Introduction

Exosomes are nano-sized lipid vesicles which are formed by the inward budding of multivesicular bodies (MVBs) in most mammalian cells and are released into the extracellular milieu by the fusion of MVBs with the plasma membrane [1]. They contain diverse cellular materials, including proteins, lipids, and nucleic acids, and functionally transfer those maternal materials to recipient cells as a means of communication between distal cells [1]. Recently, emerging evidence has demonstrated that exosomes carry distinctive sets of RNAs, including small non-coding RNAs, tRNAs, and microRNAs (miRNAs) [2]. The transfer of exosomal miRNAs results in the direct regulation of expression of a subset of genes in recipient cells, thereby mediating a variety of cellular responses [3]. One of the most important features of exosomes is that they are found in most body fluids, including blood, urine, breast milk, and saliva and have received attention due to their potential for use as diagnostic biomarkers and as candidates for developing genetic drugs [4]. Among body fluids, saliva is practically optimized for liquid biopsy. Collecting saliva is simple and non-invasive and is undeterred by place and time. Recently, miRNAs in salivary exosomes were reported as potential diagnostic biomarkers for pancreatic cancer [5], and Saito et al. demonstrated the differential expression of exosomal miRNAs in gingival crevicular fluid of periodontitis [6], indicating the potential role of salivary exosomal miRNAs in the pathogenesis of periodontitis.

Gingivitis, a mild form of periodontal disease and periodontitis are most prevalent inflammatory diseases in periodontium. Chronic inflammation in periodontium results in loss of connective and alveolar bone tissue, which eventually causes tooth loss. Direct infection by bacteria and chronic exposure to bacterial metabolites such as lipopolysaccharides lead to immune-inflammatory responses, mediating serial pathogenesis of gingivitis [7]. Chronic gingivitis and periodontitis are usually asymptomatic except for mild bleeding from gums during tooth brushing. Although various biomarkers have been developed for predicting inflammation in periodontium, clinical diagnosis of periodontal diseases is still based on visual assessment, including radiography of periodontium and probing the pocket depth of sulcus. Although an alternative and emerging diagnostic approach by profiling salivary exosome-derived miRNAs has been implicated, pathological mechanisms of salivary exosomal miRNAs in periodontal diseases is not well understood. In this study, we analyzed the expression profile of salivary exosomal miRNAs divided according to the severity of periodontitis based on single-case research design. Furthermore, the biological function of the miRNA candidates (hsa-miR-4487, -4532, and -7108-5p) in human gingival fibroblasts was examined. Our findings suggest that differentially expressed salivary exosomal miRNAs are involved in the pathogenic responses in gingiva and could be potential diagnostic biomarkers for periodontitis.

Materials and Methods

Clinical subject and ethics statement

A 62-year-old male patient with periodontitis, who visited the Department of Periodontics at the Wonkwang University Dental Hospital, was informed about the purpose of the current study and gave written consent for participation. Diagnosis of periodontitis was determined based on the evaluation of pocket depth (PD), bleeding on probing (BOP), and bone resorption. The study was approved by the Institutional Review Board of the Wonkwang University Dental Hospital (IRB number: WKDIRB 201708-01) and was performed in accordance with relevant guidelines and regulations.

Saliva collection and isolation of salivary exosomes

Unstimulated whole saliva was collected just prior to clinical treatment and collected samples in each phase were divided by the severity of periodontitis; among them, clinically severe phase (referred to as A1) and mild phase (referred to as A3) were compared. To obtain cell-free saliva, the sample was centrifuged at 2600 ×g for 15 min at 4°C and the supernatant was collected. Each sample was treated with RNase inhibitor (Superase-In, Ambion Inc., Austin, TX, USA) and total RNA, including microRNAs was, isolated using the mirVana miRNA isolation kit (ThermoFisher, MA, USA).

Reagents

Dulbecco’s modified Eagle’s medium (DMEM), Opti-MEM alpha, and fetal bovine serum (FBS) were obtained from Gibco (Carlsbad, CA, USA). Total exosome isolation reagent was purchased from Invitrogen (Foster city, CA, USA). Antibodies for ERK1/2, p-ERK1/2, JNK, p-JNK, p38, p-p38, and β-actin were purchased from Cell Signaling Technology (Danvers, MA, USA). miRNA mimics (hsa-miR-4487, -4532, and -7108-5p) were obtained from ThermoFisher Scientific (mirVana Mimics & Inhibitors, Eugene, OR, USA).

Cell culture and transfection of miRNA mimics

Human gingival fibroblast (HGF) cells (ATCC, Manassas, VA, USA) were cultured in DMEM supplemented with 10% FBS and 1% antibiotics, in an incubator maintained at 37°C with 5% CO2. For delivery of miRNA mimics, cells were seeded in 35 mm culture dishes at approximately 70-80% confluency. Next day, miRNA mimics were transfected into cells using the Lipofectamine RNAiMAX reagent (Invitrogen, Foster city, CA, USA) according to the manufacturer’s instructions. After an additional incubation for 3 days, cells were used for RT-PCR and western blotting analyses.

Affymetrix miRNA arrays

RNA purity and integrity were evaluated by ND-1000 Spectrophotometer (NanoDrop, Wilmington, USA) and Agilent 2100 Bioanalyzer, respectively (Agilent Technologies, Palo Alto, USA). The Affymetrix Genechip miRNA 4.0 array was used according to the manufacturer's protocol. RNA samples (7.56 ng) were labeled with the FlashTag™ Biotin RNA Labeling Kit (Genisphere, Hatfield, PA, USA). The labeled RNA was quantified, fractionated and hybridized onto the miRNA microarray according to the standard procedures provided by the manufacturer. The labeled RNA was heated to 99°C for 5 minutes and then at 45°C for 5 minutes. RNA-array hybridization was performed with agitation at 60 rotations per minute for 16 hours at 48°C on an Affymetrix® 450 Fluidics Station. The chips were washed and stained using a Genechip Fluidics Station 450 (Affymetrix, Santa Clara, California, United States). The chips were then scanned using an Affymetrix GCS 3000 scanner (Affymetrix, Santa Clara, California, United States). Signal values were computed using the Affymetrix® GeneChip™ Command Console software. Raw data were extracted automatically using Affymetrix data extraction protocol provided by Affymetrix GeneChip® Command Console® Software (AGCC). The CEL files import, miRNA level RMA+DABG-All analysis and results were exported using Affymetrix® Power Tools (APT) Software. Array data were filtered by probes annotated species. The comparative analysis between test sample and control sample was carried out using fold change. All statistical tests and visualization of differentially expressed genes were conducted using R statistical language 3.0.2.

MicroRNA target prediction and pathway analysis

DIANA Tools (DIANA-mirPath V.3; http://microrna.gr/ mirpath) was employed for the enrichment analysis of predicted target genes and relevant signaling pathways based on the KEGG pathways and microT-CDS algorithm [8]. The p values to the identified pathways was calculated by the mirPath.

Reverse-transcription polymerase chain-reaction (PCR)

Total RNA from HGF cells was extracted using Trizol reagent (Invitrogen, CA, USA) according to the manufacturer’s protocol. cDNA was then synthesized using PrimeScript RT reagent kit (Takara, Shiga, Japan). The following primers were used for PCR amplification: IL-6: forward primer 5'-GATTCC AAAGATGTAGCCGCCC-3', reverse primer 5'-GCTGGACTG CAGGAACTCCTTA-3'; IL-8; forward primer 5'-TCTTGGC AGCCTTCCTGATTTCTG-3', reverse primer 5'-AGTTTTCC TTGG GGTCCAGACAG-3'; GAPDH: forward primer 5'-AGG GCTGCTTTTAACTCTGGT-3', reverse primer 5'- CCCCAC TTGATTTTGGAGGGA-3'.

Western blot

Cells were lysed using RIPA buffer (Pierce Biotechnology, Rockford, IL, USA) containing protease and phosphatase inhibitors. Extracted proteins were quantified using the BCA assay, and 10 μg of total protein was subjected to SDS-PAGE. Separated proteins were transferred onto a PVDF membrane for 1 h. The membrane was then incubated with primary antibodies, including anti-ERK1/2, p-ERK1/2, p38, p-p38, JNK, p-JNK, and β-actin at a dilution of 1:1000, overnight at 4°C. After incubation with HRP-conjugated IgG (secondary antibody) (1:3000) for 1 hr at RT, immunoreactive proteins were detected using enhanced chemiluminescence.

Statistical analysis

Results were analyzed using the SPSS software version 14.0 (SPSS Inc., Chicago, IL, USA) and data were presented as mean ± SEM of the stated number of observations obtained from more than three experiments. Statistical differences were analyzed through one-way ANOVA test followed by Tukey’s post hoc test. Values of p < 0.05 were considered as statistically significant (**p < 0.05).

Results

Clinical characteristics

At the time of referral, the patient reported having suffered from chronic periodontitis in the past and was still receiving supportive periodontal maintenance. He had suffered from gingival pain, bleeding, and swelling in #24 tooth and periodontal examination revealed a periodontal pocket depth (PD) of 10, 4, 4 / 8, 3, 3 mm; positive (+) bleeding on probing; and positive (+++) tooth mobility (Table 1). A deep alveolar defect was observed in the medial surface in the periapical radiographic view. The medical history did not mention smoking, alcohol, medication, or systemic diseases except mild, high blood pressure. The details of the clinical characteristics of the subject enrolled in the current study are presented in Table 1. Prior to periodontal therapy, a sample of unstimulated whole saliva was collected during the follow-up, and saliva collected in clinically severe phase was referred to as A1. After sampling of approximately 1 mL of whole saliva, he was treated with subgingival root planing and curettage and was prescribed antibiotics and analgesics for 5 days. Upon his next visit 2 months later for regular examination, oral hygiene status was shown to be clinically stable, and saliva collected in mild phase referred to as A3. The pocket depth was 5, 3, 3 / 5, 3, 3 mm and bleeding on probing was negative (Table 1).

Characteristics of differentially expressed miRNAs in salivary exosomes

Exosomes were isolated from each saliva sample and an adequate amount of exosomal RNA including miRNAs was used to perform microarray using Affymetrix miRNA arrays. A set of miRNA microarray profiles identified 256 mature miRNAs present in salivary exosomes, and in all, 13 and 43 of these 256 miRNAs showed more than 2-fold difference (|FC| ≥ 2) in A1 and A3 phases, respectively (Table 2). Particularly, miRNAs hsa-miR-4532 (|FC| = 18.322697) and -4487 (|FC| = 9.292005) were markedly increased in the A1 phase, whereas hsa-miR-7108-5p (|FC| = 12.20601) and -6800-5p (|FC| = 15.25109) were up-regulated in the A3 phase. By utilizing the DIANA-mirPath tool based on microT-CDS target prediction algorithm [9], we found that KEGG pathways (p < 0.05, FDR corrected) including the ErbB signaling pathway, tyrosine metabolism, central carbon metabolism, transcriptional misregulation in cancer, endocytosis, prostate cancer, biosynthesis of unsaturated fatty acids, pancreatic cancer, estrogen signaling pathway, and glioma were enriched among these 4 miRNAs (Table 3). Notably, the endocytosis KEGG pathway was commonly enriched among these 4 miRNAs (Figure 1). A subset of 18 genes was predicted as the target of these 4 miRNAs (Figure 2).

Overexpression of miRNA mimics (hsa-miR-4487, -4532, and 7108-5p) in human gingival fibroblasts results in induction of IL-6 expression and p38 activation

According to the clinical diagnosis of recurring inflammation in gingiva, we assumed that differentially expressed exosomal miRNAs in saliva somehow determine inflammatory responses in gingival fibroblasts. To confirm this, we examined the effects of hsa-miR-4487, -4532, and 7108-5p on the induction of pro-inflammatory cytokines, such as IL-6, IL-8, IL-1β, and TNF-α. Mimics of three miRNAs were delivered into human gingival fibroblasts and mRNA transcript of each gene was evaluated by RT-PCR. Interestingly, overexpression of miRNAs significantly induced only IL-6 expression (Figure 3A), but not IL-1β and TNFα (data not shown). Expression of IL-8 showed a slightly increasing tendency without significance (Figure 3B). During pathogenesis of lipopolysaccharides (LPS) derived from microbes, it is well known that IL-6, a pro-inflammatory cytokine, is induced by the activation of MAP kinases including ERK, p38, and JNK [10]. Considering this, we measured the phosphorylation of these MAP kinases by western blotting. Results showed that ERK and JNK were not affected by overexpression of 3 miRNAs, whereas phosphorylation of p38 was significantly up-regulated by miRNA mimics (Figure 3C). These results suggest that these exosomal miRNAs in saliva can self-induce inflammatory responses in gingiva.

Discussion

Periodontitis is characterized by chronic inflammatory responses in the periodontium, caused by multiple pathogenic factors [11]. Although bacterial plaque has been known as the primary etiologic factor, epidemiologic evidence indicates that other risk factors including tobacco use, diabetes, age, and composition of saliva are closely associated with the pathogenesis of periodontitis [12, 13]. Recently, emerging studies on extracellular vesicles, also known as exosomes, revealed that exosomes in saliva, and its contents such as miRNAs, are involved in maintaining oral homeostasis [14]. Exosomal miRNAs are now widely accepted as promising biomarkers for the diagnosis and treatment of oral diseases. For instance, the exosomal miRNA miR-4484 was strongly up-regulated in saliva from patients with oral lichen planus [15], and distinctive expression patterns of 40 miRNAs, including miR-223-3p, -205-5p and -203a, were confirmed in gingival crevicular fluid of periodontitis patients [6]. These strategies for discovering candidate miRNAs from human saliva have been generally based on a group research design that is limited due to the requirement of a considerably large group of similar cases and normal subjects as control. Besides, results obtained by comparison between groups easily overlooks the clinical differences between individual patients. To minimize the differences between individuals, alternative research design, namely single-case experimental design (SCED), has been applied in the field of psychology and education. SCEDs are useful studies to demonstrate clinical phenomena of a single or a small group of subjects by validating variation differences over time [16].

In this study, as a pilot experiment, we applied SCED to validate the differences in exosomal miRNAs between saliva samples. Comparing the expression of salivary exosomal miRNAs with severity, variation differences in 4 miRNAs (hsa-miR-4487, -4532, -7108-5p, and -6800-5p) were observed. Unfortunately, very little is known about the biological functions of these 4 miRNAs except that miR-4487 is known as a negative regulator of Unc-51 like kinase 1 (ULK1) involved in cell autophagy [17] and miR-4532 might target potassium voltage-gated channel J11 (KCNJ11) and its variant (rs60432575 G>A) appears to attenuate the down-regulation of KCNJ11 expression [18]. However, prediction of target genes and pathway analysis indicates that the endocytosis pathway is commonly enriched among these 4 miRNAs. It is well known that biogenesis and release of exosomes are associated with endosomal trafficking [19]. Especially, Rab5b, predicted as the target gene of hsa-miR-4487, is well known as an exosomal marker gene [20] and EGFR, predicted as the target gene of hsa-miR-7108-5p, is involved in exosome production and wound healing [21]. In agreement with these reports, our findings suggest that delivery of salivary exosomes and its contents to periodontal tissue cells can affect diverse cellular responses and can be considered as prominent biomarkers for periodontitis.

The main feature of this clinical case was periodic recurrence of inflammatory responses in periodontal tissues despite the patient’s efforts in managing oral health. Based on the hypothesis that salivary exosomal miRNAs might affect and cause the inflammatory responses in gingiva, we examined the biological effects of candidate miRNAs (hsa-miR-4487, 4532, and 7108-5p) on inflammatory responses in human gingival fibroblasts. Intriguingly, overexpression of miRNA mimics in HGFs significantly induced only IL-6 expression, but not IL-8, IL-1β, and TNF-α. Lipopolysaccharides (LPS) released by periodontal pathogenic bacteria mediate the induction of pro-inflammatory cytokines IL-6, IL-8, IL-1β, and TNFα through the activation of several signaling pathways, one of them being the MAPK pathway [22]. Meanwhile, our data indicated that overexpression of miRNA mimics in gingival fibroblasts resulted in increased induction of IL-6 expression, and these miRNAs except miR-4532 led to activation of p38 kinase. IL-6 is a multifunctional cytokine and known to mediate diverse cellular responses, including proliferation of T-lymphocytes, secretion of immunoglobulin, and differentiation of Blymphocytes and osteoclasts [23, 24]. Considering these, our results suggest that up-regulation of miRNAs (hsa-miR-4487, 4532, and 7108-5p) in salivary exosomes might be closely related to recurrence of inflammation and bone loss in periodontal tissues, in this patient.

In conclusion, the present results demonstrated that hasmiR- 4487, -4532, and -7108-5p in salivary exosomes can be regarded as useful biomarkers for periodontitis, especially during periodic recurrence of inflammation and bone loss in periodontium. In addition, these miRNAs could be target molecules to attenuate the induction of IL-6 and activation of p38 kinase.