of quantities such as the RMSD was monitored. The particle mesh Ewald algorithm and LINCS algorithm were used in MD simulation. The processes and methods are described in detail in the Supporting Information.

After 200 ns simulation, the binding energy between the protein and corilagin was analyzed using the MM-GBSA method.[33−35] The MM-GBSA method can be summarized as followswhere ΔEMM is composed of van der Waals (ΔEvdw), the electrostatic (ΔEele), and internal energy (ΔEinternal).Moreover, ΔGsol can be divided into the polar (ΔGGB) and nonpolar contributions (ΔGNP), as shown in 3At the same time, the free energy of the per-residue can be calculated using the MM-GBSA method.The entropy term TΔS is needed to explain the conformation entropy change of the two binding partners in complexes. In this study, the nmode module of Amber 10 software package was used to calculate the entropy change of the complex. Other parameters referred to previous literature.[36]

Fluorescence Quenching

The binding constant (KA) of the ligand to the wild-type (WT) and mutant Stp1 binding sites was calculated through fluorescence quenching, and the calculation method used was based on the following equation[37]