Original Article

Biochemical Pathway and Protein-Network Analysis of Dental Caries Based on Systems Biology Approaches

Please cite this article as follows: Mohamadi Z, Khamverdi Z, Taherkhani A. Biochemical pathway and protein-network analysis of dental caries based on systems biology approaches. Avicenna J Med Biochem. 2023; 11(1):19-32. doi:10.34172/ajmb.2023.2408

Background

According to a report about tooth decay (TD), the Global Burden of Disease study estimated that more than two billion of the world’s population are suffering from this highly prevalent chronic disorder (1). TD has remained a dominant global health issue in most modern and developing countries, in which 60%–90% of schoolchildren and a considerable number of adults have already been affected by this disease (2,3). TD is a comprehensive disorder, and various factors are involved in its occurrence. The pathogenesis of the disease can be elucidated by a logic diagram, which includes three circles and the interactions between them. The circles represent microbial load (plaque), intake of food, as well as host factors (e.g., genetic factors). The interaction among these circles could lead to dental caries (DCs). The parameter of “time” has also been added to the above circles, which displays the duration of interaction between caries-associated factors. Previous studies have shown that Streptococcus mutans and lactobacilliare primarily involved in the beginning and development of DC, respectively. Fermentable sugars are the primary substrates for these bacteria, and the carbohydrates produced from bacteria are mostly stable in the biofilm. DC could cause other oral disorders such as tooth pain, tooth defects, loss of teeth, and tooth crowding (4-12). However, the dentine–pulp complex has been reported to have regenerative characteristics by secreting a tertiary dentine extracellular matrix following a tooth injury, protecting the cells beneath the lesion, and maintaining the vitality of the tissues (13). Since previous studies have documented that the inflammatory reaction occurs due to bacterial infection in DC (14-21), further investigation is inevitable to illustrate the exact molecular mediation.

Reanalyzing the microarray datasets containing enormous gene expression data provides an opportunity to discover new genes previously not linked with the disorder and, therefore, could identify new potential biomarkers for diagnosis and curative aims. In addition, identifying novel caries-associated genes could help understand the etiology of this disorder (22,23). It may be hypothesized that the significant alteration in the expression of various genes in patients with DC compared to the individuals with sound teeth results in the deregulation of several signaling pathways and biological processes (BPs) associated with the pathogenesis of DC. Moreover, the most critical genes and their regulating markers involved in the etiology of DC could be identified by analyzing the protein-protein interaction (PPI) network and gene regulatory network (GRN) associated with the disease, respectively (24). Kinase enrichment analysis also provides an opportunity to illustrate protein kinases involved in the phosphorylation of transcription factors (TFs) regulating the critical genes (25).

The present study, therefore, aimed to determine (1) differentially expressed genes (DEGs) within the pulpal tissue obtained from healthy and severely carious human teeth, (2) the hub genes in the PPI network, (3) the most significant TFs regulating the hub genes, (4) protein kinases involved in the TF’s activity, and (5) signaling pathways and gene ontology (GO) in patients with tooth caries compared to individuals with sound teeth. This was performed by reanalyzing the gene expression dataset GSE1629 (22).

The gene dataset GSE1629, containing almost 15 000 human sequences, was established by McLachlan et al (22) to identify genes associated with carious lesions to promote diagnostic and therapeutic approaches. McLachlan et al (22) performed their study by utilizing pooled RNA extracted from patients with DC and individuals with sound teeth extracted from the pulpal region. The completely decayed and healthy premolar and molar teeth, collected for orthodontic aims from individuals from age 20 to 30 years, were extracted from the Birmingham Dental Hospital. DC illustrated the carious lesions varying from enamel to decays. Next, the extracted teeth were instantly overwhelmed in RNA stabilizing buffer using the RNA Later (Sigma, UK), and the complete tissue of the pulp was cautiously eliminated by using a sterile dental probe and forceps (26).

Materials and Methods

Microarray Expression Dataset Analysis

The dataset GSE1629 (2) was chosen for analyses from the Gene Expression Omnibus (GEO) (NCBI GEO, http://www.ncbi.nlm.nih.gov/geo) (27). The GSE1629 consisted of four pooled pulpal tissue samples from 11 patients with clinically DC showing deep dentinal lesions/pulp exposure (n = 2) and 12 with sound teeth (n = 2). This dataset was developed on the platform of Affymetrix Human Genome U133A Array (GPL96). DEGs between pulpal tissue of patients with DC compared to individuals with sound teeth were identified using the GEO2R tool (28) with the benchmark of |Log2 fold change(FC)| > 1 as well as the false discovery rate (FDR) < 0.001. Then, the Benjamini and Hochberg procedure was applied to calculate corrected P values.

The Protein-Protein Interaction Network and Clustering Analyses

The Search Tool for the Retrieval Of Interacting Genes (STRING) knowledge database (version 11.0 STRING, http://string-db.org) (29) was utilized for identifying possible connections between DEGs. Next, unconnected genes were removed from the network. The network analyzer tool within the Cytoscape 3.9.0 (http://www.cytoscape.org/) (30) was used for calculating the centralities (e.g., degree and betweenness) of nodes, while clustering analysis was carried out using the molecular complex detection (MCODE) plugin. The benchmarks considered for each module to be statistically significant included the number of nodes > 10, Degree cutoff = 2, MCODE score > 3, Max depth = 100, and the k-score = 2. MCODE is frequently used for determining condensed zones in PPI networks named clusters (modules) (31). It has been demonstrated that clusters include genes that take part in common pathways or BPs. Moreover, most of the clusters contain a seed node known to be a vertex of the modules based on its existing biological role in living systems (32-34).

Functional Analysis

The GO analysis is frequently used for mining particular biological properties from a set of high-throughput data, including genome, transcriptome, and proteome (35). The Reactome database (https://reactome.org/) provides perceptive bioinformatics tools for illustrating and analyzing the pathways to support theoretical and experimental study and genomics data mining, which has been developed at the Ontario Institute for Cancer Research, Cold Spring Harbor Laboratory, the New York University of Medicine, and The European Bioinformatics Institute (36). In addition, Database for Annotation, Visualization, and Integrated Discovery (DAVID) v6.8 (https://david.ncifcrf.gov/) is an online bioinformatics database providing an entire set of functional annotation tools for researchers to realize the biological concepts of multiple proteins and genes (37,38). DAVID was used to illustrate the DEGs enrichment of cellular components (CCs) and molecular functions (MFs) as well as the main clusters enrichment of BPs. In addition, the Reactome database was used to demonstrate pathways significantly associated with the salient modules in the PPI network. The cutoff conditions for GO annotations and pathways were set to FDR < 0.05 and the enriched gene numbers ≥ 2.

Master Regulators of the Hub Genes

The iRegulon plugin was utilized for the prediction of upstream TFs linked to the hubs. Only TFs with a normalized enrichment score (NES) > 3.0 were considered to be statistically significant (39). The NES shows the significance of the detected TFs and a correlation with the area under the curve value associated with the motifs (40-42). Subsequently, significant motifs (e.g., GRN) consisting of TFs and their downstream hub genes were built using the iRegulon plugin. A GRN provides valuable information associated with the upstream regulators of critical genes involved in the pathogenesis of the disease. Therefore, GRN analysis might lead to identifying several crucial genes regulated with common TFs. Thus, future studies could recognize TFs as novel drug targets (43).

Consensus Sequences of Transcription Factors

JASPAR is an open-access database (https://jaspar.genereg.net/) (44) providing helpful information about TFs binding sites across six taxonomic classes, including fungi, insect, nematode, Plantae, Urochordata, and Vertebrata. This useful web server includes consensus sequences Logos and position frequency matrixes of several TFs. Position frequency matrixes illustrate the occurrences of nucleotides at each position in a set of demonstrated TF-DNA interactions and could be used for constructing position-specific scoring matrixes. The consensus sequences of these matrixes are made of nucleotides with the highest frequency at each position. The consensus sequences matching scores for top-ranked TFs were calculated in R programming (version 4.0.0) (45) based on the method described by Xiong (46). The probability of consensus sequences matching the binding sites of TFs is described as 2^(matchscore) times more than that by random chance (46).

Protein Kinases Enrichment Analysis

Protein kinases add a phosphate group from adenosine triphosphate to three residues of proteins, including Thr, Ser, and Tyr, while phosphatases catalyze the reverse reaction. Phosphorylation and dephosphorylation of proteins affect the activity, localization, stability, and interaction of substrates with other molecules in biological networks. The abnormal expression or activity of these two classes of enzymes could affect several signaling pathways and BPs in human cells, resulting in many disorders (47-49). Here, the online kinase enrichment analysis 3 (KEA3) knowledge database (https://maayanlab.cloud/kea3/) (50) was utilized for uncovering upstream protein kinases potentially involved in the phosphorylation of TFs involved in the hub gene regulation. The KEA3 uses several types of resources to illustrate upstream TFs for a set of genes (proteins). These resources are based on PPI, kinase-substrate interaction, co-occurrence, and transcript co-expression (50). In this regard, the PPI libraries are as follows: BioGRID (51), mentha (52), huMAP (53), prePPI (54), MINT (55), HIPPIE (56), PIPs (57,58), PSOPIA (59), REACTOME (60), ChengPPI (61), and STRING (62). In addition, PhosphoSitePlus (63), PhosD (64), PhosphoNetworks (65), PTMsigDB (66), ChengPPI (61), and Phospho.ELM (67) are libraries used to study kinase-substrate interactions. After a gene set is uploaded into the KEA3, upstream protein kinases are scored and ranked based on different algorithms used in different libraries. KEA3 then calculates the mean and the sum of the ranks in different libraries.

Results

The Identification of Differentially Expressed Genes in Patients With Dental Caries

In our present study, we obtained the data of four pooled samples from the pulp tissue of 11 patients with DC showing deep dentinal lesions/pulp exposure (n = 2 ) and 12 individuals with sound teeth (n = 2). A total of 196 DEGs, including 146 overexpressed and 50 underexpressed genes, with an FDR less than 0.001 and the |Log2 FC| > 1 were identified using the GEO2R online tool and considered for further analysis in the present study. Top-ranked over- and underexpressed genes in carious teeth compared to those in the sound teeth are presented in Figure 1, and Figure 2a demonstrates the volcano plot of DEGs. Figure 2b illustrates the hierarchical clustering of the hub genes, and Figure 2c demonstrates the interactions between hub genes obtained from the DAVID database.

Close

Figure 1.

Top-10 Up- and Downregulated Genes in Pulpal Tissue of Patients With Carious Lesion Compared With the Healthy Pulp Tissue Regarding Their Log2 FC. Note. FC: Fold change. The x-axis shows the genes’ name, while the y-axis demonstrates the value of Log2 FC

Figure 1.

Top-10 Up- and Downregulated Genes in Pulpal Tissue of Patients With Carious Lesion Compared With the Healthy Pulp Tissue Regarding Their Log2 FC. Note. FC: Fold change. The x-axis shows the genes’ name, while the y-axis demonstrates the value of Log2 FC

Close

Figure 2.

(a) Volcano Plot of Genes in the Dataset GSE1629. (b) Hierarchical Clustering of the Hub Genes. (c) Interactions between the Hub Genes. Note. PPI: Protein-protein interaction. More connected genes in the main PPI network are presented larger, while the color of the nodes is linked to the betweenness centrality value

Figure 2.

(a) Volcano Plot of Genes in the Dataset GSE1629. (b) Hierarchical Clustering of the Hub Genes. (c) Interactions between the Hub Genes. Note. PPI: Protein-protein interaction. More connected genes in the main PPI network are presented larger, while the color of the nodes is linked to the betweenness centrality value

Protein-Protein Interaction Network, Clustering, and Functional Analyses

A PPI network was achieved based on the STRING database’s DEGs, and the cut-off for confidence score was set to ≥ 0.4. After removing disconnected nodes, a PPI network with 141 genes and 917 connections was transferred into the Cytoscape software for advanced analyses.

The MCODE plugin identified two substantial clusters (i.e., cluster No. 1 and cluster No. 2) within the PPI network (Figure 3). These clusters were involved in the pathways and BPs associated with DC. They consisted of several genes mainly enriched in pathways and BPs linked to the human immune system, including ‘antigen processing and presentation of peptide or polysaccharide antigen via major histocompatibility complex (MHC) class II (BP)’, ‘immune response (BP)’, ‘antigen processing and presentation of exogenous peptide antigen via MHC class II (BP)’, ‘antigen processing and presentation (BP)’, ‘inflammatory response (BP)’, ‘leukocyte migration (BP)’, ‘interferon-gamma-mediated signaling pathway (BP)’, ‘T cell costimulation (BP)’, ‘chemotaxis (BP)’, ‘positive regulation of T cell activation (BP)’, ‘MHC class II antigen presentation (Pathway), ‘T cell receptor (TCR) signaling (Pathway), and ‘interferon-gamma signaling (Pathway)’.

Close

Figure 3.

Module Analysis. Note. PPI: Protein-protein interaction; DEG: Differentially expressed genes; MCODE: Molecular complex detection. The network was built based on the DEGs in patients with dental caries compared to patients with sound teeth. The MCODE plugin discovered two salient clusters in the PPI network. The Hexagons illustrate seed nodes in each module. The size of the nodes is directly correlated with the clustering coefficient value of the genes in the leading PPI network

Figure 3.

Module Analysis. Note. PPI: Protein-protein interaction; DEG: Differentially expressed genes; MCODE: Molecular complex detection. The network was built based on the DEGs in patients with dental caries compared to patients with sound teeth. The MCODE plugin discovered two salient clusters in the PPI network. The Hexagons illustrate seed nodes in each module. The size of the nodes is directly correlated with the clustering coefficient value of the genes in the leading PPI network

At an FDR less than 0.05, a total of 35 pathways and 14 BPs were found to be significantly affected inDC. Moreover, DAVID analysis revealed that a total of 26 CCs and 10 MFs were significantly dysregulated in patients with DC. Figure 4 illustrates the top 10 significant pathways, BPs, CCs, and MFs enriched in DC. Accordingly, the most meaningful BPs and pathways were linked to the human immune system, including ‘antigen processing and presentation of peptide or polysaccharide antigen via MHC class II’ and ‘MHC class II antigen presentation (Pathway)’. Furthermore, the average centrality value within the PPI network for betweenness and degree was calculated as 0.012620178 and 13.0071, respectively. Accordingly, a total of 10 genes were revealed to have a degree and betweenness values more than twice of the nodes within the network, and therefore, were considered hub genes associated with the pathogenesis of TD (Table 1).

Close

Figure 4.

Top-10 (a) Pathways, (b) Biological Processes, (c) Cellular components, and (d) Molecular Functions Enriched in Patients with DCs Concerning Their FDR. Note. DCs: Dental caries; FDR: False discovery rate. The genes’ names are presented at the top left corner of the Logos. The consensus sequences scores are also demonstrated below the names. The x-axis shows the name of the term, and the y-axis corresponds to the Log10 FDR

Figure 4.

Top-10 (a) Pathways, (b) Biological Processes, (c) Cellular components, and (d) Molecular Functions Enriched in Patients with DCs Concerning Their FDR. Note. DCs: Dental caries; FDR: False discovery rate. The genes’ names are presented at the top left corner of the Logos. The consensus sequences scores are also demonstrated below the names. The x-axis shows the name of the term, and the y-axis corresponds to the Log10 FDR

Master Regulators of Hub Genes

The iRegulon plugin was executed to predict TFs regulating the hub genes. Only TFs with the criteria of NES > 3.0 were considered to be statistically meaningful (39). Accordingly, there were 50 TFs significantly associated with tooth caries. The most enriched TF was bromodomain and PHD finger-containing transcription factor (BPTF) with an NES equal to 7.167. A total of four hubs were found to be downstream genes of BPTF including ITGB2, C1QB, CD44, PTPRC, and C1QB. The statistics of the significant TFs are presented in Table 2.

Binding Sites Logos and Matching Scores of Transcription Factors

The top-ranked TFs with the criteria of NES > 5 were investigated in the JASPAR webserver. The consensus sequences Logos of nine TFs, including TCF12, SPI1, ESR2, STST3, RELB, PAX5, RXRA, TCF15, and HNF4A, were available in the database. The maximum and minimum scores for TFs binding sites were calculated at 22.91 and 7.5 for ESR2 and SPI1, respectively (Figure 5).

Close

Figure 5.

Consensus Sequences Logos of Top-ranked TFs Regulating the Hub Genes in DCs. Note. TF: Transcription factor; DCs: Dental caries; FDR: False discovery rate. The genes’ names are presented at the top left corner of the Logos. The consensus sequences scores are also demonstrated below the names

Figure 5.

Consensus Sequences Logos of Top-ranked TFs Regulating the Hub Genes in DCs. Note. TF: Transcription factor; DCs: Dental caries; FDR: False discovery rate. The genes’ names are presented at the top left corner of the Logos. The consensus sequences scores are also demonstrated below the names

Upstream Protein Kinases

Top-10 protein kinases based on the mean of the ranks in different libraries were considered significant. Moreover, all top-10 ranked kinases had a P value less than 0.05. Mitogen-activated protein kinase 1 (MAPK1) demonstrated the best result with the mean rank score and a P value of 19.64 and 6.96E-05, respectively (Figure 6 and Table 3). Finally, a GRN was constructed based on the eight hub genes, 50 TFs, and 10 protein kinases. This regulatory network included 440 edges (Figure 7).

Close

Figure 6.

The Mean Rank Visualization from KEA3 for TFs Regulating the Hub Genes. Note. KEA3: Kinase enrichment analysis 3; TF: Transcription factor. The x-axis demonstrates the sum of ranks in different libraries, while the y-axis shows the protein kinases’ names

Figure 6.

The Mean Rank Visualization from KEA3 for TFs Regulating the Hub Genes. Note. KEA3: Kinase enrichment analysis 3; TF: Transcription factor. The x-axis demonstrates the sum of ranks in different libraries, while the y-axis shows the protein kinases’ names

Close

Figure 7.

Gene Regulatory Network Associated with the Pathogenesis of DCs. Note. DCs: Dental caries. The yellow (hexagon), blue (circle), and violet (octagon) nodes represent hub genes, transcription factors, and protein kinases, respectively

Figure 7.

Gene Regulatory Network Associated with the Pathogenesis of DCs. Note. DCs: Dental caries. The yellow (hexagon), blue (circle), and violet (octagon) nodes represent hub genes, transcription factors, and protein kinases, respectively

Discussion

DC is a common chronic disease affecting more than 60% of childhood and a considerable percentage of adults within the most industrialized and developing countries. It is a multifactorial disorder, and several agents are involved in its pathogenesis, including cariogenic bacteria and genetic factors (4-9). Despite many types of study in this scope, our knowledge regarding the molecular mechanisms and most essential genes that participate in the beginning and development of TD is limited (22).

In the present study, we tried to determine potential biomarkers, master regulators, protein kinases regulating TFs, signaling pathways, and BPs associated with the pathogenesis of DC. Therefore, gene expression data using pooled RNA from the pulp tissue of 12 individuals with clinically sound teeth and 11 patients with DC exhibiting deep dentinal lesions/pulp exposure were obtained from the GEO web tool and analyzed. A total of 196 DEGs, including 146 overexpressed and 50 underexpressed genes, with the criteria of FDR < 0.001 and the value of |Log2 FC| > 1 were identified. Further, the most significant pathways and BPs associated with DC were identified using the Reactome and DAVID online databases, respectively. Moreover, a PPI network associated with DC was built and analyzed based on DEGs.

The MHC includes a set of genes within the jawed vertebrates taking part in the immune system (68). The MHC is positioned at chromosome-6 (6p21.3) in humans and contains more than 200 genes and many alleles (69,70) and encodes glycoproteins named human leukocyte antigens (HLAs) that take part in the body’s immune defense by the presentation of short peptides to T cells (71).

Various HLA alleles are involved in different human disorders and conditions such as autoimmune disease, inflammation, cancer, social behavior, as well as a shorter lifetime. However, the exact underlying mechanism has not been clearly understood. Overall, HLAs are classified into two main subclasses (i.e., HLA class I and HLA class II) based on their different structures, biological roles, and the expression on the surface of cells.

All nucleated cells contain different amounts of HLA class I proteins on their cell membrane. These proteins contribute to presenting peptides from degraded proteins or abnormal ones from hazardous viruses inside the cytoplasm to the TCR of the CD8 ⁺ T cells, contributing to activating the defense system. Furthermore, the HLA class II proteins are exhibited on the surface of B cells and the other antigen-presenting cells, known as active immune cells. These proteins are involved in presenting extracellular antigens to CD4 ⁺ T cells. Additionally, T cells produce HLA-II main isoforms, including HLA-DP, HLA-DR, and HLA-DQ, almost 3-5 days after activation. Furthermore, interferon-γ and interleukin (IL)-4 are produced just after a few hours following the activation of TCR (72-80).

TD occurs after the deterioration of the tooth’s organic and/or mineral texture. Matrix metalloproteinases (MMPs) take part in collagen loss, while bacteria contribute to damaging the inorganic compounds within the healthy tissues of the teeth (81). S. mutans is the most prevalent caries-associated bacteria, but Lactobacillus acidophilus and Lactobacillus casei have also been linked with dental plaque (82,83).

A significant correlation has been observed between caries resistance and a robust immune system. In this regard, caries-free individuals revealed higher serum antibody titers to streptococcal antigens than caries-susceptible individuals (84). Further, caries-free individuals were found to produce T cells with a salient ability to increase at the time of exposure to S. mutans antigens (85).Moreover, caries-resistant subjects can potentially generate more active T-helper-cells upon exposure to a minor amount of S. mutans antigens compared with patients exhibiting TD (86). Therefore, HLA-related genes from cluster 2 might enhance the immune system against S. mutans antigens and vice-versa.

It has also been reported that different HLA alleles may affect bacterium-induced TD. In this case, a significant DR-positive cell aggregation was observed within the pulp tissue of the patients with superficial caries (87). Moreover, the DR-positive cell aggregation was expanded to odontoblast cells in patients with deeper caries lesions (86).

Acton et al (88) performed a study to examine the possible correlation between MHC alleles at the HLA-DRB1, DQB1, and TNFa microsatellite loci and the number of oral bacteria associated with DC, in addition to the decayed, missing, and filled teeth index in 186 African-American females. The patients had an average age and median decayed, missing, and filled surface indexes of 20.8 years and 9 (range 0–68), respectively. Acton et al (88), reported a significant correlation between enhanced levels of S. mutans and the elevated expression of DRB1*4 and DRB1*3 (P-value = 0.005). Furthermore, the enhanced level of DRB1*8 was linked to the increased level of S. mutans(P= 0.04). A positive correlation was also observed between the levels of DRB1*1 and L. casei (P = 0.04). Moreover, the TNFa allele 103 and TNFa 117 were negatively and positively associated with the levels of L. acidophilus, respectively. Moreover, Acton et al (88) demonstrated a significant correlation between the production of HLA class II alleles, TNFa genetic, and the colonization of DC-associated oral bacteria, including S. mutans, L. casei, and L. acidophilus.

Based on the obtained results, a total of 25 nodes were involved in cluster 1 such as HLA-DRB1 and HLA-DRA. Further, cluster 2 consisted of 19 genes, including HLA-DPB1, HLA-DQA1, HLA-DMB, HLA-DQB1, HLA-DPA1, and HLA-DMA. Further analysis revealed that the HLA isotypes mentioned above were significantly over-expressed in the pulp tissue of patients with tooth caries compared to patients with sound teeth with the benchmark of FDR < 0.001 and the absolute value of Log2 FC > 1. Accordingly, it may be hypothesized that these genes contribute to the beginning and/or developing of TD. Otherwise, the enhanced expression of these genes may be due to the response of increased tooth caries. However, more experiments are necessary to approve our findings.

A total of 10 genes demonstrated degree and betweenness values higher than twice the average of the genes within the PPI network and, therefore, were found to be hub genes playing a significant role in the etiology of DC. The top-ranked genes based on the degrees were PTPRC (CD45), ITGB2, TYROBP, and MMP9, while PTPRC, CD44, MMP9, and ITGB2 revealed the highest betweenness centrality in the PPI network, respectively.

CD45 is a tyrosine-protein phosphatase antigen encoded by the PTPRC gene. It positively regulates T cells through binding to the dipeptidyl-peptidase 4 (DPP4) (89-91). Lacerda-Pinheiro et al (92) designed a study to understand the principles of the differentiation of pulp cells promoted by the bioactive molecules. The authors implanted the agarose beads (alone or covered with the products of the amelogenin gene [A + 4 and A-4]) in the mucosa of the cheeks in mice. They reported that agarose increased the recruitment of CD45⁺ cells, leading to the enhanced recruitment of leukocytes from the vascular compartment, resulting in the enhanced re-differentiation of leukocytes into osteo-chondrogenic lineages. In addition, the coated beads with A + 4 induced the production of osteo-chondrogenic markers, including RP59, SOX9, and bone sialoprotein. It is worth mentioning that RP59 is an antigen produced in the bone marrow and young osteoblasts (93,94). Moreover, SOX9 is a TF that participates in chondrocyte differentiation and skeletal development (95). Our results revealed that CD45 was significantly overexpressed in the pulp tissue of patients with DC compared to patients with sound teeth (FDR = 3.46E-05; Log2 FC = 2.8), which may be due to the response of increased TD.

MMPs are a family of zinc-dependent proteolytic enzymes taking part in the degradation of extracellular matrix proteins such as collagens in their native and denatured forms. MMPs are activated by decreasing the pH to 4.5, followed by neutralization. Previous studies have demonstrated that the dentin matrix mainly contains type I collagen, and MMPs have a specific role in DC pathogenesis (96). Several types of MMPs are involved in dentin matrix elimination, including MMP8 (collagenase-2), MMP2 (72-kD gelatinase), MMP9 (92-kD gelatinase), MMP13 (collagenase-3), and MMP20, leading to TD (81,97); therefore, the systemic application of MMP inhibitors may be helpful in the prevention of TD (96).

Wang et al (98) studied the correlation between the saliva levels of MMP2/MMP9 and different stages of DC in childhood. The levels of MMP2 and MMP9 were measured by the enzyme-linked immunosorbent assay. Wang et al (98) reported that the saliva levels of MMP9 in severe and mild-caries groups were significantly higher compared with the caries-free group (P value < 0.05). However, no significant difference was observed between the severe caries group and the mild caries group (P value > 0.05). Wang et al (98) concluded that the salivary levels of MMP9 might be correlated with TD in children. According to our results, MMP9 was significantly overexpressed in the pulp tissue of patients with TD compared to patients with sound teeth (FDR = 5.78E-05; Log2 FC = 2.21), suggesting that MMP9 is involved in the etiology of DC.

Integrin beta 2 (ITGB2) is a type of integrin chain and has been reported to be produced explicitly in leukocytes, promoting the connection between leukocytes and the endothelium (99,100). Zhang et al (101) compared the expression level of ITGB2 in cancer-associated fibroblasts (CAFs) and normal fibroblasts in patients with oral squamous cell carcinoma (OSCC) by using the reverse transcription polymerase chain reaction and western blot analyses, demonstrating that the ITGB2 expression significantly increased in CAFs compared with the matched normal fibroblasts. The authors reported that the over-expression of ITGB2 caused the hyper-activation of glycolysis through PI3K/AKT/mTOR pathways, leading to enhanced lactate excretion in CAFs and more proliferation of OSCC cells. The hyperactivity of glycolysis was found and confirmed through bioinformatics approaches and gas chromatography/mass spectrometry analysis, respectively. According to the present results, a significant increase was found in the expression of ITGB2 in the pulp tissue of patients with DC compared with patients with sound teeth (P value = 1.6E-04; Log2 FC = 1.97). We hypothesize that some of the mechanisms that mediate DC may be similar to those in OSCC, leading to more production and excretion of lactate from the pulp tissue into the dentin region, resulting in a lower pH; as a consequence, causing enhanced dental decay, although this requires confirmation.

A total of 50 TFs were determined as master regulators of eight of the hub genes. Bromodomain and BPTF were the most significant enriched TFs with the criteria of NES = 7.167, regulating ITGB2, CD44, PTPRC, and C1QB. BPTF is the central member of the human nucleosome remodeling factor (102). The misregulation of BPTF has been demonstrated in many human cancers such as bladder cancer (103), hepatocellular carcinoma (104), glioma (105), and lung cancer (106). However, the etiological role of BPTF in DC needs more studies in the future.

Moreover, MAPK1 demonstrated the best score among protein kinases regulating TFs (mean rank = 19.64). MAPKs are intracellular protein kinases necessary for inflammatory bone loss by activating MMPs and inflammatory cytokines (107). In this regard, p38 MAPK is an upstream activator of TNF-a, IL-1b and -6, and prostaglandin E2 (108), leading to the overexpression of prostaglandins, MMPs, and receptor activator of nuclear factor kappa beta (109,110), resulting in osteoclastogenesis and bone loss procedures (107). Similar mechanisms might be involved in the aberrant expression of MAPK1 in DC, leading to tooth loss; however, confirmation is needed.

Our study had certain limitations. Only individuals from the Birmingham Dental Hospital were included in the GSE1629 dataset; therefore, the present results may not entirely justify the data obtained from the other nationalities. Further, only four pooled pulpal tissue samples from 11 patients with clinically DC having deep dentinal lesions/pulp exposure and 12 individuals with sound teeth were included in the GSE1629 dataset, and the sample size was small. Hence, a greater number of individuals with DC and sound teeth may promote the reliability of statistical approaches and probably illustrate more DEGs that are significantly linked to the pathogenesis of carious teeth. Moreover, the genes profiled in this study were developed from the Affymetrix Human Genome U133A Array GPL96 platform, which probably does not represent all the genomes. In addition, our results were achieved based on bioinformatics analyses; therefore, considerable in vitro experiments are necessary to approve our findings. In future studies, molecular experiments with large targeted cohorts are obligatory to support our data.

Conclusion

Overall, the present study successfully identified 196 genes differentially expressed in the pulpal tissue of patients with tooth caries compared to patients with sound teeth (adjusted P value < 0.001; |Log2 FC > 1|). Furthermore, a total of 10 hub genes including PTPRC, ITGB2, TYROBP, MMP9, CXCL8, CD44, CCL2, C1QB, C3, and SPP1 were determined as proteins of potential importance to the pathogenesis of DC in the PPI network. Further, BPTF and MAPK1 were found to be the most significant master regulator and protein kinase potentially involved in the etiology of DC, respectively. Moreover, substantial modules in the PPI network were mostly enriched in the pathways and BPs linked to the immune system. However, more studies and wet-lab experiments must be carried out in the future to confirm these findings and demonstrate their definite role in the etiology of DC.

Acknowledgments

The authors would like to appreciate the Deputy of Research and Technology, Dental Research Center, and Research Center for Molecular Medicine, Hamadan University of Medical Sciences, Hamadan, Iran, for their support.

Authors’ Contribution

Conceptualization: Amir Taherkhani, Zeinab Mohamadi, Zahra Khamverdi.

Data curation: Amir Taherkhani.

Formal analysis: Amir Taherkhani, Zeinab Mohamadi.

Investigation: Amir Taherkhani.

Methodology: Amir Taherkhani, Zeinab Mohamadi.

Project administration: Amir Taherkhani.

Resources: Amir Taherkhani, Zeinab Mohamadi, Zahra Khamverdi.

Supervision: Amir Taherkhani.

Validation: Amir Taherkhani.

Visualization: Amir Taherkhani.

Writing – original draft: Amir Taherkhani.

Writing – review & editing: Zahra Khamverdi, Zeinab Mohamadi.

Competing Interests

The authors declare that they have no competing interests.

Data Availability Statement

The data sets included in the study were ‘The Gene Expression Omnibus (GEO) database (NCBI GEO, http://www.ncbi.nlm.nih.gov/geo)’, ‘The Search Tool for the Retrieval of Interacting Genes (STRING) online database (version 11.0 STRING, http://string-db.org)’, ‘Database for Annotation, Visualization, and Integrated Discovery (DAVID) v6.8 (https://david.ncifcrf.gov/)’, and ‘Reactome online database (https://reactome.org/)’.

Ethical Approval

The present study has been confirmed by the Ethics Committee of Hamadan University of Medical Sciences, Hamadan, Iran (Ethics no. IR.UMSHA.REC.1399.493).

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

1. Bernabé E, Marcenes W. Can minimal intervention dentistry help in tackling the global burden of untreated dental caries?. Br Dent J 2020; 229(7):487-91. doi: 10.1038/s41415-020-2155-9 [Crossref] [ Google Scholar]
2. Petersen PE, Bourgeois D, Ogawa H, Estupinan-Day S, Ndiaye C. The global burden of oral diseases and risks to oral health. Bull World Health Organ 2005; 83(9):661-9. [ Google Scholar]
3. World Health Organization (WHO). WHO Expert Consultation on Public Health Intervention Against Early Childhood Caries: Report of a Meeting, Thailand, 26-28 January 2016. WHO; 2017.
4. Selwitz RH, Ismail AI, Pitts NB. Dental caries. Lancet 2007; 369(9555):51-9. doi: 10.1016/s0140-6736(07)60031-2 [Crossref] [ Google Scholar]
5. Featherstone JD, Adair SM, Anderson MH, Berkowitz RJ, Bird WF, Crall JJ. Caries management by risk assessment: consensus statement, April 2002. J Calif Dent Assoc 2003; 31(3):257-69. [ Google Scholar]
6. Krol DM. Dental caries, oral health, and pediatricians. Curr Probl Pediatr Adolesc Health Care 2003; 33(8):253-70. doi: 10.1016/s1538-5442(03)00093-2 [Crossref] [ Google Scholar]
7. Hassell TM, Harris EL. Genetic influences in caries and periodontal diseases. Crit Rev Oral Biol Med 1995; 6(4):319-42. doi: 10.1177/10454411950060040401 [Crossref] [ Google Scholar]
8. Anderson M. Risk assessment and epidemiology of dental caries: review of the literature. Pediatr Dent 2002; 24(5):377-85. [ Google Scholar]
9. Thomson WM. Dental caries experience in older people over time: what can the large cohort studies tell us?. Br Dent J 2004; 196(2):89-92. doi: 10.1038/sj.bdj.4810900 [Crossref] [ Google Scholar]
10. Pitts NB, Zero DT, Marsh PD, Ekstrand K, Weintraub JA, Ramos-Gomez F. Dental caries. Nat Rev Dis Primers 2017; 3:17030. doi: 10.1038/nrdp.2017.30 [Crossref] [ Google Scholar]
11. Rathee M, Sapra A. Dental caries. In: StatPearls. Treasure Island, FL: StatPearls Publishing; 2020.
12. Usha C, Sathyanarayanan R. Dental caries-a complete changeover (Part I). J Conserv Dent 2009; 12(2):46-54. doi: 10.4103/0972-0707.55617 [Crossref] [ Google Scholar]
13. Smith AJ. Pulpal responses to caries and dental repair. Caries Res 2002; 36(4):223-32. doi: 10.1159/000063930 [Crossref] [ Google Scholar]
14. D’Souza R, Brown LR, Newland JR, Levy BM, Lachman LB. Detection and characterization of interleukin-1 in human dental pulps. Arch Oral Biol 1989; 34(5):307-13. doi: 10.1016/0003-9969(89)90102-7 [Crossref] [ Google Scholar]
15. Huang GT, Potente AP, Kim JW, Chugal N, Zhang X. Increased interleukin-8 expression in inflamed human dental pulps. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999; 88(2):214-20. doi: 10.1016/s1079-2104(99)70118-6 [Crossref] [ Google Scholar]
16. Jiang Y, Russell TR, Schilder H, Graves DT. Endodontic pathogens stimulate monocyte chemoattractant protein-1 and interleukin-8 in mononuclear cells. J Endod 1998; 24(2):86-90. doi: 10.1016/s0099-2399(98)80083-6 [Crossref] [ Google Scholar]
17. Matsushima K, Ohbayashi E, Takeuchi H, Hosoya S, Abiko Y, Yamazaki M. Stimulation of interleukin-6 production in human dental pulp cells by peptidoglycans from Lactobacillus casei. J Endod 1998; 24(4):252-5. doi: 10.1016/s0099-2399(98)80107-6 [Crossref] [ Google Scholar]
18. McLachlan JL, Sloan AJ, Smith AJ, Landini G, Cooper PR. S100 and cytokine expression in caries. Infect Immun 2004; 72(7):4102-8. doi: 10.1128/iai.72.7.4102-4108.2004 [Crossref] [ Google Scholar]
19. Zehnder M, Delaleu N, Du Y, Bickel M. Cytokine gene expression--part of host defence in pulpitis. Cytokine 2003; 22(3-4):84-8. doi: 10.1016/s1043-4666(03)00116-9 [Crossref] [ Google Scholar]
20. Baumgardner KR, Sulfaro MA. The anti-inflammatory effects of human recombinant copper-zinc superoxide dismutase on pulp inflammation. J Endod 2001; 27(3):190-5. doi: 10.1097/00004770-200103000-00014 [Crossref] [ Google Scholar]
21. Rutherford RB, Gu K. Treatment of inflamed ferret dental pulps with recombinant bone morphogenetic protein-7. Eur J Oral Sci 2000; 108(3):202-6. doi: 10.1034/j.1600-0722.2000.108003202.x [Crossref] [ Google Scholar]
22. McLachlan JL, Smith AJ, Bujalska IJ, Cooper PR. Gene expression profiling of pulpal tissue reveals the molecular complexity of dental caries. Biochim Biophys Acta 2005; 1741(3):271-81. doi: 10.1016/j.bbadis.2005.03.007 [Crossref] [ Google Scholar]
23. Li J, Wang Y, Wang X, Yang Q. CDK1 and CDC20 overexpression in patients with colorectal cancer are associated with poor prognosis: evidence from integrated bioinformatics analysis. World J Surg Oncol 2020; 18(1):50. doi: 10.1186/s12957-020-01817-8 [Crossref] [ Google Scholar]
24. Junker BH, Schreiber F. Analysis of Biological Networks. Vol 2. John Wiley & Sons; 2011.
25. Mahfuz A, Zubair-Bin-Mahfuj AM, Podder DJ. A network-biology approach for identification of key genes and pathways involved in malignant peritoneal mesothelioma. Genomics Inform 2021; 19(2):e16. doi: 10.5808/gi.21019 [Crossref] [ Google Scholar]
26. McLachlan JL, Smith AJ, Sloan AJ, Cooper PR. Gene expression analysis in cells of the dentine-pulp complex in healthy and carious teeth. Arch Oral Biol 2003; 48(4):273-83. doi: 10.1016/s0003-9969(03)00003-7 [Crossref] [ Google Scholar]
27. Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C. NCBI GEO: mining tens of millions of expression profiles--database and tools update. Nucleic Acids Res 2007; 35(Database issue):D760-5. doi: 10.1093/nar/gkl887 [Crossref] [ Google Scholar]
28. Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics 2007; 23(14):1846-7. doi: 10.1093/bioinformatics/btm254 [Crossref] [ Google Scholar]
29. Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res 2015; 43(Database issue):D447-52. doi: 10.1093/nar/gku1003 [Crossref] [ Google Scholar]
30. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003; 13(11):2498-504. doi: 10.1101/gr.1239303 [Crossref] [ Google Scholar]
31. Saito R, Smoot ME, Ono K, Ruscheinski J, Wang PL, Lotia S. A travel guide to Cytoscape plugins. Nat Methods 2012; 9(11):1069-76. doi: 10.1038/nmeth.2212 [Crossref] [ Google Scholar]
32. Zali H, Rezaei Tavirani M. Meningioma protein-protein interaction network. Arch Iran Med 2014; 17(4):262-72. [ Google Scholar]
33. Bader GD, Hogue CW. An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics 2003; 4:2. doi: 10.1186/1471-2105-4-2 [Crossref] [ Google Scholar]
34. Taherkhani A, Kalantari S, Arefi Oskouie A, Nafar M, Taghizadeh M, Tabar K. Network analysis of membranous glomerulonephritis based on metabolomics data. Mol Med Rep 2018; 18(5):4197-212. doi: 10.3892/mmr.2018.9477 [Crossref] [ Google Scholar]
35. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM. Gene ontology: tool for the unification of biology The Gene Ontology Consortium. Nat Genet 2000; 25(1):25-9. doi: 10.1038/75556 [Crossref] [ Google Scholar]
36. Croft D, O’Kelly G, Wu G, Haw R, Gillespie M, Matthews L. Reactome: a database of reactions, pathways and biological processes. Nucleic Acids Res 2011; 39(Database issue):D691-7. doi: 10.1093/nar/gkq1018 [Crossref] [ Google Scholar]
37. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009; 4(1):44-57. doi: 10.1038/nprot.2008.211 [Crossref] [ Google Scholar]
38. Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol 2003; 4(5):P3. [ Google Scholar]
39. Zhu FF, Yang LZ. Bioinformatic analysis identifies potentially key differentially expressed genes and pathways in orbital adipose tissues of patients with thyroid eye disease. Acta Endocrinol (Buchar) 2019; 15(1):1-8. doi: 10.4183/aeb.2019.1 [Crossref] [ Google Scholar]
40. Janky R, Verfaillie A, Imrichová H, Van de Sande B, Standaert L, Christiaens V. iRegulon: from a gene list to a gene regulatory network using large motif and track collections. PLoS Comput Biol 2014; 10(7):e1003731. doi: 10.1371/journal.pcbi.1003731 [Crossref] [ Google Scholar]
41. Taherkhani A, Kalantari S, Nafar M. Prediction of molecular signature, potential biomarkers, and molecular pathways associated with membranous nephropathy based on protein protein interactions. Rev Invest Clin 2018; 70(4):184-91. doi: 10.24875/ric.18002551 [Crossref] [ Google Scholar]
42. Hong W, Hu Y, Fan Z, Gao R, Yang R, Bi J. In silico identification of EP400 and TIA1 as critical transcription factors involved in human hepatocellular carcinoma relapse. Oncol Lett 2020; 19(1):952-64. doi: 10.3892/ol.2019.11171 [Crossref] [ Google Scholar]
43. Bayat Z, Mirzaeian A, Taherkhani A. Potential biomarkers and signaling pathways associated with the pathogenesis of primary ameloblastoma: a systems biology approach. Int J Dent 2022; 2022:3316313. doi: 10.1155/2022/3316313 [Crossref] [ Google Scholar]
44. Castro-Mondragon JA, Riudavets-Puig R, Rauluseviciute I, Lemma RB, Turchi L, Blanc-Mathieu R. JASPAR 2022: the 9th release of the open-access database of transcription factor binding profiles. Nucleic Acids Res 2022; 50(D1):D165-D73. doi: 10.1093/nar/gkab1113 [Crossref] [ Google Scholar]
45. Gentleman R. R Programming for Bioinformatics. CRC Press; 2008.
46. Xiong J. Essential Bioinformatics. Cambridge University Press; 2006.
47. Coussens L, Parker PJ, Rhee L, Yang-Feng TL, Chen E, Waterfield MD. Multiple, distinct forms of bovine and human protein kinase C suggest diversity in cellular signaling pathways. Science 1986; 233(4766):859-66. doi: 10.1126/science.3755548 [Crossref] [ Google Scholar]
48. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science 2002; 298(5600):1912-34. doi: 10.1126/science.1075762 [Crossref] [ Google Scholar]
49. Fabbro D, Cowan-Jacob SW, Moebitz H. Ten things you should know about protein kinases: IUPHAR review 14. Br J Pharmacol 2015; 172(11):2675-700. doi: 10.1111/bph.13096 [Crossref] [ Google Scholar]
50. Kuleshov MV, Xie Z, London ABK, Yang J, Evangelista JE, Lachmann A. KEA3: improved kinase enrichment analysis via data integration. Nucleic Acids Res 2021; 49(W1):W304-W16. doi: 10.1093/nar/gkab359 [Crossref] [ Google Scholar]
51. Salwinski L, Miller CS, Smith AJ, Pettit FK, Bowie JU, Eisenberg D. The database of interacting proteins: 2004 update. Nucleic Acids Res 2004; 32(Database issue):D449-51. doi: 10.1093/nar/gkh086 [Crossref] [ Google Scholar]
52. Keshava Prasad TS, Goel R, Kandasamy K, Keerthikumar S, Kumar S, Mathivanan S. Human protein reference database--2009 update. Nucleic Acids Res 2009; 37(Database issue):D767-72. doi: 10.1093/nar/gkn892 [Crossref] [ Google Scholar]
53. Bader GD, Betel D, Hogue CW. BIND: the biomolecular interaction network database. Nucleic Acids Res 2003; 31(1):248-50. doi: 10.1093/nar/gkg056 [Crossref] [ Google Scholar]
54. Pagel P, Kovac S, Oesterheld M, Brauner B, Dunger-Kaltenbach I, Frishman G. The MIPS mammalian protein-protein interaction database. Bioinformatics 2005; 21(6):832-4. doi: 10.1093/bioinformatics/bti115 [Crossref] [ Google Scholar]
55. Güldener U, Münsterkötter M, Oesterheld M, Pagel P, Ruepp A, Mewes HW. MPact: the MIPS protein interaction resource on yeast. Nucleic Acids Res 2006; 34(Database issue):D436-41. doi: 10.1093/nar/gkj003 [Crossref] [ Google Scholar]
56. Fabregat A, Jupe S, Matthews L, Sidiropoulos K, Gillespie M, Garapati P. The reactome pathway knowledgebase. Nucleic Acids Res 2018; 46(D1):D649-D55. doi: 10.1093/nar/gkx1132 [Crossref] [ Google Scholar]
57. Hu J, Rho HS, Newman RH, Zhang J, Zhu H, Qian J. PhosphoNetworks: a database for human phosphorylation networks. Bioinformatics 2014; 30(1):141-2. doi: 10.1093/bioinformatics/btt627 [Crossref] [ Google Scholar]
58. Loizou JI, El-Khamisy SF, Zlatanou A, Moore DJ, Chan DW, Qin J. The protein kinase CK2 facilitates repair of chromosomal DNA single-strand breaks. Cell 2004; 117(1):17-28. doi: 10.1016/s0092-8674(04)00206-5 [Crossref] [ Google Scholar]
59. Dinkel H, Chica C, Via A, Gould CM, Jensen LJ, Gibson TJ. PhosphoELM: a database of phosphorylation sites--update 2011. Nucleic Acids Res 2011; 39(Database issue):D261-7. doi: 10.1093/nar/gkq1104 [Crossref] [ Google Scholar]
60. Brittain J, Darwin IF. Tomcat: The Definitive Guide: The Definitive Guide. O’Reilly Media; 2007.
61. Hornbeck PV, Zhang B, Murray B, Kornhauser JM, Latham V, Skrzypek E. PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Res 2015; 43(Database issue):D512-20. doi: 10.1093/nar/gku1267 [Crossref] [ Google Scholar]
62. Oughtred R, Stark C, Breitkreutz BJ, Rust J, Boucher L, Chang C. The BioGRID interaction database: 2019 update. Nucleic Acids Res 2019; 47(D1):D529-D41. doi: 10.1093/nar/gky1079 [Crossref] [ Google Scholar]
63. Licata L, Briganti L, Peluso D, Perfetto L, Iannuccelli M, Galeota E. MINT, the molecular interaction database: 2012 update. Nucleic Acids Res 2012; 40(Database issue):D857-61. doi: 10.1093/nar/gkr930 [Crossref] [ Google Scholar]
64. Lonsdale J, Thomas J, Salvatore M, Phillips R, Lo E, Shad S. The genotype-tissue expression (GTEx) project. Nat Genet 2013; 45(6):580-5. doi: 10.1038/ng.2653 [Crossref] [ Google Scholar]
65. Merkel D. Docker: lightweight Linux containers for consistent development and deployment. Linux J. 2014(239):2.
66. Clarke DJB, Kuleshov MV, Schilder BM, Torre D, Duffy ME, Keenan AB. eXpression2Kinases (X2K) Web: linking expression signatures to upstream cell signaling networks. Nucleic Acids Res 2018; 46(W1):W171-W9. doi: 10.1093/nar/gky458 [Crossref] [ Google Scholar]
67. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 2008; 9:559. doi: 10.1186/1471-2105-9-559 [Crossref] [ Google Scholar]
68. Klein J. Natural History of the Major Histocompatibility Complex. Wiley; 1986.
69. Janeway CA Jr, Travers P, Walport M, Shlomchik MJ. The major histocompatibility complex and its functions. In: Immunobiology: The Immune System in Health and Disease. 5th ed. 2001. New York: Garland Science; 2001.
70. Hosomichi K, Shiina T, Suzuki S, Tanaka M, Shimizu S, Iwamoto S. The major histocompatibility complex (MHC) class IIB region has greater genomic structural flexibility and diversity in the quail than the chicken. BMC Genomics 2006; 7:322. doi: 10.1186/1471-2164-7-322 [Crossref] [ Google Scholar]
71. Holoshitz J. The quest for better understanding of HLA-disease association: scenes from a road less travelled by. Discov Med 2013; 16(87):93-101. [ Google Scholar]
72. Mosaad YM. Clinical role of human leukocyte antigen in health and disease. Scand J Immunol 2015; 82(4):283-306. doi: 10.1111/sji.12329 [Crossref] [ Google Scholar]
73. Dyer P, McGilvray R, Robertson V, Turner D. Status report from ‘double agent HLA’: health and disease. Mol Immunol 2013; 55(1):2-7. doi: 10.1016/j.molimm.2012.08.016 [Crossref] [ Google Scholar]
74. Li XC, Raghavan M. Structure and function of major histocompatibility complex class I antigens. Curr Opin Organ Transplant 2010; 15(4):499-504. doi: 10.1097/MOT.0b013e32833bfb33 [Crossref] [ Google Scholar]
75. Rock KL, Reits E, Neefjes J. Present yourself! By MHC class I and MHC class II molecules. Trends Immunol 2016; 37(11):724-37. doi: 10.1016/j.it.2016.08.010 [Crossref] [ Google Scholar]
76. Flajnik MF, Kasahara M. Origin and evolution of the adaptive immune system: genetic events and selective pressures. Nat Rev Genet 2010; 11(1):47-59. doi: 10.1038/nrg2703 [Crossref] [ Google Scholar]
77. Gaspari AA, Jenkins MK, Katz SI. Class II MHC-bearing keratinocytes induce antigen-specific unresponsiveness in hapten-specific Th1 clones. J Immunol 1988; 141(7):2216-20. [ Google Scholar]
78. Frasca L, Marelli-Berg F, Imami N, Potolicchio I, Carmichael P, Lombardi G. Interferon-gamma-treated renal tubular epithelial cells induce allospecific tolerance. Kidney Int 1998; 53(3):679-89. doi: 10.1046/j.1523-1755.1998.00800.x [Crossref] [ Google Scholar]
79. Bal V, McIndoe A, Denton G, Hudson D, Lombardi G, Lamb J. Antigen presentation by keratinocytes induces tolerance in human T cells. Eur J Immunol 1990; 20(9):1893-7. doi: 10.1002/eji.1830200904 [Crossref] [ Google Scholar]
80. Ko HS, Fu SM, Winchester RJ, Yu DT, Kunkel HG. Ia determinants on stimulated human T lymphocytes Occurrence on mitogen- and antigen-activated T cells. J Exp Med 1979; 150(2):246-55. doi: 10.1084/jem.150.2.246 [Crossref] [ Google Scholar]
81. Taherkhani A, Moradkhani S, Orangi A, Jalalvand A, Khamverdi Z. Molecular docking study of flavonoid compounds for possible matrix metalloproteinase-13 inhibition. J Basic Clin Physiol Pharmacol 2020; 32(6):1105-19. doi: 10.1515/jbcpp-2020-0036 [Crossref] [ Google Scholar]
82. Keyes PH. The infectious and transmissible nature of experimental dental caries Findings and implications. Arch Oral Biol 1960; 1:304-20. doi: 10.1016/0003-9969(60)90091-1 [Crossref] [ Google Scholar]
83. Caufield PW, Cutter GR, Dasanayake AP. Initial acquisition of mutans streptococci by infants: evidence for a discrete window of infectivity. J Dent Res 1993; 72(1):37-45. doi: 10.1177/00220345930720010501 [Crossref] [ Google Scholar]
84. Challacombe SJ, Lehner T. Serum and salivary antibodies to cariogenic bacteria in man. J Dent Res 1976; 55 Spec No:C139-48. doi: 10.1177/002203457605500306011 [Crossref] [ Google Scholar]
85. Lehner T, Challacombe SJ, Wilton JM, Ivanyi L. Immunopotentiation by dental microbial plaque and its relationship to oral disease in man. Arch Oral Biol 1976; 21(12):749-53. doi: 10.1016/0003-9969(76)90065-0 [Crossref] [ Google Scholar]
86. Lehner T, Lamb JR, Welsh KL, Batchelor RJ. Association between HLA-DR antigens and helper cell activity in the control of dental caries. Nature 1981; 292(5825):770-2. doi: 10.1038/292770a0 [Crossref] [ Google Scholar]
87. Yoshiba N, Yoshiba K, Nakamura H, Iwaku M, Ozawa H. Immunohistochemical localization of HLA-DR-positive cells in unerupted and erupted normal and carious human teeth. J Dent Res 1996; 75(8):1585-9. doi: 10.1177/00220345960750081001 [Crossref] [ Google Scholar]
88. Acton RT, Dasanayake AP, Harrison RA, Li Y, Roseman JM, Go RC. Associations of MHC genes with levels of caries-inducing organisms and caries severity in African-American women. Hum Immunol 1999; 60(10):984-9. doi: 10.1016/s0198-8859(99)00088-9 [Crossref] [ Google Scholar]
89. Charbonneau H, Tonks NK, Walsh KA, Fischer EH. The leukocyte common antigen (CD45): a putative receptor-linked protein tyrosine phosphatase. Proc Natl Acad Sci U S A 1988; 85(19):7182-6. doi: 10.1073/pnas.85.19.7182 [Crossref] [ Google Scholar]
90. Wu L, Fu J, Shen SH. SKAP55 coupled with CD45 positively regulates T-cell receptor-mediated gene transcription. Mol Cell Biol 2002; 22(8):2673-86. doi: 10.1128/mcb.22.8.2673-2686.2002 [Crossref] [ Google Scholar]
91. UniProt Consortium. UniProt: a hub for protein information. Nucleic Acids Res 2015; 43(Database issue):D204-12. doi: 10.1093/nar/gku989 [Crossref] [ Google Scholar]
92. Lacerda-Pinheiro S, Septier D, Tompkins K, Veis A, Goldberg M, Chardin H. Amelogenin gene splice products A + 4 and A-4 implanted in soft tissue determine the reorientation of CD45-positive cells to an osteo-chondrogenic lineage. J Biomed Mater Res A 2006; 79(4):1015-22. doi: 10.1002/jbm.a.30912 [Crossref] [ Google Scholar]
93. Wurtz T, Krüger A, Christersson C, Lundmark C. A new protein expressed in bone marrow cells and osteoblasts with implication in osteoblast recruitment. Exp Cell Res 2001; 263(2):236-42. doi: 10.1006/excr.2000.5114 [Crossref] [ Google Scholar]
94. Krüger A, Ellerström C, Lundmark C, Christersson C, Wurtz T. RP59, a marker for osteoblast recruitment, is also detected in primitive mesenchymal cells, erythroid cells, and megakaryocytes. Dev Dyn 2002; 223(3):414-8. doi: 10.1002/dvdy.10067 [Crossref] [ Google Scholar]
95. Matsushita M, Kitoh H, Kaneko H, Mishima K, Kadono I, Ishiguro N. A novel SOX9 H169Q mutation in a family with overlapping phenotype of mild campomelic dysplasia and small patella syndrome. Am J Med Genet A 2013; 161A(10):2528-34. doi: 10.1002/ajmg.a.36134 [Crossref] [ Google Scholar]
96. Sulkala M, Wahlgren J, Larmas M, Sorsa T, Teronen O, Salo T. The effects of MMP inhibitors on human salivary MMP activity and caries progression in rats. J Dent Res 2001; 80(6):1545-9. doi: 10.1177/00220345010800061301 [Crossref] [ Google Scholar]
97. Taherkhani A, Orangi A, Moradkhani S, Khamverdi Z. Molecular docking analysis of flavonoid compounds with matrix metalloproteinase-8 for the identification of potential effective inhibitors. Lett Drug Des Discov 2021; 18(1):16-45. doi: 10.2174/1570180817999200831094703 [Crossref] [ Google Scholar]
98. Wang X, Wang X, Qin M. [A preliminary study of saliva matrix metalloproteinases (MMP-2 and MMP-9) in children with caries]. Beijing Da Xue Xue Bao Yi Xue Ban 2018;50(3):527-31. [Chinese].
99. Lee SH, Corry DB. Homing alone? CD18 in infectious and allergic disease. Trends Mol Med 2004; 10(6):258-62. doi: 10.1016/j.molmed.2004.04.002 [Crossref] [ Google Scholar]
100. Tan SM. The leucocyte β2 (CD18) integrins: the structure, functional regulation and signalling properties. Biosci Rep 2012; 32(3):241-69. doi: 10.1042/bsr20110101 [Crossref] [ Google Scholar]
101. Zhang X, Dong Y, Zhao M, Ding L, Yang X, Jing Y. ITGB2-mediated metabolic switch in CAFs promotes OSCC proliferation by oxidation of NADH in mitochondrial oxidative phosphorylation system. Theranostics 2020; 10(26):12044-59. doi: 10.7150/thno.47901 [Crossref] [ Google Scholar]
102. Zahid H, Olson NM, Pomerantz WCK. Opportunity knocks for uncovering the new function of an understudied nucleosome remodeling complex member, the bromodomain PHD finger transcription factor, BPTF. Curr Opin Chem Biol 2021; 63:57-67. doi: 10.1016/j.cbpa.2021.02.003 [Crossref] [ Google Scholar]
103. Bi J, Liu H, Cai Z, Dong W, Jiang N, Yang M. Circ-BPTF promotes bladder cancer progression and recurrence through the miR-31-5p/RAB27A axis. Aging (Albany NY) 2018; 10(8):1964-76. doi: 10.18632/aging.101520 [Crossref] [ Google Scholar]
104. Zhao X, Zheng F, Li Y, Hao J, Tang Z, Tian C. BPTF promotes hepatocellular carcinoma growth by modulating hTERT signaling and cancer stem cell traits. Redox Biol 2019; 20:427-41. doi: 10.1016/j.redox.2018.10.018 [Crossref] [ Google Scholar]
105. Green AL, DeSisto J, Flannery P, Lemma R, Knox A, Lemieux M. BPTF regulates growth of adult and pediatric high-grade glioma through the MYC pathway. Oncogene 2020; 39(11):2305-27. doi: 10.1038/s41388-019-1125-7 [Crossref] [ Google Scholar]
106. Dai M, Hu S, Liu CF, Jiang L, Yu W, Li ZL. BPTF cooperates with p50 NF-κB to promote COX-2 expression and tumor cell growth in lung cancer. Am J Transl Res 2019; 11(12):7398-409. [ Google Scholar]
107. Rogers JE, Li F, Coatney DD, Otremba J, Kriegl JM, Protter TA. A p38 mitogen-activated protein kinase inhibitor arrests active alveolar bone loss in a rat periodontitis model. J Periodontol 2007; 78(10):1992-8. doi: 10.1902/jop.2007.070101 [Crossref] [ Google Scholar]
108. Lee HJ, Kang IK, Chung CP, Choi SM. The subgingival microflora and gingival crevicular fluid cytokines in refractory periodontitis. J Clin Periodontol 1995; 22(11):885-90. doi: 10.1111/j.1600-051x.1995.tb01788.x [Crossref] [ Google Scholar]
109. Assuma R, Oates T, Cochran D, Amar S, Graves DT. IL-1 and TNF antagonists inhibit the inflammatory response and bone loss in experimental periodontitis. J Immunol 1998; 160(1):403-9. [ Google Scholar]
110. Graves DT, Delima AJ, Assuma R, Amar S, Oates T, Cochran D. Interleukin-1 and tumor necrosis factor antagonists inhibit the progression of inflammatory cell infiltration toward alveolar bone in experimental periodontitis. J Periodontol 1998; 69(12):1419-25. doi: 10.1902/jop.1998.69.12.1419 [Crossref] [ Google Scholar]