Trends Sci. 2025; 22(10): 10420

Oral health plays a crucial role in maintaining general health [1]. Preventing dental problems including cavities and periodontal diseases, as well as their wider effects on systemic health, requires maintaining good oral health [2,3]. Research has demonstrated that poor oral hygiene increases the risk of serious conditions, including cardiovascular disease, diabetes, and respiratory infections [4-6]. Among these, certain bacteria and fungi play significant roles in either maintaining or compromising oral health. Streptococcus mutans is often associated with its capacity to create acids from the fermentation of carbohydrates, which results in tooth damage and dental caries [7]. By altering the composition of biofilms and interacting with other oral pathogens, S. sanguinis can affect the oral environment [8]. Furthermore, a fungus called Candida albicans can cause oral infections such as thrush, particularly in people on antibiotics or those with weakened immune systems [9-11].

The management of oral diseases often involves the use of antibiotics and antifungal agents to combat bacterial and fungal infections [12,13]. However, the increasing prevalence of antibiotic and antifungal resistance poses a significant challenge in this context [14]. Commonly used antibiotics, such as penicillins and cephalosporins, are resistant to bacteria such as Streptococcus and Staphylococcus, while macrolides and fluoroquinolones are also resistant in certain strains [15,16]. Similarly, antifungal agents like azoles are becoming less effective against fungi such as C. albicans, which can develop resistance through mechanisms involving biofilms and genetic mutations [17,18]. As a result, strategies to combat resistance, including the development of new antimicrobial agents, are urgently needed [19-22]. This study aims to explore this challenge by searching for compounds abundant in nature that can be potential antimicrobials with multitarget bioactivity against S. mutans, S. sanguinis, and C. albicans.

Among natural compounds, -sitosterol, a phytosterol commonly found in plant-based foods, has garnered attention for its potential health benefits, including lowering cholesterol levels and managing benign prostatic hyperplasia [23,24]. Recent studies have also investigated its antimicrobial properties, revealing moderate antibacterial activity against certain pathogens [24,25]. While its effectiveness as a standalone antimicrobial agent is limited compared with that of conventional antibiotics, -sitosterol shows promise when used in synergistic combinations with existing drugs [26]. The antimicrobial potential of -sitosterol against C. albicans, S. mutans, and S. sanguinis remains underexplored in current research.

The formation of biofilms by both bacteria and fungi further complicates treatment by protecting these microorganisms from the effects of antibiotics and antifungals [27]. An essential stage in the development of dental caries is the creation and maintenance of biofilms on the tooth surface, which are facilitated by the GbpC enzyme from S. mutans [28]. Sap5 facilitates adhesion and proteolysis during C. albicans biofilm formation, whereas SrtC contributes to protein anchoring and matrix synthesis in S. sanguinis [29,30].

Materials and methods

General experimental procedures

NMR spectra were captured using a Bruker 700 MHz device. Additionally, HR-ESI-MS spectra were acquired by attaching a quadrupole time of flight mass spectrometer (Xevo G2-XS QTOF, Waters Corp.) to a Waters ACQUITY UPLC system (Waters Corp., Milford, MA). RP-18 gel (63 - 212 μm, Fujifilm Wako, Osaka, Japan) and silica gel (0.063 - 0.2 mm, Merck, Darmstadt, Germany) were used for column chromatography (CC). The RP-18 F254S (0.25 mm, Merck) and 60 F254 (0.25 mm, Merck) precoated silica gel plates were utilized for thin layer chromatography (TLC). Sample spots were visualized by heating and spraying with 10% aqueous H₂SO₄.

Plant materials

Fresh leaves of Piper crocatum (Ruiz and Pav.) were gathered in September 2022 from Cikarang Barat, Bekasi, Jawa Barat, Indonesia. The plant material was identified in the Biosystematics and Molecular Laboratory at the Department of Biology, Faculty of Mathematics and Natural Sciences, Universitas Padjadjaran, Sumedang, Indonesia with number 93/LBM/IT/9/2022.

Extraction and isolation

The dried and cut leaves of P. crocatum (15 kg) were extracted with 150 L of methanol through maceration at room temperature (25 °C) for 3 days. The collected methanol filtrate was evaporated under reduced pressure (Buchi Rotavapor R-100, Switzerland), to yield a solid extract (25 g). The concentrated P. crocatum methanol extract was then separated by column chromatography using Silica G 60 stationary phase (0.063 - 0.200 mm) with solvent n-hexane and ethyl acetate with an increasing 10% gradient, from 100% n-hexane to 100% ethyl acetate to afford in a total of 11 fractions (Fr.A-K). With the bioactivity-guided method, each fraction was then tested against S. mutans, S. sanguinis, and C. albicans microorganisms by disc diffusion [31,32]. The results obtained are listed in Table S1 in the Supplementary information, showing that Fr.C has the most active activity against all microbes. Fr.C was then purified by column chromatography, eluted with a solvent gradient system of 100% n-hexane to a mixture of n-hexane (95%) in ethyl acetate with a gradient of 0.5%. until 21 subfractions (Fr.C-1 to Fr.C-21) were obtained. Subfractions Fr.C.13 (29.9 mg) and Fr.C.14 (55.5 mg) were combined, and purified with methanol-water solvent, and -sitosterol (1) was eluted as a pure compound in 10% methanol in water.

Microbial strains and culture media

The American Type Culture Collection (ATCC) standard strains Streptococcus mutans ATCC 25175, S. sanguinis ATCC 10556, and Candida albicans ATCC 10231 were employed. Mueller-Hinton agar (MHA) and brain heart infusion (BHI) were used as the culture and assay media for bacteria. Potato dextose broth (Sigma-Aldrich, US) and potato dextrose agar (PDA) were used as the media for the fungi C. albicans. Before testing, stock cultures of bacterial strains were grown in the proper medium at 37 °C for 24 h. The turbidity of the bacterial suspension was then adjusted to 0.5 McFarland standard using a Biochrom microplate reader set to 620 nm in wavelength. This resulted in a final concentration of 10⁷ CFU/mL [33].

Antimicrobial assay

The fractions (Fr.A-K) with antibacterial activity were screened via the disk diffusion method. Fractions were prepared with a concentration of 2%, and the positive control, chlorhexidine (Sigma-Aldrich, US), was utilized. A 5 mL brain heart infusion (BHI) medium (Sigma-Aldrich, US) was inoculated with 100 μL of bacteria. Following a 48 h incubation period at 37 °C, bacterial strains were tested at 620 nm to meet the 0.5 McFarland standard. An aliquot of 100 μL bacterial suspension was spread evenly onto agar medium (Sigma-Aldrich, USA) to assess the zone of inhibition. Each of the fractions and chlorhexidine was dripped for 20 μL on a paper disk (6 mm, Grainger approved, Origin, USA). The paper disks were subsequently cultured for 24 h at 37 °C on a nutritional agar medium [34].

The MIC of compound 1 was assessed via the broth microdilution method. To perform the MIC assay, 100 μL of broth medium was added to each well of a 96-well microplate (NEST Biotechnology, Wuxi, China). A solvent volume of 100 μL was added to columns A-1, B-1, E-1, and F-1, while 100 μL of compound 1 with an initial concentration of 2500 g/mL was added to columns C-1, D-1, G-1, and H-1. Serial dilutions were performed from columns 1 to 12 via a microdilution technique. In addition, 5 μL of bacterial suspension was added to columns E to H. After incubation for 48 h at 37 °C, the optical density was measured using a microplate reader (Biochrom Ltd., Cambridge, UK) at 620 nm. Then, to determine the minimum bactericidal concentration (MBC) or minimum fungicidal concentration (MFC), each solution including the media, the compound, and bacteria in the microplate (columns G and H) was subsequently distributed over an agar medium and incubated for 48 h at 37 °C [32].

Ramachandran plot for enzyme validation

Ramachandran’s plots were used to assess the structural accuracy of the enzymes in this study. Each 3-dimensional structure Gbp C (PDB ID: 6CAM), Srt C (PDB ID: 8GR6), Sap 5 (PDB ID: 2QZX), and CYP51 (PDB ID: 5TZ1) were input in .pdb format on the PROCHECK SAVES v6.1 by the UCLA web server (https://saves.mbi.ucla.edu/) [35].

Molecular docking study

The crystal structures of the target enzymes, including their crystal ligands were retrieved from the RCSB Protein Data Bank (PDB) (https://www.rcsb.org/) in .pdb format. The PDB IDs for each enzyme are as follows: Glucan binding protein C from Streptococcus mutans (PDB ID: 6CAM), sortase C from Streptococcus sanguinis (PDB ID: 8GR6), Sap 5 from Candida albicans (PDB ID: 2QZX), and sterol 14-alpha demethylase enzyme from C. albicans (PDB ID: 5TZ1). The receptors are separated from water, ions, and other impurities and saved in .pdb format. -sitosterol, a compound isolated in P. crocatum, that was tested as a ligand (CID 222284) and its derivatives, -hydroxysitosterol (CID 12309569), β-sitosterol-3-O- -d-glucoside (CID 12309057), and stigmastanol (CID 241572) was obtained from Pubchem (https://pubchem.ncbi.nlm.nih.gov/) and the energy is minimized with Chemdraw 3D. Kollman partial charges for protein molecules were applied by Autodock 4.0 to create the search grid box and.pdbqt files before the structures were saved as.pdbqt files [36].

The coordinates of the receptor target were determined using a grid box search method, with a focus on the crystal ligands. Redocking was conducted after separating the ligands from the macromolecules to optimize the molecular docking method. The optimization results demonstrated that the docking parameters yielded a Root Mean Square Deviation (RMSD) value of ≤ 2.0 Å, confirming the validity of the grid box [37,38]. The GbpC grid box was X: 236.063, Y: –26.596, Z: 12.498 with box size 40×40×40 Å. The Srt C grid box was X: –2.378, Y: –10.837, Z: –12.165 with box size 40×40×40 Å. The Sap5 grid box was X: –2.378, Y: –10.837, Z: –12.165 with box size 40×50×40 Å. The CYP51 was X: 70.486, Y: 65.237, Z: 4.453 with box size 40×40×40 Å. Subsequently, docking was performed using a genetic algorithm with 100 runs with Autodock 4.0. After docking, the protein and ligand complexes with the best energy were visualized with BIOVIA Discovery Studio [39].

ADMET prediction and drug-likeness analysis

ADME (absorption, distribution, metabolism, and excretion) predictions were analyzed on the pkCSM web server (biosig.lab.uq.edu.au/pkcsm/), while drug-likeness predictions were completed from the SwissADME web server (http://www.swissadme.ch/) [40,41]. Further ADMET profiling was conducted using ADMETlab 3.0 (https://admetmesh.scbdd.com/), to predict toxicity-related parameters, including plasma protein binding (PPB), volume of distribution (VD), clearance (CL), hERG inhibition, AMES toxicity, hepatotoxicity [42,43].

Molecular dynamics simulation

Among the enzyme-ligand complexes evaluated through molecular docking, the complex with lowest binding free energy and the most stable docking pose compared to the other candidates was selected for molecular dynamics (MD) simulations. The partial atomic charges of β-sitosterol-3-O-glucoside was calculated using the semi-empirical quantum mechanical Austin Model 1-Bond Charge Correction (AM1-BCC) method, implemented in the antechamber module of AmberTools21. The topology for the ligand was generated using the Generalized Amber Force Fields 2 (GAFF2), and the complex was solvated in a box of 10 Å with TIP3P water molecules. Counter ions (Na⁺ and Cl⁻) were added to attain a salt concentration of 0.15 M using the tleap utility in AmberTools21.

Molecular dynamics simulations for each protein-ligand complex were conducted using GPU-accelerated Particle-Mesh Ewald Molecular Dynamics (PMEMD) with periodic boundary conditions in Amber20 software27. The procedure began with 2 sequential energy minimization steps: Initially, a positional restraint of 25 kcal mol⁻¹ Å⁻² was applied to the protein-ligand complex, followed by a reduction of the restraint to 5 kcal mol⁻¹ Å⁻² in the 2^nd step. The system temperature was then increased to 300 K under an NVT (Number-Volume-Temperature) ensemble for 50 ps. Subsequently, the simulation switched to an NPT (Number-Pressure-Temperature) ensemble to equilibrate the system density to 1 g cm⁻³ over 50 ps. During the following NVT equilibration, restraints on the solute were decreased incrementally by 1 kcal mol⁻¹ Å⁻² every 50 ps until completely removed. Production molecular dynamics simulations were performed for 100 ns at 300 K under NPT conditions to generate trajectories for each system. Root Mean Square Deviation (RMSD) and Root Mean Square Fluctuation (RMSF) analyses were performed using R programming language to evaluate the complex’s stability and inhibitory potential throughout the simulation.

Results and discussion

Chemical characterization of -sitosterol (1)

The isolated compound -sitosterol (1), was white crystal (65 mg) with ES- m/z 413.27; IR (in KBr) ν_max 3439, 3072, 2937, 2851, 1639, 1464, 1378, 1054, and 924 cm⁻¹. The ¹H and ¹³C-NMR shifts of compound 1 are summarized in Table 1. The ¹³C-NMR (Figure S3 in Supplementary Information) and DEPT displayed 29 carbon signals for 6 methyl carbons ( _C 11.9 C-18, 19.1 C-19, 18.9 C-21, 21.3 C-26, 19.5 C-27, and 12.0 C-29), 11 methylene carbons ( _C 37.3 C-1, 31.7 C-2, 42.2 C-4, 31.9 C-7, 21.2 C-11, 39.8 C-12, 24.4 C-15, 28.3 C-16, 33.9 C-22, 26.1 C-23, and 23.1 C-28), 8 methine carbons ( _C 71.9 C-3, 31.9 C-8, 50.2 C-9, 56.0 C-14, 56.1 C-17, 36.2 C-20, 45.9 C-24, and 29.2 C-25), 3 quaternary carbons ( _C 140.8 C-5, 36.6 C-10, and 42.4 C-13), and 1 olefinic methine carbon ( _C 121.8 C-6). ¹H-NMR (Figure S4) revealed 6 methyl protons ( 0.84 H-18, 0.82 C-19, 0.92 H-21, 0.83 H-26, 0.83 H-27, and 0.83 H-29), 11 methylene protons ( 1.85 H-1, 1.95 H-2, 2.30 H-4, 1.99 H-7, 1.02 H-11, 1.16 H-12, 1.58 H-15, 1.09 H-16, 1.34 H-22, 1.16 H-23, and 1.25 H-28), and 9 methine protons ( 3.52 H-3, 5.36 H-6, 2.02 H-8, 0.95 H-9, 1.01 H-14, 1.14 H-17, 1.35 H-20, 0.94 C-24, and 1.67 H-25).

The presence of 1 primary methyl at δ_C 12.0/C-29 (δ_H 0.83, t, J = 7 Hz, H-29) and an additional 1 methylene at δ_C 25.5/C-28 (δ_H 1.15, t, J = 3.1 Hz, H-28) confirmed the chain moiety of the stigmastane skeleton. Moreover, the attachment of a hydroxyl at C-3 (δ_C 71.9) with a hydroxyl proton (δ_H 4.53) and another double bond pair at C-5 (δ_C 140.8)/C-6 (δ_C 121.8) afforded the whole structure of 1. A comparison of the NMR data of compound 1 with those in the literature revealed that it was nearly identical which enabled us to identify it as stigmast-5-ene, known as β-sitosterol, with the structure shown in Figure 1 [44,45].

In vitro antibacterial and antifungal activity

Using the Kirby-Bauer method, the potential antibacterial activity of -sitosterol compound was evaluated by determining the zones of inhibition against S. mutans, S. sanguinis, and C. albicans. The antibacterial and antifungal data for -sitosterol are presented in Table 2. According to the MIC value classification, concentrations ranging from 101 to 500 μg·mL⁻¹ indicate strong antibacterial activity, whereas those ranging from 500 to 1000 μg·mL⁻¹ denote moderate antibacterial activity, and MIC more than 1000 μg·mL⁻¹ indicates weak antibacterial activity [46,47]. Table 2 shows that -sitosterol (1) was most active against S. mutans, with an MIC of 312.5 ± 0.16 µg∙mL^−1, which falls into the strong category. However, S. sanguinis has moderate activity, with low bactericidal (MBC) ability. Compared with S. sanguinis, compound 1 also has moderate fungicidal ability against C. albicans. Therefore, it can be concluded that compound 1 has the strongest activity against S. mutans. This could be due to several factors, one of which is the different specificities of virulence factors in S. mutans and S. sanguinis, one of which is the biofilm-forming virulence enzyme expressed by S. mutans, which is highly dependent on glucose, so that if the compound is an inhibitor of the enzyme, the compound can disrupt the biofilm of S. mutans without affecting S. sanguinis, which does not depend on the gene or enzyme [48-50]. Therefore, the mechanism of the compound was predicted via in silico molecular docking to determine the interaction of the compound with key enzymes in S. mutans, S. sanguinis, and C. albicans, which cause pathogenicity [51].

Table 1 ¹H-NMR (700 MHz) and ¹³C-NMR (175 MHz) data of compound 1 in CDCl₃ compared to the reference.