Identification of Streptococcus pneumoniae Sortase A Inhibitors and the Interactive Mechanism

Abstract

Streptococcus pneumoniae (S. pneumoniae) Sortase A (SrtA) anchors virulence proteins to the surface of the cell wall by recognizing and cleaving the LPXTG motif. These toxins help bacteria adhere to and colonize host cells, promote biofilm formation, and trigger host inflammatory responses. Therefore, SrtA is an ideal target for the development of new preparations for S. pneumoniae. In this study, we found that phloretin (pht) and phlorizin (phz) exhibited excellent affinities for SrtA based on virtual screening experiments. We analyzed the interactive mechanism between pht, phz, and alnusone (aln, a reported S. pneumoniae SrtA inhibitor) and SrtA based on molecular dynamics simulation experiments. The results showed that these inhibitors bound to the active pocket of SrtA, and the root mean square deviation (RMSD) and distance analyses showed that these compounds and SrtA maintained stable configuration and binding during the assay. The binding free energy analysis showed that both electrostatic forces (ele), van der Waals forces (vdw), and hydrogen bonds (Hbonds) promoted the binding between pht, phz, and SrtA; however, for the binding of aln and SrtA, the vdw force was much stronger than ele, and Hbonds were not found. The binding free energy decomposition showed that HIS141, ILE143, and PHE119 contributed more energy to promote pht and SrtA binding; ARG215, ASP188, and LEU210 contributed more energy to promote phz and SrtA binding; and HIS141, ASP209, and ARG215 contributed more energy to promote aln and SrtA binding. Finally, the transpeptidase activity of SrtA decreased significantly when treated with different concentrations of pht, phz, or aln, which inhibited S. pneumoniae biofilm formation and adhesion to A549 cells without affecting normal bacterial growth. These results suggest that pht, phtz, and aln are potential materials for the development of novel inhibitors against S. pneumoniae infection.

1. Background

Streptococcus pneumoniae (S. pneumoniae) is one of the main pathogens causing community-acquired pneumonia. Other diseases caused by this pathogen include but are not limited to otitis media, septicemia, and meningitis [1]. Children and individuals with poor immune systems have a higher probability of infection by this pathogen [2]. In developing countries, nearly a million children die from pneumococcal diseases annually [3]. S. pneumoniae often asymptomatically colonizes the upper respiratory tract of hosts, and over half of the affected children have pneumococcal colonization in the nasopharynx [4]. The development of antibiotic resistance in S. pneumoniae has posed a challenge for humans in combating its infections, which has also resulted in significant social medical expenses [5,6]. Therefore, exploring new strategies and developing new drugs to combat S. pneumoniae infections is urgent.

Sortase A (SrtA) is a cysteine protease found in S. pneumoniae. It can recognize the LPXTG motif of surface proteins in bacteria, cleave the peptide bond between threonine (T) and glycine (G) in the motif, and then covalently anchor the surface protein to the cell wall peptidoglycan [7,8,9,10]. S. pneumoniae expresses more than ten surface proteins containing LPXTG, most of which have enzymatic or adhesive functions related to pathogenic processes, such as adhesion, colonization, and invasion [8,9,11,12]. In addition to enhancing bacterial pathogenicity, SrtA also affects the retention and immunomodulation of probiotics in the host intestinal tract and is involved in the assembly of bacterial pili [13]. Benedetta Maggio and colleagues identified some compounds that reduced S. pneumoniae biofilm formation by interacting with SrtA [14]. Therefore, SrtA influences the virulence of S. pneumoniae, making it an ideal target for developing drugs against infections caused by this bacterium.

Phloretin (pht) and phlorizin (phz) are mainly found in apples and have been demonstrated to possess various pharmacological functions, such as antioxidant, anti-tumor, anti-inflammation, and diabetes-controlling activities [15,16,17]. However, they have not been reported to suppress S. pneumoniae growth by interacting with SrtA. Alnustone (Aln) mainly exists in the rhizomes of turmeric and ginger, and it has been demonstrated to have anti-inflammatory, anti-cancer, and anti-hepatitis virus effects [18,19]. A study reported that aln reduced the ability of purified SrtA to split its substrate based on a fluorescence polarization method, suggesting that aln directly interacts with SrtA [19]; however, the exact interactive mechanism was unclear. This work explores the interactive mechanisms of pht, pht, aln, and SrtA. Electrostatic forces (ele), van der Waals forces (vdw), and hydrogen bonds (H-bonds) were generated during their binding to SrtA. These compounds interact with the critical active site residues (HIS141, CYS207, and ARG215) of SrtA to inhibit its biological function. These results were confirmed by the inhibition of pht, phz, and aln on SrtA transpeptide activities. In addition, pht, phz, and aln decreased the adhesion of S. pneumoniae to human lung epithelial cells without affecting the growth of bacteria.

2. Methodology

2.1. Strain, Compounds and Reagents

The S. pneumoniae (NCTC 7466) strain was stored in our laboratory and cultured in Todd-Hewitt broth supplemented with 1% yeast extract (THY) at 37 °C. Pht, phz, and aln were purchased from Chengdu Herpurify Co., Ltd. (Chengdu, China).

2.2. Structure-Based Virtual Screening and Molecular Docking

The structure of S. pneumoniae SrtA was obtained from the Protein Data Bank website with PDB ID 4O8L. This crystal structure was obtained from an X-ray assay with a separation rate of 2.70 angstroms (Å). The residue sequence of the protein ranged from 82 to 247, including the binding sites of the SrtA receptor, and there were no residue mutations. Therefore, this crystal structure model has a complete structure and higher accuracy, which meets the standards for virtual screening. The original SDF file used for virtual screening was obtained from Selleck (L6800), which contained 1219 compounds. Based on our research objectives, a sub-library containing around 800 compounds was generated from this library and used for screening. SrtA was set as the receptor, and the small-molecule compound was set as a ligand. The pdbqt files of the compounds were stored in our laboratory, and the pdbqt file of SrtA was obtained using AutoDock Tools. For screening, a docking box (40 × 40 × 40) that contained the active center of SrtA was generated with a spacing of 1 Å; other parameters were as follows: exhaustiveness = 16, num_modes = 9, energy_range = 4; compounds with an affinity greater than −7.0 kcal/mol were chosen for the next assay. AutoDock Vina was used to perform the screening task [20,21,22,23]. Based on the scores obtained from the screening between the compound and SrtA, pht and phz were chosen. Docking calculations between aln and SrtA were performed using Autodock Vina.

2.3. Molecular Simulation

Based on the complex structures of pht-SrtA, phz-SrtA, and aln-SrtA obtained from screening or docking calculations, GROMACS 2020.6 was used to perform molecular dynamics simulations, and the Amber14sb force field and TIP3P water model were used as the solvent model [24,25,26]. A cubic box was used, and after adding counter ions, the complex system was initially energy-relaxed with 2000 steps of the steepest-descent energy minimization and 2000 subsequent steps of conjugate-gradient energy minimization. The structure was equilibrated at a constant temperature and pressure. Subsequently, a 100 ns production simulation was conducted.

2.4. Parameter Analysis

The root mean square deviation (RMSD) and distance between pht, phz, aln, and SrtA against time were analyzed to evaluate the equilibrium state of the simulation and the stability of the binding. The distances between pht, phz, aln, and each residue in SrtA were analyzed using GROMACS.

2.5. Binding Free Energy Analysis

The molecular mechanics Poisson–Boltzmann surface area (MMPBSA) method was used to analyze the binding free energies between pht, phz, aln, and SrtA. The equilibrated trajectory from 70 to 100 ns was used to calculate the binding free energy by referring to the method described previously [27,28,29], and the formula used here was as follows: ΔG_bind = ΔE_vdw + ΔE_ele + ΔG_sol − TΔS.

2.6. Hydrogen Bond Analysis

Based on GROMACS and LigPlot+ [30], an analysis was conducted on the H-bond count and lifetime during the binding process of pht, phz, and aln with SrtA to evaluate the contributions of H-bonds to their binding.

2.7. Visualization of Weak Interaction Analysis

The visualization of average non-covalent interactions (aNCI) can intuitively show the interactions between molecules, such as H-bonds and vdw. An average reduced density gradient (aRDG) method has been demonstrated could be used to show the aNCI of two molecules in a dynamic trajectory by calculating the averaged density gradient ($\overline{\nabla \mathrm{\rho }}$) and averaged density ($\overline{\mathrm{\rho }}$). Here, the aNCI between pht, phz, aln, and SrtA was analyzed based on this method [31]. Briefly, the original results used for the aNCI analysis were obtained via molecular simulations. All the atoms of the compounds were fixed, and another 1 ns molecular simulation was performed using the same parameters. After the assays were completed, a trajectory containing 1001 frames was collected. The aRDG was calculated using the following equation and Multiwfn software [32,33], and visualization analysis was performed using VMD software of 1.9.3 version [34].

$$aRDG\left(r\right)={\displaystyle \frac{1}{{2\left(3{\pi }^2\right)}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$3$}\right.}}}{\displaystyle \frac{\left|\overline{\nabla \rho }\left(r\right)\right|}{{\left[\overline{\rho }\left(r\right)\right]}^{\raisebox{1ex}{$4$}\!\left/ \!\raisebox{-1ex}{$3$}\right.}}}$$

(1)

2.8. SrtA Activity Inhibition Assay

SrtA protein (8 µg) stored in our laboratory was co-incubated with various concentrations of pht (0, 8, 16 µg/mL), phz (0, 8, 16 µg/mL), and aln (0, 16, 32 µg/mL) at 37 °C for 30 min, and then Dabcyl-QALPETGEE-Edans (GL Biochem, Shanghai), the substrate peptide of SrtA, was added to the samples (5 µg/mL) and co-incubated for 60 min in the dark. The fluorescence values of each well were obtained using a microplate reader (excitation wavelength 350 nm, emission wavelength 520 nm), and the inhibitory effects of pht, phz, and aln against SrtA activity were evaluated based on the fluorescence values [35].

2.9. Growth Curve Assay

S. pneumoniae strains in the initial stage of logarithmic growth were co-cultured with different concentrations of pht (0, 16, 32 µg/mL), phz (0, 16, 32 µg/mL), and aln (0, 16, 32 µg/mL). The optical density (OD) at 600 nm was measured every two hours until the bacteria reached the plateau phase to evaluate whether these compounds affected the growth of the bacteria.

2.10. Adhesion Inhibition Assay

Human lung epithelial cells A549 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) overnight. Clean DMEM medium containing S. pneumoniae and different concentrations of pht (0, 16, 32 µg/mL), phz (0, 16, 32 µg/mL), and aln (0, 16, 32 µg/mL) were used to replace the medium and culture for two hours. The medium was then removed, and the cells were washed three times with sterile phosphate buffer. Samples were harvested and coated onto an agar culture medium to obtain clones, and the inhibitory effects of the compounds against S. pneumoniae’s adhesion to cells were measured based on the clones.

2.11. Bacterial Biofilm Assay

S. pneumoniae with various concentrations (0, 16, 32 µg/mL) of pht, phz, or aln in 96 well cell plates (5 × 10⁶ colony-forming units per well) were co-cultured at 37 °C for 24 h. The culture medium was then removed, and the samples were washed three times with a sterile phosphate buffer. Samples were stained with 0.1% crystal violet solution for 20 min after drying, and 33% acetic acid was used to treat each well after discarding the staining solution and washing. Samples were used to measure the OD570 values to evaluate biofilm formation.

2.12. Statistical Analysis

An unpaired t-test was conducted for statistical analysis to determine the significance of the data using Prism software (version 9.5.0). The results are presented as mean ± standard deviation (SD). Significant differences were defined as p < 0.05.

3. Results

3.1. Analysis of the Results

3.1.1. pht, phz, and aln Bind to the Active Center of SrtA

The molecular structures of phz, pht, and aln are shown in Figure 1a–c. Based on the results of virtual screening and docking calculations, it was found that phz, pht, and aln were located at the active center of SrtA (red line was Chain A, yellow line was Chain B) with affinities of −8.63 kcal/mol, −8.53 kcal/mol, and −6.53 kcal/mol, respectively (Figure 1d–f). Binding site analysis showed that residues in SrtA chain A and chain B both interact with phz, pht, and aln,; these residues mainly include but are not limited to HIS141, HIS142, THE143, PHE144, ASP188, GLU208, ASP209, LEU210, ARG215, and THR213 (Figure 1g–i). HIS141, CYS207, and ARG215 have been identified as the critical active sites of S. pneumoniae SrtA, the interactive sites between phz, pht, aln, and SrtA involved in HIS141 and/or ARG215, suggesting phz, pht, aln maybe inhibit the virulence of S. pneumoniae by targeting SrtA.

3.1.2. pht, phz, aln, and SrtA Maintained Stable Binding During the Simulation

The RMSD values of free SrtA fluctuated around 0.13 nm when the simulation time reached 40 ns (Figure 2a). When bound with pht or aln, the RMSD values of SrtA fluctuated around 0.15 nm or 0.12 nm during the entire simulation (Figure 2b,c); when the simulation time reached 60 ns, the RMSD values of pht and aln stabilized at 0.17 nm or 0.20 nm (Figure 2b,c). When bound to phz, the RMSD values of SrtA fluctuated around 0.20 nm when the simulation time reached 50 ns, while the RMSD of phz fluctuated around 0.25 nm when the simulation time reached 30 ns (Figure 2d). The distances between pht, aln, phz, and SrtA against the simulation time suggest that these compounds maintained stable binding during the simulation (Figure 2e), which was confirmed by the structural overlap of the trajectories at different moments (Figure 2f–h).

3.1.3. Confirmation of the Interactive Residues

The total binding free energy between phz, pht, aln, and SrtA was −39.17 kJ/mol, −59.49 kJ/mol, and −97.72 kJ/mol, respectively, including electrostatic action (ele) −86.65 kJ/mol, −68.24 kJ/mol, and −8.23 kJ/mol; van der Waals interactions (vdw) −157.73 kJ/mol, −164.75 kJ/mol, and −188.87 kJ/mol; and solvation interaction 171.91 kJ/mol, 154.25 kJ/mol, and 80.8 kJ/mol (Figure 3a). Large areas of green or blue isosurfaces between pht (Figure 3b), phz (Figure 3c), aln (Figure 3d), and the adjacent residues also demonstrate the existence of vdw and H-bonds interactions.

To confirm the exact interactive residues, residue energy decomposition was conducted. For the pht-SrtA complex system, GLU208, ASP209, and HIS142 provided more ele energy with values of −11.41 kJ/mol, −23.83 kJ/mol, −7.31 kJ/mol (Figure 4a), HIS141, ILE143, and PHE119 contributed more vdw energy with values of − 12.13 kJ/mol, − 14.37 kJ/mol, − 13.80 kJ/mol (Figure 4b), the solvation free energy (sol) that composed of polar or nonpolar interaction is shown in Figure 4c,d, and the LEU118, PHE119, and ILE143 contributed total binding free energy (Figure 4e). For the phz-SrtA complex system, ASP188, ASP191, and ARG215 provided more ele energy, with values of −11.67 kJ/mol, −24.71 kJ/mol, and − 10.63 kJ/mol, respectively (Figure 4f). VAL190, ASP188, and LEU210 contributed more vdw energy with values of −8.80 kJ/mol, −9.10 kJ/mol, and −9.40 kJ/mol (Figure 4g). The sol energies are shown in Figure 4h,i. ASP209, LEU210, and GLU192 contributed more to the total binding free energy (Figure 4j). When SrtA was bound to aln, HIS141, GLU208, and ARG215 contributed more ele energy (Figure 4k), and HIS141, ASP209, and PHE119 provided more vdw energy (Figure 4l). The sol energy is shown in Figure 4m,n, and PHE119, GLU208, and TYR247 contributed more to the total binding free energy (Figure 4o). These results indicate that although these three compounds bind to the active center of SrtA, the interacting residues involved are not identical.

For further confirmation, the distance between each residue in SrtA and phz, pht, and aln was analyzed, and it was found that the residues that contributed more energy between the binding showed lower distance to phz (Figure 5a), pht (Figure 5b), and aln (Figure 5c).

3.1.4. Hbonds Are Generated Between SrtA and pht, phz

H-bonds are important weak interactions that promote the binding of compounds to proteins. Here, we analyzed whether H-bonds were generated in the interactions between pht, phz, aln, and SrtA. It was found that there are more than twenty-five pairs of H-bonds were detected between SrtA and phz (Figure 6a), but these Hbonds have a lower presence except one pair with an occupancy of 94.5% (Figure 6b), an oxygen atom of ASP191 in SrtA Chain B was the acceptor, an oxygen atom of number 10 in phz was the donor (Figure 6c). For the pht-SrtA system, about ten pairs of H-bonds were formed (Figure 6d), three of which had higher occupancies with values of 44.4%, 50.4%, and 61.8%, respectively (Figure 6e). PHE144, HIS142, and ASP209 in Chain B of SrtA were involved in these Hbonds (Figure 6f). No Hbonds were measured between aln and SrtA.

3.1.5. pht, phz, aln Inhibit the Activities of SrtA

To verify the reliability of the docking and simulation results, SrtA activity inhibition experiments were performed. The activity of SrtA was defined as 100% when no pht, phz, or aln were present in the reaction system. The activities of SrtA decreased to 58.53%, and 44.30% (Figure 7a), 73.74%, and 54.57% (Figure 7b), and 63.45% and 49.54% (Figure 7c) when treated with different concentrations of pht, phz, and aln. These results indicate that pht, phz, and aln can significantly inhibit the activity of SrtA and confirm the reliability of the docking and simulation results.

3.1.6. pht, phz and aln Does Not Affect the Growth of S. pneumoniae

A growth curve assay was performed to evaluate whether pht, phz, or aln affected the growth of S. pneumoniae. The bacteria showed a normal growth trend when the system did not contain pht, phz, or aln; however, the bacteria showed similar growth trends when treated with different concentrations of pht (Figure 8a), phz (Figure 8b), or aln (Figure 8c) treatment, suggesting that these compounds do not affect the growth of S. pneumoniae.

3.1.7. pht, phz and aln Inhibits the Adhesion of S. pneumoniae to A549 Cells

Clones measured from A549 cells that received S. pneumoniae treatment but without pht, phz, or aln were defined as 100%, while the adhesion of S. pneumoniae to A549 cells decreased to 58.70% and 42.80% (Figure 9a), 64.27% and 56.66% (Figure 9b), 71.47% and 59.65% (Figure 9c) when A549 cells received different concentrations of pht, phz, or aln treatments, indicating that pht, phz, or aln could significantly reduce the adhesion of S. pneumoniae to human lung epithelial cells.

3.1.8. pht, phz and aln Inhibits S. pneumoniae Biofilm Formation

As shown in

Table 1. The inhibitory effects of pht, phz, and aln on S. pneumoniae biofilms.

Concentrations (µg/mL)	pht	phz	aln
0	99.98 ± 4.25	100.00 ± 2.59	100.02 ± 3.34
16	50.95 ± 1.91 **	55.60 ± 2.76 **	68.46 ± 3.85 **
32	37.70 ± 3.95 **	47.14 ± 2.29 **	59.99 ± 2.89 **

Notes: ** represents p ≤ 0.01.

, the S. pneumoniae biofilm obtained from samples that did not receive compound treatment was set as the control, but the formation of biofilm was reduced to 50.95% and 37.70%, 55.60% and 47.14%, 68.46% and 59.99% when the bacteria were treated with different concentrations of pht, phz, or aln, suggesting that these compounds significantly inhibited the formation of S. pneumoniae biofilms.

4. Discussion

Given the important role of SrtA in the pathogenicity of S. pneumoniae, there has been increasing interest in its research. To date, three crystal structures of SrtA have been reported [36,37], and CYS207, HIS141, and ARG215 residues were identified as the critical active sites, as SrtA almost completely loses its activity when these residues are mutated [36]. In this study, pht and phz were identified as potential inhibitors of SrtA using 4O8L as the initial structure. Then, pht, phz, and aln, which have been identified as inhibitors of SrtA, were used to analyze the interaction mechanism with SrtA separately. For molecular simulation assays, simulation convergence or system equilibrium is the basis for subsequent analysis, and RMSD is usually used to evaluate the convergence or system equilibrium of the protein-ligand system. Herein, the three complex systems reached equilibrium based on the RMSD parameter. Further, although pht, phz, and aln bound to the active center of SrtA, the interaction energies and residues involved in the interaction were not exactly the same; ele, vdw, and Hbonds interactions promoted the binding between pht, phz, and SrtA, but the main interaction between aln and SrtA was vdw, the ele interaction was weakened, and Hbonds were not found. These differences may be attributed to the hydroxyl groups in pht and phz. Residues in SrtA that interacted with pht, phz, or aln included but were not limited to HIS141, CYS207, and ARG215, which are the active site residues.

SrtA can recognize and specifically cleave the LPXTG sequence and anchor surface proteins to the bacterial cell wall [38,39,40]. To verify the reliability of the docking and simulation results, we synthesized substrate peptide sequences carrying fluorescent groups at both ends, which could be used to quantitatively detect the activity of SrtA by detecting the fluorescence values. It was found that pht, phz, or aln significantly reduced the activity of SrtA after co-incubation, which confirmed the reliability of the docking and simulation results.

Selective growth stress is a key factor in the generation and development of bacterial resistance [41,42]. Anti-virulence strategies aim to weaken bacterial pathogenicity by inhibiting or neutralizing key virulence factors in bacterial infections [43,44,45]. A key prerequisite of the anti-virulence strategy is that inhibitors do not exert selective growth pressure on bacteria and do not trigger the development of bacterial resistance upon application. Herein, we confirmed that pht, phz, and aln did not affect the normal growth of S. pneumoniae at the tested concentrations, indicating that these three compounds will not cause bacterial resistance when applied to the prevention and control of S. pneumoniae infection in the future.

The surface proteins that are anchored to the cell wall by SrtA are involved in the biofilm formation, adhesion, and colonization of S. pneumoniae to host cells, which is important for promoting the bacteria to achieve a successful infection [7]. In this study, pht, phz, or aln treatment significantly inhibited the formation of biofilms, adhesion, and colonization of S. pneumoniae to human lung epithelial cells, which may result from the inhibition of SrtA activity, leading to a reduction in the surface protein that anchored to the cell wall, thus reducing the adhesion and colonization of bacteria to host cells.

Computer-aided drug screening is important for exploring and developing new drugs. However, due to the complexity of the actual biological environment system, off-target effects are universal. Some factors can trigger off-target effects, such as insufficient target selectivity, binding of drugs to other non-target molecules, diversity and dynamics of targets, and affinity selectivity of drugs. Therefore, it is important to verify the reliability of these compounds in actual biological environments, followed by computer-aided screening. Herein, we confirmed the effectiveness of the target compounds in a practical biological environment based on SrtA activity inhibition, growth curve, adhesion, and biofilm assays, providing support for the application of the compounds.

In this study, pht and phz were identified as the inhibitors of SrtA, aln is an inhibitor of SrtA that was used as the reference group. Pht and phz share similar structures; there are four phenolic hydroxyl groups in pht, and in phz, the hydrogen atom of the ortho phenolic hydroxyl group of pht is replaced by a pyran ring, and each of the four chiral carbon atoms has an alcohol hydroxyl group. At 16 µg/mL, pht exhibited the most potent inhibitory effect on SrtA activity, suggesting that the pyran ring contributes to this inhibitory effect. Although there are three alcoholic hydroxyl groups in the pyran ring, the phenolic hydroxyl group is more active than the alcoholic hydroxyl group and is easier to ionize out hydrogen ions, which may contribute to the difference in the inhibition of SrtA activity. Therefore, Pht may possess a higher drug-making potential.

5. Conclusions

Pht, phz, and aln bound to the active center of S. pneumoniae SrtA, ele interaction, vdw interaction, and Hbonds promoted the binding of pht, phz, and SrtA; the interaction between aln and SrtA was mainly vdw energy. Pht, pht, and aln interact with HIS141 and ARG215 in SrtA, which are critical active site residues. As a result, SrtA lost its bioactivity significantly when treated with pht, phz, or aln. Pht, phz, and aln did not show anti-S. pneumoniae characteristics but significantly alleviated the formation of S. pneumoniae biofilms and the adhesion of S. pneumoniae to human lung epithelial cells. These results suggest that these compounds would not exert growth pressure on S. pneumoniae and would not cause the development of drug resistance when used to combat infections from this pathogen, suggesting that they have great potential for application in the prevention and control of S. pneumoniae infection.