Periodontal Implications of Porphyromonas gingivalis-Derived Metabolites in Host Antioxidant and Anti-Inflammatory Mechanisms: A Computational Analysis

Introduction

The term “periodontic diseases” describes chronic inflammatory conditions that impact the gingival tissue, alveolar bone, cementum, and periodontal ligament, which make up the periodontium that surrounds the tooth [1]. Gingival inflammation is the main characteristic of periodontitis and is often due to accumulation of bacteria and debris, also known as dental plaque, between the gum line and the tooth. Surprisingly, plaque usually forms on rough tooth surfaces, where gingival inflammation is exacerbated by bacterial colonization and growth [2]. The pathophysiology and development of periodontitis are prominently caused by Porphyromonas gingivalis (P. gingivalis) [3]. However, despite similar microbial loads, the severity of periodontitis varies significantly among individuals, suggesting that systemic, environmental, and genetic factors influence disease susceptibility [4,5]. Apart from its local effects, P. gingivalis is increasingly recognized as a cause of systemic diseases, including non-alcoholic fatty liver disease (NAFLD), diabetes mellitus, Alzheimer’s disease, rheumatoid arthritis, coronary artery disease, chronic kidney disease, and cancer [6–11]. For instance, periodontal disease is a risk factor that contributes to the development of NAFLD, and Th17 induces inflammation and promotes intracellular lipid accumulation in hepatocytes [10]. Notably, the correlation between periodontitis and these systemic diseases is believed to involve oxidative stress and chronic inflammation due to the reactive oxygen species (ROS) levels [12]. Specifically, Hoare et al demonstrated that during periodontitis, inflammatory processes in the oral cavity can spread throughout the body, resulting in bacterial components that can cause oxidative stress, cell cycle regulation, apoptosis, cell proliferation, transport mechanisms, and proinflammatory responses [13]. Physiologically, the endogenous antioxidants neutralize ROS to maintain homeostasis, but any disruptions in this antioxidant defense system can exacerbate oxidative stress and produce several pathological consequences [14].

The P. gingivalis exerts its pathogenicity through a range of virulence factors, including lipopolysaccharides, fimbriae, capsules, and proteases, which trigger host immunological responses and disrupt the antioxidant system [15]. As evidenced by Fernández-Veledo et al (2024) and Krishnamurthy et al (2023), it also has an effect on the development of various conditions, including metabolic syndromes, inflammatory disorders, neurodegenerative diseases, and cancers [16,17]. Although the pathophysiology of periodontitis has been thoroughly investigated in relation to virulence factors of P. gingivalis, the impact of its metabolic byproducts on host antioxidant systems remains largely unexplored. Earlier research by Sheets et al (2012) and Yu et al (2014) have shown that P. gingivalis-derived metabolites, including short-chain fatty acids (SCFAs), lipopolysaccharides (LPS), and proteolytic enzymes like gingipains can disrupt the host’s antioxidant defense mechanisms [18,19]. However, their potential disruption of specific host antioxidant proteins could offer novel insights into the pathogenesis of periodontitis. Paraoxonase 1 (PON1) is a critical antioxidant enzyme associated with high-density lipoproteins (HDL) and plays a key role in mitigating oxidative stress by hydrolysing lipid peroxides. PON1 is a 43-kDa protein consisting of 354 amino acids, and encoded by the PON1 gene located on chromosome 7. PON1 enzyme plays a key role in protecting lipids from oxidation while also exerting anti-inflammatory effects [20]. Reduced PON1 activity has been implicated in several diseases, including atherosclerosis, neurodegenerative disorders, and cancer [21,23]. While studies have implicated metabolites like P. gingivalis metabolites in disrupting host redox homeostasis, the precise molecular mechanisms by which they impair PON1 activity have not been clearly defined. In recent decades, computational approaches have demonstrated groundbreaking potential, offering high-throughput, cost-effective, and time-efficient alternatives to experimental methods. These approaches enable the prediction and visualization of complex host–microbe molecular interactions at atomic and molecular levels [24]. Therefore, this study aimed to explore the potential interaction between P. gingivalis-derived metabolites and the host antioxidant enzyme PON1, with an emphasis on understanding how these microbial metabolites impair PON1 function and contribute to oxidative stress. Initially, PON1 expression in the oral epithelium and saliva was evaluated, followed by analyzing the functional relevance through protein interaction networks, molecular pathways, and disease associations. Subsequently, P. gingivalis-derived metabolites were computationally screened to identify potential inhibitors of PON1 using molecular docking and dynamic simulation techniques, enabling the prediction of binding affinities, interaction patterns, and structural impacts of key metabolites on PON1.

Material and Methods

PON1 TISSUE EXPRESSION AND NETWORK ANALYSIS:

The GeneCards database was accessed using the search term “human PON1” to determine PON1 expression data across human tissues. GeneCards is a comprehensive human gene compendium integrated from over 80 digital sources, enabling researchers to effectively explore and connect data on human genes, variants, proteins, diseases, cells, and biological pathways based on experimental evidence. Subsequently, the PON1 interacting proteins were examined using the STRING (version 12.0) database, which compiles both predicted and experimental interaction data from text mining, co-expression, high-throughput lab experiments, and curated pathway databases. A high confidence score of 0.7 was set to ensure the reliability of protein interactions. Following this, ShinyGO version 0.77 was employed for functional analysis, using the Reactome pathway module to determine the significance of PON1 network proteins and the Jensen Disease module to explore disease associations, both with a false discovery rate (FDR) cutoff of 0.05.

PROTEIN PREPARATION FOR DOCKING:

The three-dimensional structure of human PON1 was retrieved from the Protein Data Bank (PDB), database, selecting the entry PDB ID: 1V04 due to its high resolution (2.16 Å), complete amino acid sequence, and inclusion of physiologically relevant calcium and phosphate ions. Prior to simulation, these bound ions were removed, and the protein structure was optimized using the OPLS3 force field within the Protein Preparation Wizard of Maestro v11.2 (Schrödinger, LLC, New York, NY, USA). During the preparation process, hydrogen atoms were added, water molecules were removed, and corrections were made to bond orders, disulfide bonds, and any missing atoms to ensure structural integrity. Then, the PrankWeb server (Masaryk University, Brno, Czech Republic; https://prankweb.cz/) was used to predict ligand-binding sites by submitting the PDB ID. The server utilizes multiple sequence alignment (MSA) data from the HSSP (homology-derived structures of proteins) database to identify potential binding pockets and displays a sequence-based view of the protein. Finally, the ligand-binding residues were identified, and a 20 Å receptor grid was generated around these residues using the Receptor Grid Generation module in Maestro v11.2 to accurately define the active site for docking studies.

:

The Virtual Metabolic Human (VMH) database provides comprehensive information on human and gut microbial metabolism, incorporating manually curated metabolic models assembled from biochemical, genomic, and physiological data. By searching for P. gingivalis within the gut microbiota panel of the VMH database, a total of 870 corresponding metabolites were identified and collected. The structural data for these metabolites were retrieved from the NCBI PubChem database in MOL format and subsequently processed using the LigPrep module in Maestro v11.2 (Schrödinger, LLC, New York, NY, USA). The ligand structure optimization was performed using the OPLS_2005 force field, which included tautomer generation, desalination, retention of specified chiralities, and 32 conformer generation per metabolite.

MOLECULAR DOCKING:

All P. gingivalis-derived metabolites were docked to the processed PON1 structure using the Glide module in Maestro v11.2 (Schrödinger, LLC, New York, NY, USA). Molecular docking was performed in standard-precision (SP) mode to calculate the binding affinity of each metabolite at the PON1 binding site. The Glide module uses the OPLS3 force field to accurately model the energetics of metabolite-protein interactions, evaluating hydrophobic, electrostatic, and hydrogen bonding interactions, ensuring precise prediction of binding affinities. Also, the algorithm assesses metabolite positions, orientations, and conformations, generating the Glide Score (G Score), which represents binding energy in kcal/mol. The metabolite exhibiting the highest binding affinity or lowest G score toward PON1 was selected for further analysis and visualized using the Ligand Interaction Diagram tool in Maestro v11.2 (Schrödinger, LLC, New York, NY, USA).

MD SIMULATION:

The MD simulation is a robust and widely used technique for examining protein dynamics and tracking conformational changes over time upon ligand binding. In this study, the simulation was performed using GROMACS 2020.4, a Linux-based high-performance computing package (GROMACS, Royal Institute of Technology and Uppsala University, Sweden) to assess the structural stability of PON1 upon binding with the selected P. gingivalis metabolite. The AMBER99SB force field was applied to the PON1 protein, while the topology of the P. gingivalis metabolite was generated using the GAFF2 force field. A cubic simulation box was constructed, positioning the protein–metabolite complex at the center and solvating it using the TIP3P water model, using counterions for neutralization. Energy minimization was carried out using the steepest descent algorithm for 50 000 steps, followed by equilibration in 2 phases: the NVT (constant number of particles, volume, and temperature) ensemble and the NPT (constant number of particles, pressure, and temperature) ensemble, each for 100 ps. The production run of the MD simulation was executed for 100 ns at a constant temperature of 300 K and pressure of 1 bar, using a 2-fs time step. The simulation protocol was adopted based on work by Nasebi et al (2024), which demonstrated that a 100 ns simulation duration with explicit solvent models like TIP3P yields reliable and quantitative insights into protein–ligand dynamics [25]. The simulation setup was executed for apo PON1 (the unbound protein) to compare the structural stability.

POST SIMULATION ANALYSIS:

The simulation trajectory was analyzed using XMGRACE software (Center for Computational Sciences, University of Tsukuba, Japan) to assess key structural parameters, including root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), radius of gyration (Rg), solvent-accessible surface area (SASA), and the number of hydrogen bonds (HB). A metabolite-protein complex was considered structurally stable if RMSD values remained below 2 nm, in accordance with the criteria reported by Wagner et al (2019) [26]. A stable complex was indicated by minimal RMSF fluctuations in critical residues, along with no significant deviations in SASA or hydrogen bonding profiles.

Results

PON1 EXPRESSION ACROSS TISSUES AND INTERACTOMIC NETWORK:

The GeneCards database analysis indicated that PON1 is highly expressed in blood plasma and serum, underscoring its vital role in systemic circulation and detoxification. Notably, elevated expression levels were also observed in monocytes and synovial fluid, suggesting involvement in immune regulation and inflammatory processes. Furthermore, moderate expression in oral epithelium and salivary glands suggests PON1 has a role in maintaining oral and periodontal health (Figure 1). To further explore its functional interactions, a protein–protein interaction (PPI) network was constructed using the STRING database, which enables identification of proteins that function along with PON1 in physiological or pathological processes. The top 100 interacting proteins were selected based on the highest confidence scores (≥0.7), ensuring the reliability of the predictions (Figure 2). The network included PON family members (PON2, PON3), proteins involved in lipid metabolism (APOA1, CLU), and oxidative stress regulators such as SOD1 and GPX3, emphasizing the multifaceted role of PON1 in cellular homeostasis.

FUNCTIONAL ANALYSIS OF PON1 INTERACTING PROTEINS:

Subsequently, pathway enrichment analysis was conducted to discover the major biological processes linked to PON1 and its interacting proteins. The results highlighted several critical pathways, including neurotransmitter clearance, indicating a potential role for PON1 in synaptic function and neuroprotection, and G protein-coupled receptor (GPCR) signaling, which suggests its involvement in modulating immune responses and metabolic processes. The detoxification of ROS pathway underscored PON1’s critical role in mitigating oxidative stress. Additionally, the involvement of pathways related to lipid metabolism, post-translational protein phosphorylation, and PPAR-α activation shows its contributions to maintaining lipid homeostasis, regulating intracellular signaling, and modulating inflammatory responses (Figure 3). We investigated the pathological relevance of PON1-associated proteins, revealing PON1’s involvement across a broad spectrum of diseases, including cardiovascular, periodontal, neurological, metabolic, and immunological conditions. The strong association with cardiovascular diseases was likely due to its role in lipid metabolism and oxidative stress regulation, while its expression in oral tissues suggests a role in periodontal diseases. The presence of neuroprotective interactions indicates a connection with neurological disorders, while its contribution to systemic inflammation and metabolic regulation links it to diabetes and obesity. Additionally, the involvement of oxidative stress and immune response pathways suggests roles in various immunological disorders (Figure 4).

VIRTUAL SCREENING AND DOCKING:

A total of 870 metabolites derived from P. gingivalis were retrieved from the VMH database. Following structural optimization and conformer generation using the LigPrep module, 1276 unique conformers were generated. These optimized metabolite structures were subsequently subjected to molecular docking using the Glide module in Schrödinger’s Maestro. Docking was conducted in standard-precision (SP) mode, and metabolites were ranked based on their binding affinity to human PON1 using the G Score. Among all conformers, the metabolite with the molecular formula C38H70O10P+demonstrated the strongest binding interaction, with the lowest G Score of −9.784 kcal/mol, docking score of −9.784 kcal/mol, and a Glide E-model energy of −110.51, indicating a highly favorable and stable complex formation (Table 1). To further investigate the binding interaction, a two-dimensional interaction diagram was generated using the Ligand Interaction Diagram panel, Maestro v11.2 (Schrödinger, LLC, New York, NY, USA). The analysis revealed that the metabolite C38H70O10P+ established 4 hydrogen bonds with key residues in the PON1 active site (Figure 5). These interactions likely play a critical role in stabilizing the PON1-C38H70O10P+ complex, supporting its high binding affinity observed in the docking analysis.

:

Molecular docking analysis identified C38H70O10P+ possessing strong binding affinity toward human PON1. However, since docking provides a static view of protein–metabolite interaction and does not account for the dynamic nature, a 100-ns simulation was performed. Two independent simulations were executed for PON1-C38H70O10P+ complex and the apo PON1 (unbound protein) systems under physiological conditions. The RMSD analysis was used to evaluate the structural stability of both systems (apo PON1 and the PON1-C38H70O10P+ complex) throughout the 100-ns simulation. As illustrated in Figure 6A, both systems maintained structural stability and reached equilibrium without significant fluctuations. Notably, the PON1-C38H70O10P+ complex exhibited better stability after 60 ns compared to apo PON1. The average RMSD between 60 and 100 ns was 0.238 nm for apo PON1, while it was slightly lower, at 0.219 nm, for the complex, indicating that binding of C38H70O10P+ contributed to increased protein stability.

The RMSF analysis revealed a slightly higher flexibility in the PON1-C38H70O10P+ complex (0.12736 nm) compared to the apo PON1 (0.11822 nm), suggesting a modest increase in local residue mobility, likely due to conformational changes induced by the metabolite (Figure 6B). The Rg remained consistent at 1.88 nm for both the apo and complex forms, indicating that the overall compactness and structural integrity of the protein were preserved (Figure 7A). A marginal increase in SASA was observed in the complex (178.7740 nm2) relative to the apo form (177.3479 nm2), suggesting slight alterations in surface exposure upon metabolite binding (Figure 7B). The number of hydrogen bonds was nearly identical between the apo form (237.97) and the PON1-C38H70O10P+ complex (236.45), indicating that metabolite binding did not significantly disrupt the protein’s internal hydrogen bonding network (Figure 7C). Therefore, docking and MD simulation results confirm the stable binding of the PON1-C38H70O10P+ complex, suggesting that the P. gingivalis-derived C38H70O10P+ metabolite impairs the antioxidant activity of PON1 through its inhibitory interaction. This inhibition could contribute to increased oxidative stress, promoting the progression of periodontitis and potentially leading to associated systemic complications.

Discussion

Periodontitis is a widespread inflammatory disease that affects individuals of all age groups, with a notably higher prevalence among the elderly [27]. The progression from early gingivitis to periodontitis is initiated by the emergence of specific oral pathogens, with P. gingivalis playing a particularly prominent role. Interestingly, beyond its well-known role in local biofilm formation and plaque accumulation, P. gingivalis is increasingly recognized for its involvement in systemic diseases. Notably, Tarantino G et al (2021) reported that P. gingivalis induces inflammation in the gut, spleen, and liver, potentially contributing to the progression of various systemic conditions [28]. Physiologically, the human oral cavity harbors one of the most diverse microbial communities, comprising several bacterial species, including P. gingivalis. In susceptible individuals, the accumulation of microbial plaque can trigger a complex host-mediated inflammatory and immune response, which serves as the primary etiological factor underlying the periodontal tissue damage observed in periodontitis [29]. According to Basic and Dahlén (2023), metabolites produced by the diverse microbial community in the dental biofilm arise from complex microbial interactions and contribute to destructive periodontal disease [30]. In particular, metabolites from P. gingivalis enhance its adaptation to oxidative stress and promote inflammation-driven periodontal tissue destruction. However, such crucial metabolites of P. gingivalis on the host’s antioxidant proteins remain largely unexplored. Recent studies, such as Masumoto et al (2018), reported that PON1 plays a protective role in periodontitis by promoting cytodifferentiation and mineralization of human periodontal ligament fibroblasts [31]. Ghorbani et al (2023) identified PON1 with anti-atherosclerotic and antioxidant properties that displayed reduced function in ischemic heart disease and periodontitis [32]. This may result from host-microbial metabolite interactions, as Jin et al (2022) found that HMGB1, PON1, MCP-1, and P. gingivalis are involved in development of meconium-stained amniotic fluid, indicating an oxidative/antioxidant imbalance accompanied by inflammation [33]. Therefore, there is a vital need to explore the molecular mechanisms underlying the interaction between P. gingivalis-derived metabolites and the antioxidant enzyme PON1 to treat periodontitis conditions.

This study investigated the interaction between P. gingivalis-derived metabolites and PON1 through a multi-step computational approach, incorporating protein expression analysis, molecular docking, and MD simulations. Initial analysis using the GeneCards database, which integrates data from sources such as GTEx and the Human Protein Atlas (HPA), revealed high PON1 expression in blood plasma and serum, elevated levels in monocytes and synovial fluid, and moderate expression in the oral epithelium and salivary glands. This moderate expression suggests a role for PON1 in oral and periodontal health. We also constructed a PPI network using the STRING database to explore the functional associations of PON1 (Figure 1). A high confidence score threshold of 0.7 was applied to reduce false positives; however, this may have excluded weaker yet potentially biologically significant interactions [34]. Subsequently, enrichment analysis of the top 100 PON1-interacting proteins (Figures 2–4) revealed their significant involvement in immune responses, metabolic detoxification of ROS, lipid metabolism, post-translational protein phosphorylation, and PPAR-α activation, which are known to be impaired under periodontitis conditions [35]. Then, the 870 P. gingivalis-derived metabolites compiled from the VMH database were screened for their interaction with PON1 using a molecular docking approach. Among them, the metabolite C38H70O10P+ or [1-[(2R)-2,3-di(hexadec-9-enoyloxy)propoxy]-3-hydroxy-2-oxopropyl]-trihydroxyphosphanium exhibited the strongest binding affinity (−9.784 kcal/mol), forming 1 hydrogen bond each with Ser193 and Phe292, and 2 hydrogen bonds with Tyr197 (Table 1, Figure 5). As reported by Madushanka et al (2023), hydrogen bonding plays a crucial role in stabilizing protein–ligand interactions. Similarly, the PON1-C38H70O10P+ complex showed stability due to the formation of 4 hydrogen bonds.

We used MD simulation to assess the structural stability and dynamic behavior of the PON1-C38H70O10P+ complex over time. The simulation provides a high-resolution view of biomolecules’ behavior at the atomic level over time, particularly for identifying and validating potential molecular targets for natural product inhibitors. This approach also provides insights into binding stability, conformational changes, and interaction consistency under physiological conditions. The simulation was executed for 2 systems – apo PON1 and PON1-C38H70O10P+ complex – to assess structural stability upon metabolite binding. Subsequently, statistical comparisons of RMSD, RMSF, Rg, and SASA were performed to enhance the reliability of the findings. The RMSD measures the average distance between the atoms of an apo protein or protein–ligand complex [37]. Low and stable RMSD values indicate that the apo protein or the protein–ligand complex maintains conformational stability during the simulation. The average RMSD for apo PON1 and complex was 0.238 and 0.219 nm, respectively, below the 2-nm threshold, indicating that the protein remained stable upon ligand binding (Figure 6A).

The RMSF measures the flexibility of individual amino acid residues over time [37]. Both the apo PON1 and PON1-C38H70O10P+ complex showed lower fluctuations of 0.12736 and 0.11822 nm, respectively, indicating enhanced metabolite-protein interactive stability (Figure 6B). The Rg assesses the overall compactness of the protein structure. Both apo and complex forms maintained a Rg value of 1.88 nm, suggesting that metabolite binding did not disrupt the protein’s structural integrity (Figure 7A). SASA quantifies the surface area of the protein exposed to solvent. Both the complex (178.7740 nm2) and the apo form (177.3479 nm2) of PON1 showed minor changes in SASA values upon metabolite interaction (Figure 7B). The similar number of hydrogen bonds in apo PON1 and the PON1-C38H70O10P+ complex indicates that metabolite binding does not disrupt the protein’s structure (Figure 7C). Previous studies have shown that P. gingivalis metabolites can disrupt host enzyme activity, leading to oxidative stress and inflammation [38]. Consistent with these findings, the current study results suggest that C38H70O10P+ can alter PON1 function, potentially increasing oxidative stress and contributing to periodontitis and other related systemic complications. However, the study has several limitations, including: (1) the use of rigid receptor conformations in docking, which may not accurately reflect the dynamic flexibility of the PON1; (2) the limited time scale of MD simulations; and (3) the absence of in vitro and in vivo experiments to validate the inhibitory effect of the C38H70O10P+ metabolite on PON1 activity. Moreover, techniques like surface plasmon resonance (SPR) and crystallography will be essential to validate metabolite binding and its thermodynamic properties, while site-directed mutagenesis can help confirm key interaction sites.

Conclusions

In recent years, multiple studies have emphasized the significant role of microbiome-derived metabolites in compromising host defense mechanisms, particularly by impairing antioxidant proteins such as PON1, thereby contributing to the progression of periodontal disease. This study provides novel insights into the interaction between P. gingivalis-derived metabolites impairing the host antioxidant enzyme PON1 in periodontitis. Initial tissue-based expression studies and enrichment analysis confirmed that PON1 is highly expressed in oral epithelium and salivary glands and plays a critical role in modulating oxidative stress and inflammation through its PPI networks. Out of 870 metabolites screened from P. gingivalis, the C38H70O10P+ demonstrated the strongest binding affinity to PON1 (−9.784 kcal/mol) in molecular docking. Furthermore, MD simulations revealed that the metabolite exhibits stable interactive behavior with PON1 at ligand binding sites, suggesting inhibition of its normal enzymatic function. Therefore, the C38H70O10P+ metabolite affects the PON1 antioxidant function and can lead to oxidative stress in periodontitis and other systemic complications. However, the computational predictions require experimental validation to confirm the inhibitory effects of C38H70O10P+ on PON1 activity.