Abstract

Background

   Periodontitis is the primary cause of tooth loss in adults and is
   associated with cardiovascular disorders and type 2 diabetes, etc.,
   significantly impairing patients’ quality of life. Salvia
   miltiorrhiza(S. miltiorrhiza), a traditional Chinese medicine,
   possesses properties such as anti-inflammatory and antioxidative
   effects and has potential in the treatment of periodontitis; however,
   its mechanism of action remains unclear. The aim of this study was to
   investigate the therapeutic mechanism of S. miltiorrhiza in
   periodontitis using an integrated approach combining network
   pharmacology, molecular docking, molecular dynamics simulations, and
   experimental validation.

Methods

   The active components and target genes of Salvia miltiorrhiza were
   collected from the TCMSP and SwissTargetPrediction databases, while the
   target genes of periodontitis were obtained from the GeneCards, OMIM,
   Disgenet, and DrugBank databases. The intersection of these targets was
   identified using jvenn, followed by network visualization using STRING
   and Cytoscape 3.10.2 software. CytoNCA plug-ins were used to calculate
   node scores and identify hub genes. The DAVID database was used to
   conduct Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes
   (KEGG) pathway enrichment analyses. Molecular docking and molecular
   dynamics simulation were employed to evaluate the affinity and
   stability between key active compounds of S.miltiorrhiza and the hub
   genes. A RAW264.7 cell model induced by Pg-LPS (Porphyromonas
   gingivalis-lipopolysaccharide) was established. A CCK-8 assay was used
   to determine the effects of S.miltiorrhiza on the viability of RAW264.7
   cells, thereby screening for appropriate drug concentrations. The
   Griess method was used to test the effect of S. miltiorrhiza on nitric
   oxide (NO) in cells. The mRNA expression levels of inflammation-related
   factors were detected by RT-qPCR.

Results

   A total of 65 compounds from S. miltiorrhiza and 132 corresponding
   target genes were identified, along with 1900 periodontitis-related
   target genes. The intersection of these targets revealed 60 common
   targets. PPI network analysis revealed that S. miltiorrhiza may
   alleviate periodontitis by modulating key genes, including IL-6, BCL2,
   STAT3, TNF, TP53, CASP3, and MMP9. Molecular docking indicated strong
   binding affinities between the bioactive compounds in S. miltiorrhiza
   and these critical targets. Functional enrichment analysis suggested
   that the anti-inflammatory action of S. miltiorrhiza in periodontitis
   may involve the regulation of pathways such as AGE-RAGE, TNF and
   PI3k/AKT1 pathways. The results of cell experiments revealed that S.
   miltiorrhiza could treat and prevent periodontitis by inhibiting NO
   production and regulating the mRNA expression of inflammatory factors,
   including IL-1β, TNF, IL-6, and IL-10.

Conclusion

   S. miltiorrhiza exerts therapeutic effects on periodontitis via
   multiple components, targets and pathways, which provides a sufficient
   theoretical and practical basis for the further study of S.
   miltiorrhiza in the treatment of periodontitis.

   Keywords: Salvia miltiorrhiza(SM), Periodontitis, Network pharmacology,
   Molecular docking method, Molecular dynamics simulation, Inflammation

Introduction

   Periodontitis, the sixth most prevalent global disease, affects 11.2%
   of the worldwide population [[32]1]. This persistent inflammatory
   condition is primarily caused by plaque accumulation. Furthermore, its
   development is influenced by several additional factors, including
   bacterial infections, irritation from inadequate dental restorations,
   endocrine abnormalities, genetic predisposition, and environmental
   influences [[33]2].The fundamental concept underlying the treatment of
   periodontitis is the use of mechanical techniques to eliminate calculus
   and plaque from teeth [[34]3]. Nevertheless, in practical scenarios,
   dental surfaces that are inaccessible to equipment exist, and
   mechanical techniques are unable to completely eliminate all surface
   bacteria [[35]4]. Consequently, antibiotics are frequently administered
   in conjunction with mechanical therapy. Notably, prolonged antibiotic
   use is associated with bacterial resistance and other adverse effects
   [[36]5].

   Traditional Chinese medicine (TCM) has been widely employed in the
   treatment of periodontitis due to its minimal side effects, substantial
   efficacy, and low propensity to induce drug resistance [[37]6]. For
   nearly two millennia, S. miltiorrhiza, an herb widely used in
   traditional Chinese medicinal formulations, has been employed in China
   to address various clinical conditions. Research has demonstrated that
   S. miltiorrhiza has potential health benefits in terms of
   anti-inflammatory, anticoagulant, and antioxidant activities, thereby
   offering new possibilities for the treatment and prevention of
   periodontitis [[38]7].

   While several studies have shown the multisystemic pharmacological
   effects of S. miltiorrhiza, the domain of measurement Chinese medicine
   is insufficiently broad to evaluate the pharmacological effects of S.
   miltiorrhiza from a singular system. Network pharmacology, an
   innovative research paradigm, focuses on integrating disease networks,
   gene networks, and therapeutic target networks [[39]8]. Molecular
   docking is a methodology that employs stoichiometry to simulate the
   geometric arrangement of molecules and intermolecular forces,
   facilitating the identification of intermolecular interactions and the
   prediction of receptor-ligand interaction complex configurations
   [[40]9]. Molecular dynamics employs computational methods to solve
   molecular motion equations, simulating the structure and properties of
   molecular systems [[41]10]. Recently, accumulating evidence has
   demonstrated that computer-aided drug design approaches can predict the
   active components, targets, and action mechanisms of natural compounds;
   this not only significantly shortens the timeline and reduces costs in
   new drug development but also opens novel research frontiers for
   natural product exploitation [[42]11].

   The purpose of this study was to explore the potential targets and
   signalling pathways of S. miltiorrhiza in treating periodontitis
   through network pharmacology, molecular docking, and molecular dynamics
   simulation, providing a robust theoretical foundation for future
   experimental investigations and clinical applications of this herb.
   Additionally, an in vitro cell model was established to assess the
   regulatory effect of S. miltiorrhiza on inflammatory cytokine secretion
   by macrophages stimulated with Pg-LPS, offering theoretical support for
   its development as a therapeutic agent for periodontitis.All precise
   procedures of the method are illustrated in Fig. [43]1.

Fig. 1.

   [44]Fig. 1
   [45]Open in a new tab

   Workflow of a pharmacology-based approach to investigate the
   therapeutic mechanisms of S. miltiorrhiza on periodontitis

Materials and methods

Evaluation of the active components and targets of S. miltiorrhiza

   To identify the main components of S. miltiorrhiza, we searched the
   Traditional Chinese Medicine Systematic Pharmacology Database and
   Analysis Platform (TCMSP). The criteria for screening active components
   in S. miltiorrhiza were formulated based on the pharmacokinetic (ADME)
   properties of drugs in traditional Chinese medicine. The goal
   parameters were a bioavailability (OB) of at least 30% and a drug-like
   property (DL) of at least 0.18. Targets associated with the candidate
   bioactive compounds were identified via the SwissTargetPrediction
   database. The standard gene names of the relevant targets were
   identified by checking the acquired target names in the UniProt
   database. (The website of the public databases was shown in
   Table [46]1)

Table 1.

   Software and database for network Pharmacology
   Name                    Website
   TCMSP                   [47]https://tcmspw.com/tcmsp.php
   SwissTargetPrediction   [48]http://www.swisstargetprediction.ch/
   Genecard                [49]https://www.genecards.org/
   Omim                    [50]https://www.omim.org/
   Disgenet                [51]http://DisGeNET.org
   Drugbank                [52]https://go.drugbank.com/
   Venny 2.1.0             [53]https://bioinfogp.cnb.csic.es/tools/venny/
   STRING                  [54]https://cn.string-db.org/
   DAVID                   [55]https://david.ncifcrf.gov/
   Bioinformatics platform [56]http://www.bioinformatics.com.cn/
   PubChem                 [57]https://pubchem.ncbi.nlm.nih.gov/
   PDB                     [58]https://www.rcsb.org/
   Uniprot                 [59]https://www.uniprot.org/
   [60]Open in a new tab

Target prediction of S. miltiorrhiza in the treatment of periodontitis

   Using the term “periodontitis” as the search term, we conducted a
   search for the databases Genecard, Omim, Disgenet, and DrugBank to
   ascertain the objective of periodontitis-related disorders. The aims
   were subsequently combined and stripped of their emphasis. A crossover
   network of targets of active ingredients in S. miltiorrhiza and
   periodontitis disease-related targets was constructed via Venny 2.1.0
   to screen the intersection targets used for the following analyses.
   (The website of the public databases and analysis tool was shown in
   Table [61]1).

Development and examination of PPI networks

   Through the use of the STRING database, the common target genes for
   periodontitis in S. miltiorrhiza were identified. The construction of
   the PPI network required the application of a filter that had a
   confidence level that was greater than 0.7. Following elimination of
   the free nodes, the nodes were imported into Cytoscape 3.10.2 for
   visual analysis. The network evaluation study employed CyoNCA to
   systematically assess the effectiveness of S. miltiorrhiza treatment
   for periodontitis based on the node degree and neutrality value.

GO function and KEGG enrichment analysis

   Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathway enrichment analyses were performed via DAVID to determine the
   role of Astragali radix in biological processes, molecular functions,
   cellular components and signalling pathways in the treatment of
   periodontitis. The results were visualized on a bioinformatics
   platform. Statistical significance was observed at p < 0.05, with a
   smaller p value indicating greater enrichment and a larger count
   indicating that more genes were enriched (Table [62]1).

Map of the relationship between S. miltiorrhiza-active ingredients and target
networks

   To acquire a better understanding of the S. miltiorrhiza-active
   ingredient-target network, we removed redundant items from the Excel
   spreadsheet containing the active ingredient and its related target
   with S. miltiorrhiza. All tables were then loaded into the Cytoscape
   3.10.2 program.

Correlation analysis of drugs with disease targets

   To generate the original network, the encouraging nodes were removed,
   and the intersecting drug and disease targets (2.2) were added to the
   String database. The PPIs were then analysed and correlated, and the
   intersecting nodes were then removed.

Molecular docking analysis

   Molecular docking was performed to validate the target prediction
   results by docking selected active ingredients with hub genes. The
   bioactive components and protein structures of the hub genes were
   retrieved from the PubChem and PDB databases.The AutoDockTools1.1.2
   software was utilized in order to establish a docking relationship
   between the component elements of the medication and the target
   proteins. Subsequently, PyMOL software was used to visualize the
   results.

Molecular dynamics simulations

   Molecular dynamics simulations were performed using Amber 24 (San
   Francisco, CA, USA) to examine the interaction mechanisms between
   molecular ligands and protein receptors and to assess binding
   stability. Simulation conditions were set at a constant temperature of
   300 K and atmospheric pressure (1 Bar). The Amber99sb-ildn force field
   was applied, with water molecules as the solvent (Tip3p model). The
   simulation system’s total charge was neutralized by adding appropriate
   Na + ions. Energy minimization was conducted via the steepest descent
   method, followed by 100,000-step equilibrations under
   isothermal-isovolumic (NVT) and isothermal-isobaric (NPT) ensembles,
   respectively. The coupling constant was 0.1 ps, and each equilibration
   lasted 100 ps.Subsequently, a 100-ns free molecular dynamics simulation
   was executed, comprising 5 million steps with a 2-fs time step.
   Post-simulation, trajectory analysis was performed using the software’s
   built-in tools to calculate root-mean-square deviation (RMSD),
   root-mean-square fluctuation (RMSF), and protein radius of gyration for
   each amino acid. These analyses were combined with free energy
   calculations (MMGBSA) and free energy landscape data.

In vitro experiment

Chemicals and materials

   S. miltiorrhiza was purchased from Tongrentang Pharmacy (Beijing,
   China). Foetal bovine serum (FBS) was obtained from Oricell (USA) and
   used as a supplement for cell culture. Dulbecco’s Modified Eagle Medium
   (DMEM), sourced from Gibco (USA), served as the primary culture medium.
   Phosphate-buffered saline (PBS), procured from Biosharp (China), was
   used for washing and dilution. For the cell viability assays, the CCK-8
   kit was supplied by Biyuntian (China). P. gingivalis lipopolysaccharide
   (P. gingivalis LPS), a key experimental agent, was purchased from
   InvivoGen (USA). Additionally, chloroform, isopropyl alcohol, and
   anhydrous ethanol, which are essential for RNA extraction and
   purification, were obtained from Guangdong Guanghua Technology (China).
   DEPC-treated water, which was used to prevent RNA degradation, was
   obtained from Blue Sky (China). To extract RNA, the TRIzol-based
   TRleasyTM Total RNA Extraction reagent manufactured by Yisheng Biology
   in China was utilized. To perform complementary DNA (cDNA) synthesis, a
   premix kit from Yisheng Biology (China) was used. Additionally,
   quantitative polymerase chain reaction (qPCR) analysis was conducted
   using the qPCR SYBR Green Kit, which was also provided by Yisheng
   Biology (China).

Preparation of S. miltiorrhiza extract

   The extraction of S. miltiorrhiza was followed a literature-based
   protocol: the material was extracted twice with water at 80°C for
   120 min per cycle. The resulting aqueous extracts were combined,
   filtered, and concentrated under vacuum at 60°C to a relative density
   of 1.18 to 1.22. Ethanol was then added to achieve 70% (v/v) alcohol
   concentration, and the solution was allowed to stand for 12 h. The
   supernatant was collected, after which ethanol was recovered under
   reduced pressure. The remaining solution was concentrated to a thick
   paste and dried in a vacuum oven to yield the S. miltiorrhiza extract
   (yield: 10%) [[63]12].

Methods of cell treatment and culture

   In this particular investigation, the mouse macrophage line RAW264.7
   was utilized. This particular line was obtained from the Laboratory of
   Oral Biology at Jilin University. To cultivate the cells, Dulbecco’s
   Modified Eagle Medium (DMEM) supplemented with 10% foetal bovine serum
   (FBS) and 1% cyocyanin was added throughout the growth process. For
   optimal development conditions, cultures were maintained in an
   incubator that was humidified and maintained at 37 °C with 5% carbon
   dioxide. To maintain cell viability and proliferation, RAW264.7
   macrophages were subcultured every two days.

Cell counting kit 8 (CCK-8) was utilized to assess cell viability

   To seed RAW264.7 cells (murine macrophage line, catalogue number:
   TCM13) into 96-well plates, a density of 0.5 × 10⁴ cells was added to
   each well. This was performed at a specific concentration. Following
   the adherence of the cells to the surface of the plate, the culture
   media was changed to complete media that contained S. miltiorrhiza
   extract, and the cells were then incubated for approximately
   twenty-four hours. Subsequently, 100 µL of DMEM was added to each well,
   into which 10 µL of CCK-8 reagent was added. The plates were incubated
   for one hour in 5% CO₂, and 450 nm absorbance was measured. The
   absorbance measurements were used to create a cell viability curve to
   identify the best S. miltiorrhiza extract concentration for future
   research.

Relationship between S. miltiorrhiza and the generation of NO in RAW264.7
cells

   For the purpose of seeding the RAW264.7 macrophage line onto 6-well
   plates, a total of 2 × 10⁵ cells were individually added to each well.
   The amount of cell suspension added to each well was 2 mL. After an
   initial incubation period of twenty-four hours, which was designed to
   allow the cells to attach to one another and become more stable, the
   cells were then treated with specialized medium that corresponded to
   the experimental groups. These groups included the blank control group,
   the LPS stimulation group (treated with lipopolysaccharide alone), and
   three experimental groups treated with low, medium, or high
   concentrations of S. miltiorrhiza extract. After 24 h, 50 µL of the
   supernatant was collected from each well and transferred to a 96-well
   plate. Subsequently, 50 µL of Griess Reagents I and II were added to
   each well at room temperature. After an incubation period of three
   minutes, the absorbance of each well was measured at 540 nm using an
   enzyme-linked immunosorbent assay reader. The absorbance of each well
   was measured. Sodium nitrite (NaNO₂) was utilized to create a standard
   curve, which facilitated the determination of nitric oxide (NO) content
   in the growth medium for each experimental group.

Real-time quantitative fluorescence PCR (qRT-PCR) was used to assess the
expression levels of TNF-α, IL-6, IL-1β and IL-10 mRNA

   RAW264.7 cells were seeded onto 6-well plates at a density of 2 × 10⁵
   cells per well when the plates were prepared. Following the adherence
   of the cells to the surface of the plate, the culture media was changed
   to complete media that contained different quantities of
   lipopolysaccharide (LPS) and S. miltiorrhiza extract. TRIzol was used
   to extract total RNA after 24 h of therapy. A NanoDrop 2000
   spectrophotometer was used to measure the RNA concentration, and a
   reverse transcription kit was used to produce cDNA. Quantitative
   real-time polymerase chain reaction (qRT-PCR) was carried out to
   evaluate the mRNA expression levels of TNF-α, IL-6, IL-1β and IL-10.
   The quantification of the expression levels of these cytokines was
   carried out by applying the 2^-ΔΔCt approach, with β-actin serving as
   the internal reference gene for standardization.The primer sequence is
   displayed in Table [64]2.

Table 2.

   Primer sequences of Real-time PCR
   Gene    Sequence of Primers (5ʹ→3ʹ)
   β-actin F: GGAGATTACTGCCCTGGCTCCTA
           R: GACTCATCGTACTCCTGCTTGCTG
   IL-6    F: GGAGCCCACCAAGAACGATA
           R: ACCAGCATCAGTCCCAAGAA
   TNF-α   F: CTCATGCACCACCATCAAGG
           R: ACCTGACCACTCTCCCTTTG
   IL-1β   F: ATGAAGGGCTGCTTCCAAAC
           R: TCTCCACAGCCACAATGAGT
   IL-10   F: GCTCTTACTGACTGGCATGAG
           R: CGCAGCTCTAGGAGCATGTG
   [65]Open in a new tab

Statistical analyses

   Data analysis and graphing of the experimental data were performed
   using GraphPad Prism 9.0 software. The results of the above - mentioned
   in vitro experiments were replicated three times, and all results were
   presented as mean ± standard deviation (± SD). For comparisons between
   groups, if the data conformed to the normal distribution and passed the
   homogeneity of variance test, one - way analysis of variance (ANOVA)
   was employed. Statistical significance was defined as p < 0.05.

Results

Evaluation of S. miltiorrhiza active components and action targets

   A total of 65 main components of S. miltiorrhiza were found through
   TCMSP database (Table [66]3)0.132 active metabolite-related targets of
   S.miltiorrhiza were obtained.

Table 3.

   Active components of S. miltiorrhiza
   MOL Compound OB DL
   MOL001601 1,2,5,6-tetrahydrotanshinone 38.74 0.35
   MOL001659 Poriferasterol 43.82 0.75
   MOL001771 poriferast-5-en-3beta-ol 36.91 0.75
   MOL001942 isoimperatorin 45.46 0.22
   MOL002222 sugiol 36.11 0.27
   MOL002651 Dehydrotanshinone II A 43.76 0.4
   MOL002776 Baicalin 40.12 0.75
   MOL000569 digallate 61.84 0.25
   MOL000006 luteolin 36.16 0.24
   MOL006824 α-amyrin 39.51 0.76
   MOL007036 5,6-dihydroxy-7-isopropyl-1 33.76 0.28
   MOL007041 2-isopropyl-8-methylphenanthrene-3,4-dione 40.86 0.22
   MOL007045 3α-hydroxytanshinoneIIa 44.92 0.44
   MOL007048 acrylic acid 48.24 0.31
   MOL007049 4-methylenemiltirone4 34.34 0.22
   MOL007050
   2-(4-hydroxy-3-methoxyphenyl)-5-(3-hydroxypropyl)-7-methoxy-3-benzofura
   ncarboxaldehyde 62.78 0.39
   MOL007051 6-o-syringyl-8-o-acetyl shanzhiside methyl ester 46.69 0.71
   MOL007058 formyltanshinone 73.44 0.41
   MOL007059 3-beta-Hydroxymethyllenetanshiquinone 32.16 0.4
   MOL007061 Methylenetanshinquinone 37.07 0.36
   MOL007063 przewalskin a 37.1 0.64
   MOL007064 przewalskin b 110.32 0.43
   MOL007068 Przewaquinone B 62.24 0.41
   MOL007069 przewaquinone c 55.74 0.4
   MOL007070
   (6 S,7R)-6,7-dihydroxy-1,6-dimethyl-8,9-dihydro-7 H-naphtho[8,7-g]benzo
   furan-10,11-dione 41.31 0.45
   MOL007071 przewaquinone f 40.3 0.45
   MOL007071 przewaquinone f 43.67 0.2
   MOL007071 przewaquinone f 52.47 0.45
   MOL007071 przewaquinone f 57.95 0.55
   MOL007071 przewaquinone f 56.96 0.52
   MOL007071 przewaquinone f 30.38 0.37
   MOL007088 cryptotanshinone 52.34 0.39
   MOL007093 dan-shexinkum d 38.88 0.55
   MOL007094 danshenspiroketallactone 50.43 0.3
   MOL007098 deoxyneocryptotanshinone 49.4 0.28
   MOL007100 dihydrotanshinlactone 38.68 0.32
   MOL007101 dihydrotanshinoneI 45.04 0.36
   MOL007105 epidanshenspiroketallactone 68.27 0.3
   MOL007107 [67]C09092 36.06 0.24
   MOL007108 isocryptotanshi-none 54.98 0.39
   MOL007111 Isotanshinone II 49.91 0.39
   MOL007115 manool 45.04 0.2
   MOL007118 microstegiol 39.61 0.27
   [68]Open in a new tab

Acquisition of gene targets for periodontitis

   Genetic targets for periodontitis were identified in the GeneCards
   database. These genes were subsequently refined using the median
   relevance score planting, resulting in a final collection of 1626 gene
   targets related to periodontitis. By merging and downsizing the gene
   targets from the OMIM database, an additional 8 gene targets related to
   periodontitis were identified. Furthermore, 682 additional gene targets
   related to periodontitis were discovered in the DisGeNET database. An
   analysis of the DrugBank database yielded twenty-eight gene targets
   associated with periodontitis. When the gene targets related to
   periodontitis were extracted from the four databases, they were
   categorized and minimized once the extraction process was complete.
   Finally, the 1900 gene targets related to periodontitis were
   identified.

Identification of the intersection of periodontitis gene targets and S.
miltiorrhiza active ingredient targets of S. miltiorrhiza

   Through the online analysis tool Venn diagram, the predicted targets of
   S. miltiorrhiza and periodontitis were analysed, and the overlapping
   targets obtained were used as the core targets of S. miltiorrhiza in
   the prevention of periodontitis. As shown in Fig. [69]2A, there were 60
   intersecting genes between S. miltiorrhiza and periodontitis.

Fig. 2.

   Fig. 2
   [70]Open in a new tab

   The network pharmacological analysis of S. miltiorrhiza and
   periodontitis (A): Venn diagram of targets of active components of S.
   miltiorrhiza and periodontitis treatment-related targets; B: Network of
   PPI C: Potential core targets of S. miltiorrhiza in the treatment of
   periodontitis D: S. miltiorrhiza component-target-pathway network for
   periodontitis therapy. E: GO pathway enrichment analysis; F: KEGG
   pathway enrichment analysis

Construction and analysis of PPI networks and acquisition of core targets

   Following the establishment of a high level of confidence and burying
   of the individual targets, the overlapping targets were loaded into the
   STRING database (Fig. [71]2B). Additionally, a PPI network map of the
   core targets was constructed via Cytoscape 3.10.2, and the top 7
   targets were selected according to their MCC values. IL-6, BCL2, STAT3,
   TNF, TP53, CASP3, and MMP9 are the core targets of S. miltiorrhiza in
   preventing periodontitis (Fig. [72]2C).

Relationship diagrams for the construction of S.miltiorrhiza active
ingredient-target networks

   The 65 active chemicals identified in Sect. 3.1 were amalgamated and
   subsequently de-emphasized, resulting in 132 probable action targets
   for S. miltiorrhiza active compounds, which were determined by
   predicting the relevant targets via the Swiss Target Prediction
   database. From the 132 possible action targets of S. miltiorrhiza, 18
   active compounds associated with these targets were identified. The
   active ingredient-target network graph had 151 nodes and 116 edges
   (Fig. [73]2D). The 132 nodes illustrated as green circles in the graph
   denote the target genes of S. miltiorrhiza for periodontitis treatment,
   whereas the remaining 18 nodes indicate the active components of S.
   miltiorrhiza linked to these targets.

   S. miltiorrhiza was produced via Cytoscape. The network diagram shows
   S. miltiorrhiza, active compounds, and target interactions.

GO and KEGG pathway enrichment analysis

   The DAVID database indicates that GO and KEGG analyses of the
   prospective therapeutic targets of Salvia miltiorrhiza in periodontitis
   have been conducted. According to the findings of the GO analysis, the
   majority of the factors that were included in MF were gene expressiona
   and apoptotic, etc. In contrast, the majority of the factors that were
   included in the CC were macromolecular complex and nucleoplasm, etc.
   However, the majority of the factors associated with BP were enzyme
   binding, identical protein binding and protein binding, among others
   (Fig. [74]2E). According to KEGG data, the primary targets of S.
   miltiorrhiza in periodontitis include the AGE-RAGE, TNF, PI3K-Akt,
   HIF-1, and IL-17 signalling pathways in diabetic complications (Fig.
   [75]2F). These findings indicate that S. miltiorrhiza exerts its
   therapeutic effect on periodontitis through a multifaceted and
   multichannel mechanism.

Molecular docking results and visualization

   Combining the degree value of protein interactions of key targets in
   the PPI network analysis, the network degree analysis of “drug-active
   ingredient-target-disease network”, literature review [[76]13], and the
   core active ingredients Baicalin (CID:64982), Digallate (CID:
   54711004), and Luteolin (CID: 5280445),

   Poriferasterol (CID: 5281330) and Sugiol (CID: 94162) were selected for
   docking with the core target proteins AKT1 (ID: 1H10), CASP3 (2CJY),
   IL-6 (ID: 1ITV), MMP-9 (ID: 1ITV), STAT3 (ID: 5AX3), TNF (ID: 5M2J) and
   TP53 (ID: 3LGF). The molecular docking results revealed that the
   binding energy of these active ingredients docked with key target
   proteins was less than − 5 kcal/mol (Fig. [77]3).

Fig. 3.

   [78]Fig. 3
   [79]Open in a new tab

   Heat map of binding energy for S. miltiorrhiza compounds docking with
   core gene molecules, analyzed individually

   The lowest binding energy using hydrogen bonds was selected from the
   molecular docking results, which are visually represented in
   Fig. [80]4. As the energy becomes increasingly close to zero, the
   likelihood of activation increases; this is because the conformations
   of the ligand and the receptor are stable when they are in this state,
   which shows that the active component has some binding energy with the
   target if the binding energy (BE) of the active ingredient is
   -4.25 kcal/mol or less. In a broader sense, this suggests that the
   active ingredient has some binding energy. A good binding energy with
   the target is indicated by a binding energy (BE) of 5.00 kcal/mol or
   higher, whereas a strong binding energy is indicated by a BE of
   -7.00 kcal/mol or higher. By forming hydrogen bonds, including
   stabilizing and aromatic hydrocarbon hydrogen bonds, as well as π-π
   interactions, the active component S. miltiorrhiza has a significant
   binding affinity for the amino acid residues of the primary target. The
   bulk of the binding energies was approximately 5.0 kcal/mol, which
   validated the favourable docking outcomes.

Fig. 4.

   [81]Fig. 4
   [82]Open in a new tab

   Molecular docking pattern diagram of S. miltiorrhiza with some main
   targets against periodontitis

Molecular dynamics results

   The baicalin-MMP9 and poriferasterol-TNF complexes were selected as
   typical pairings for molecular dynamics during 100 ns of simulation to
   further validate the stability of protein-ligand interactions. These
   parameters included the root mean square deviation (RMSD), the root
   mean square fluctuation (RMSF), the radius of gyration (Rg), the
   solvent-accessible surface area (SASA), and the hydrogen bonds
   (H-bonds)(Table [83]4and[84]5).As shown in Figs. [85]5A and [86]6A,
   RMSD analysis of the baicalin-MMP9 and poriferasterol-TNF complexes
   indicated that the protein–ligand systems reached equilibrium after 20
   and 50 ns, suggesting stable binding between the bioactive compounds
   and their target atoms.

Table 4.

   Associative free energy analysis (kcal/mol)
   VDWAALS EEL EPB ENPOLAR EDIPESAR ELTAG gas ELTAG solve Delta total
   Baicilin-MMP9 -40.99 -10.90 34.39 -3.21 0 -51.89 31.18 -20.71
   Poriferasterol-TNF -49.07 -2.20 18.87 -4.65 0 -51.27 14.17 -37.09
   [87]Open in a new tab

Table 5.

   Average of RMSD, RMSF, RG, SASA and H hydrogen
   RMSD RMSF Rg SASA H-hydrogen
   ligand protein ligand protein
   Baicilin-MMP9 0.540725 0.57235 0.3561 0.8607 3.79655 6.36195 2–3
   Poriferasterol-TNF 0.6147 0.9846 0.32825 0.977525 30.571575 86.588275 2
   [88]Open in a new tab

Fig. 5.

   [89]Fig. 5
   [90]Open in a new tab

   MD of baicalin and MMP9, (A) Baicilin-MMP9-simulated RMSD changes; (B)
   Baicilin-MMP9-simulated RMSF changes; (C) Baicilin-MMP9 Rama Plot
   dynamic models; (D) Baicilin-MMP9 dynamic Rg models; (E) Baicilin-MMP9
   hydrogen bond number dynamic models; (F) Baicilin-MMP9 2D free energy
   morphology; (G) Baicilin-MMP9 3D free energy morphology; (H) Residue
   energy of baicilin-MMP9

Fig. 6.

   [91]Fig. 6
   [92]Open in a new tab

   MD of poriferasterol and TNF (A) Poriferasterol-TNF-simulated RMSD
   changes; (B) RMSF changes simulated by poriferasterol-TNF; (C) Rama
   Plot poriferasterol-TNF dynamic models; (D) Poriferasterol-TNF dynamic
   RG models; (E) Poriferasterol-TNF hydrogen bond number dynamic models;
   (F) Poriferasterol-TNF 2D free energy morphological distribution; (G)
   Poriferasterol-TNF 3D free energy morphological distribution; (H)
   Residue energy of Poriferasterol-TNF

   Throughout the simulation, RMSF was able to provide information
   regarding the degree of variability that each amino acid residue
   demonstrated. In the baicalin-MMP9 complex, residues 60–80 of MMP9
   exhibited greater flexibility compared to other regions, suggesting
   localized variations in residue dynamics (Fig. [93]5B). In the
   poriferasterol-TNF complex, multiple regions of TNF exhibited residue
   flexibility, indicating dynamic structural changes during the
   simulation (Fig. [94]6B).SASA values for the baicalin-MMP9 complex were
   low throughout the simulation, indicating a stable binding interaction
   (Fig. [95]5C). In contrast, the SASA values for the poriferasterol-TNF
   complex were higher, suggesting a more flexible binding interaction
   (Fig. [96]6C).Rg values have been used to determine the degree of
   compactness. Values for the baicalin-MMP9 complex exhibited a decrease
   at 42 ns, followed by an increase at 50 ns; subsequently, they
   exhibited relative stability, indicating the overall structural
   integrity of the complex (Fig. [97]5D). In contrast, Rg values for the
   poriferasterol-TNF complex decreased at 5 ns and continued to decline,
   indicating relatively poor structural stability (Fig. [98]6D).

Results of in vitro experiments

S. miltiorrhiza’s effect on RAW264.7 cell viability

   RAW264.7 cells were subjected to different concentrations of S.
   miltiorrhiza (0, 0.01, 0.1, 1, 2, 3, 4, 5, and 9 mg/ml) to assess its
   impact on cellular activity. The findings demonstrated that lower
   concentrations (0.01 mg/ml, 0.1 mg/ml, and 1 mg/ml) markedly improved
   cell activity, but higher concentrations (2 mg/ml, 3 mg/ml, 4 mg/ml,
   and 5 mg/ml) exhibited no significant impact on cell activityP < 0.05).
   Treatment with the greatest quantity (9 mg/ml) suppressed cellular
   activityP < 0.05) (Fig. [99]7A). To further investigate the mechanism
   of S. miltiorrhiza in antiinflammatt periodontitis through cell
   experiments, we selected three concentrations of 0.01 mg/ml S.
   miltiorrhiza, 0.1 mg/ml S. miltiorrhiza and 1 mg/ml S. miltiorrhiza for
   subsequent experiments.

Fig. 7.

   [100]Fig. 7
   [101]Open in a new tab

   RAW264.7 inflammation-related factors were reduced by S. miltiorrhiza.
   (A)Effects of S. miltiorrhiza on RAW264.7 cell viability; (B) Impact of
   S. miltiorrhiza on RAW264.7 NO (C、D、E、F) Effects of S. miltiorrhiza on
   the levels of inflammatory factors (IL-1, 、IL-6, TNF-α, IL-10) in
   RAW264.7. Repeat three times per group (n = 3). *p < 0.05, **p < 0.01,
   ***p < 0.001, suggesting a considerable difference from osteogenic
   induction medium-only cells

S. miltiorrhiza prevented LPS-induced NO release in RAW264.7 cells

   NO content in the cell culture supernatant of S. miltiorrhiza treated
   samples was indirectly measured using the Griess method after 24 h of
   treatment. As illustrated in the figure, lipopolysaccharide (LPS), used
   as a positive control, markedly elevated NO levels. In contrast, S.
   miltiorrhiza treatment significantly suppressed NO production P < 0.05)
   (Fig. [102]7B).

Impact of S. miltiorrhiza on inflammatory factor levels in RAW264.7 cells

   Based on quantitative real-time PCR results, pretreatment with S.
   miltiorrhiza at concentrations of 0.01 mg/mL, 0.1 mg/mL, and 1 mg/mL
   significantly reduced the mRNA expression levels of pro-inflammatory
   cytokines TNF-α, IL-6, and IL-1βP < 0.05). Conversely, the expression
   of the anti-inflammatory cytokine IL-10 was significantly
   increasedP < 0.05).The above results indicate that S. miltiorrhiza can
   not only inhibit the secretion of pro-inflammatory factors but also
   promote the expression of anti-inflammatory factors, suggesting that S.
   miltiorrhiza has potential anti-inflammatory properties(Fig. [103]7C
   and F).

Discussion

   Periodontitis, a common chronic infectious condition affecting 50% of
   the global population, can severely compromise masticatory function,
   result in tooth loss, and induce facial collapse, considerably reducing
   quality of life [[104]14, [105]15]. The primary treatment for
   periodontitis focuses on eliminating and controlling inflammation in
   periodontal tissues to facilitate cell and tissue repair [[106]16]. One
   study concluded that drug therapy effectively treats periodontitis
   [[107]17]. Long-term drug treatment may result in drug resistance,
   bacterial imbalance, bone loss, and other adverse effects. S.
   miltiorrhiza, a traditional Chinese medicine that has been used for
   more than 2,000 years, contains bioactive components such as phenolic
   acids and tanshinones, which exhibit antioxidant and anti-inflammatory
   properties by inhibiting inflammatory mediators and enzymes [[108]18].
   However, conclusive evidence for its efficacy in periodontitis remains
   lacking, prompting this study to employ network pharmacology, molecular
   docking, molecular dynamics, and cellular assays to explore its
   therapeutic potential. Network pharmacological analysis is extensively
   utilized to clarify the ways by which natural plant constituents
   address periodontitis and associated target investigations [[109]19].
   This study identified 65 natural plant components (OB ≥ 30%, DL ≥ 0.18)
   and 132 active metabolite-related targets in S.miltiorrhiza.
   Cross-analysis identified 60 common targets between the selected
   natural plant components and periodontitis, suggesting potential
   activity against S. miltiorrhiza.

   The active ingredient-target-disease network indicated that Baicalin,
   Digallate, Luteolin, Poriferasterol, and Sugiol have higher degree
   values. Consequently, these key components were further analysed for
   their therapeutic effects on periodontitis. PPI network analysis
   revealed six potential key targets—AKT1, CASP3, IL-6, MMP9, STAT3, TNF,
   and TP53—that are significantly involved in the treatment of
   periodontitis with S.miltiorrhiza. Previous research has indicated that
   TNF-α is crucial for the inflammatory response in periodontitis,
   resorption of alveolar bone, and loss of connective tissue connection.
   TNF may cause periodontitis by weakening the oral mucosal barrier
   [[110]20]. In addition to its role in activating the NF-κB signalling
   pathway, the TNF molecule plays a significant role in facilitating the
   generation of inflammatory factors such as IL-6 and IL-11, and it also
   plays a role in exerting an influence on the inflammatory process that
   occurs within the body [[111]21, [112]22]. IL-6 acts as a
   proinflammatory agent in adaptive immunity and serves as an effective
   enhancer of inflammation [[113]23]. There is a high frequency of AKT1
   activation in periodontitis, which serves to regulate inflammation and
   immunological responses [[114]24]. MMP9, which is primarily secreted by
   neutrophils and macrophages, is a proteolytic enzyme responsible for
   degrading collagen fibres and extracellular matrix components, thereby
   regulating tissue inflammation and disease. MMP9 primarily activates
   osteoclasts by removing collagen from demineralized bone, facilitating
   bone destruction. Consequently, its expression in periodontitis results
   in the destruction and reconstruction of periodontal tissue
   [[115]25–[116]27].

   Additionally, computational simulations were used to explore the
   binding affinity and.

   mechanisms of core bioactive ingredients with key periodontitis-related
   targets. The molecular docking results revealed that the docking scores
   for the five components with the seven targets were all less than
   − 5 kcal/mol. Molecular dynamics simulations have demonstrated that the
   complexes formed between baicalin and MMP9, as well as between
   stigasterol and TNF, exhibit remarkable resistance to dissociation.
   Additional research, which included the examination of hydrogen bonds
   and the computation of binding free energy, has provided evidence that
   the interactions between these molecules are both feasible and
   powerful. Importantly, the simulations did not predict any key factors
   that would significantly violate the drug-like properties or lead to
   serious adverse and toxic reactions; this suggests a greater likelihood
   of success for these complexes in clinical settings.

   GO and KEGG enrichment analyses suggested that S. miltiorrhiza may
   target numerous signalling pathways for periodontitis therapy. These
   include the AGE-RAGE, TNF, PI3K-Akt, HIF-1, and IL-17 signalling
   pathways. The AGE-RAGE signalling pathway is activated by the
   inflammatory response, which promotes the release of inflammatory
   mediators, induces cell apoptosis, and leads to alveolar bone
   destruction [[117]28]. PI3K-Akt is a classic signalling pathway that
   regulates inflammation and cell differentiation. Studies have shown
   that bone morphogenetic protein-2 (BMP-2) can promote osteoblast
   differentiation by activating this pathway, providing a potential
   therapeutic strategy for periodontal tissue regeneration [[118]29].
   Therefore, the effectiveness of S.miltiorrhiza is achieved through the
   regulation of various biological pathways.

   By evaluating the expression of inflammatory factors in RAW264.7 cells,
   S. miltiorrhiza was examined to determine whether it possessed the
   ability to prevent periodontitis.It has been demonstrated that S.
   miltiorrhiza effectively reduces the levels of TNF-α, IL-6, and IL-1β
   while upregulating the mRNA expression of the anti-inflammatory factor
   IL-10 in RAW264.7 cells. As a result, S. miltiorrhiza has the potential
   to regulate cellular inflammatory processes, which in turn may lessen
   the inflammatory response caused by LPS in RAW264.7 cells.

   S.miltiorrhiza’s ability to regulate periodontitis-related
   anti-inflammatory processes will be our main focus of future research,
   which will help us provide experimental and theoretical evidence as to
   how S.miltiorrhiza can be used to successfully treat and prevent
   periodontitis. This investigation should reveal further therapeutic
   potential of S. miltiorrhiza, leading to better periodontitis treatment
   strategies and a firmer theoretical base.

   Although this study provides valuable insights, several limitations
   should be acknowledged. First, the predictive results derived from
   network pharmacology and molecular docking require further validation
   through in vivo experiments. Second, the in vitro experiments focused
   exclusively on macrophages and did not assess other cell types
   critically involved in periodontitis pathogenesis, such as fibroblasts
   and epithelial cells. Future research should prioritize the following
   directions: (1) conduct animal studies to evaluate the therapeutic
   efficacy of S. miltiorrhiza in established in vivo periodontitis
   models; (2) investigate its antioxidant mechanisms through dedicated in
   vivo and in vitro experimental validation; and (3) explore its
   translational potential for treating other inflammatory diseases.

Conclusion

   Within the context of the treatment of periodontitis, this study
   combines the investigation of S.miltiorrhiza with the emerging field of
   network science in order to investigate the active components of
   S.miltiorrhiza as well as the underlying molecular mechanisms that are
   involved. In light of the complexity of Chinese herbal medicine,
   particularly the multifaceted nature of its constituents and the
   therapeutic effects they produce, theoretical findings were further
   validated through the use of in vitro experiments. These experiments
   provide significant evidence of S.miltiorrhiza’s periodontitis therapy
   mechanism.

Acknowledgements