Therapeutic potency and the related mechanism of deinoxanthin in experimental animal and cell models of periodontitis

Abstract

Periodontitis is a chronic inflammatory disease that leads to the destruction of periodontal tissue, ultimately resulting in tooth loss. Studies have aimed to develop biomaterials that effectively prevent inflammatory responses and oxidative stress in the progression of periodontitis, without adverse side effects. Here, we explored whether Deinococcus radiodurans-derived deinoxanthin (DEIX) protects against alveolar bone loss and connective tissue degradation in an experimental rat model of periodontitis and investigated the related mechanisms using human-derived periodontal ligament cells (hPDLCs) and THP-1 cells. Oral supplementation with DEIX (25 mg/kg body weight, once per day for 14 consecutive days) protected rats against ligature-mediated periodontal destruction. That protection involved the DEIX-induced restoration of the ligature-stimulated disorders, including overproduction of inflammatory mediators, accumulation of reactive oxygen species, and imbalance between osteoclast and osteoblast activity in the inflamed periodontium. In vitro experiments supported the associated mechanisms by which the direct addition of DEIX (20 µM) recovers lipopolysaccharide (LPS, 2 µg/mL)-stimulated inflammatory responses in hPDLCs and THP-1 cells. RNA sequence profiling from the DEIX and/or LPS-exposed hPDLCs further supported the protective mechanisms of DEIX on LPS-stimulated inflammatory and oxidative damage. Collectively, this study highlights the potential of DEIX to protect against inflammatory periodontal tissue destruction and demonstrates its clinical utility for patients with chronic periodontitis.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-36514-1.

Introduction

Periodontitis is a prevalent immune-inflammatory disorder that stimulates inflammatory cells to secrete proinflammatory cytokines, ultimately leading to the destruction of connective tissues, including the gingiva, periodontal ligament (PDL), and alveolar bone¹. The inflammation in periodontitis is triggered by oral bacteria and their byproducts, as well as by the host’s immune response to these pathogens². Generally, a local and systemic inflammatory immune response contributes to the induction and accumulation of reactive oxygen species (ROS) in cells³. Specifically, the inflammatory process in periodontitis can produce a large amount of ROS in a prolonged and persistent manner, and that production triggers oxidative damage in cells and tissues of the periodontium⁴. Regarding this, the production of inflammatory cytokines and the accumulation of intracellular ROS might be the main events occurring in the process of periodontitis. Alternatively, this suggests that the use of anti-inflammatory and antioxidant molecules is a promising approach to mitigate and/or protect against alveolar bone loss and degenerative tissue damage in patients with periodontitis^5–7.

As dysregulated host immune responses play a vital role in exacerbating inflammation during the progression of periodontitis, many investigators have made efforts in developing anti-inflammatory drugs or inhibitors specific to matrix metalloproteinases (MMPs) and proinflammatory cytokines that are closely associated with periodontal degenerative diseases^8–10. However, prolonged exposure to and/or persistent use of antibiotics may contribute to the host’s immune resistance to various antibiotics. The use of several commercial drugs is also known to cause side effects to patients¹¹. Therefore, studies have explored whether naturally occurring antioxidants exert beneficial effects using experimental animal models of periodontitis or cells exposed to inflammatory mediators without any side effects. Regarding this, numerous studies have indicated that natural antioxidants can exhibit various biological and pharmacological properties, protecting against periodontitis-related tissue degradation^12–14. Studies also indicate that the beneficial effects of antioxidants on periodontitis are closely associated with their ability to diminish the production of proinflammatory cytokines and intracellular ROS and recover an imbalanced activation between osteoclasts and osteoblasts. Accordingly, an antioxidant-based therapy is valuable in controlling the inflammatory process and maintaining periodontal structure from periodontitis-induced degenerative damage.

Carotenoids are potent antioxidants that have biological, medicinal, and pharmacological activities¹⁵. Deinoxanthin (DEIX; (2R)-2,1′-dihydroxy-3′,4′-didehydro-1′,2′-dihydro-β,ψ-caroten-4-one) is a unique xanthophyll carotenoid synthesized by Deinococcus microorganisms¹⁶. DEIX was first characterized from Deinococcus radiodurans (D. radiodurans) and showed more substantial antioxidant potential in scavenging hydrogen peroxide and singlet oxygen than other xanthophyll carotenoids such as carotene, lutein, lycopene, and zeaxanthin¹⁷. Its great antioxidant potential is associated with a unique structural feature¹⁶. These reports indicate that DEIX can prevent or attenuate periodontitis-mediated inflammatory and oxidative damage. However, the protective effects of DEIX on periodontal inflammation, bone metabolism, and immune-oxidative pathways have not previously been investigated.

Here, we examined whether oral supplementation with DEIX protects against periodontal tissue destruction in an experimental rat model of periodontitis, along with the associated mechanisms. To address this, we administered DEIX orally to rats once daily for 14 consecutive days, starting seven days after ligatures were placed into the gingiva of the right maxillary second molar. The ligatures were sub-cultured with Porphyromonas (P.) gingivalis overnight before their placement on the gingiva. To verify the mechanisms involved in DEIX-induced protection against periodontitis, we evaluated the direct effect of DEIX using human-originated PDL cells (hPDLCs) and a human monocytic cell line, THP-1 cells, in the presence and absence of lipopolysaccharide (LPS). This study demonstrates the protective potential and related mechanisms of DEIX on periodontitis-related periodontal degradation for the first time. Our results from in vitro biological assays and RNA sequence profiling support the molecular mechanisms by which DEIX recovers inflammatory and oxidative periodontal destruction. Overall, our findings suggest that DEIX is a potent anti-inflammatory and antioxidant that inhibits the initiation and progression of periodontitis, while also promoting the recovery from degenerative periodontal destruction.

Materials and methods

Chemicals and laboratory equipment

We produced DEIX from D. radiodurans as described previously¹⁸ and used it in all experiments of this study. Ligature was purchased from TP Orthodontics, Inc. (Seoul, Republic of Korea), and P. gingivalis-produced LPS was obtained from InvivoGen (San Diego, CA, USA). Toll-like receptor 4 (TLR4) antibody was purchased from Novus Biologicals (Centennial, CO, USA). Antibodies specific to bone morphogenetic protein-2 (BMP2; BS90141), cyclooxygenase-2 (COX-2; BS1076), nuclear factor erythroid 2-related factor 2 (Nrf2; BS1258), osteocalcin (OCN; BS7961), runt-related transcription factor 2 (RUNX2; BS2831), and tumor necrosis factor-α (TNF-α; BS6432) were purchased from Bioworld Technology, Inc. (St. Louis Park, MN, USA). Heme oxygenase-1 (HO-1; ab13248), Ki-67 (ab833), osterix (ab209484), osteopontin (OPN; ab8448), and receptor activator of nuclear factor (NF)-κB ligand (RANKL; ab9957) antibodies were obtained from Abcam (Cambridge, UK). Cathepsin K (sc-4835), β-actin (sc-47778), c-Fos (sc-166940), matrix metalloproteinase-9 (MMP-9; sc-393859), and nuclear factor of activated T-cells, cytoplasmic 1 (NFATc1; sc-7294) antibodies were provided by Santa Cruz Biotechnology (Santa Cruz, CA, USA). While RANKL (ALX-804-243) was purchased from Enzo Life Sciences, Inc. (Farmingdale, NY, USA), γ-H₂AX (ab26350) and 2’,7’-dichlorodihydrofluorescein-diacetate (DCF-DA; ab273640) were from Abcam. Fetal bovine serum (FBS) and antibiotic-antimycotic (2441713) were purchased from HyClone Laboratories (Logan, UT, USA) and Gibco (Life Technologies, Carlsbad, CA, USA), respectively. Unless specified otherwise, other chemicals and laboratory consumables were purchased from Sigma-Aldrich Co. LLC (St. Louis, MO, USA), Falcon Labware (Becton-Dickinson Biosciences, Franklin Lakes, NJ, USA), or SPL Life Sciences (Pocheon, Republic of Korea).

Animal and ethical statement

Male Sprague-Dawley rats (7 weeks old, n = 16) were purchased from Damul Science (Daejeon, Republic of Korea) and equilibrated for seven days before use. During the experimental periods, all animals were housed at 22 ± 1 °C, 55% ± 5% humidity, and a 12-h light/dark cycle with ad libitum access to food in the Animal Center of Jeonbuk National University School of Dentistry. This study was carried out in accordance with the recommendations in the Animal Care and Use Guide of Jeonbuk National University and ARRIVE guidelines 2.0 (https://arriveguidelines.org). The University Committee on Ethics approved all experimental procedures in the Care and Use of Laboratory Animals (NON2024-043-003).

Treatment of ligatures withP. gingivalis

P. gingivalis (W50 wild-type strain) was grown on 5% Sheep Blood Agar Media Plate (HiMedia Laboratories LLC, Kennett Square, PA, USA) in a Don Whitley MACS MG-500 Anaerobic Chamber (Akribis Scientific Supplies Ltd., Manchester, UK) with 80% N₂, 10% H₂, and 10% CO₂ at 37 °C. A single colony of P. gingivalis was isolated from the plate and inoculated into 10 mL of brain heart infusion broth media supplemented with 5 mg/L hemin. After 24 h of incubation, the seed culture was mixed with 90 mL of the same fresh medium, followed by an additional six days of incubation. Before the ligature treatment with the bacteria, the number of P. gingivalis in cultures was determined by comparing the optical density to a curve derived from a standard plate count. Ligatures were soaked in the subcultures of P. gingivalis (1 × 10⁹/mL) overnight before placing them into the experimental rats’ right maxillary second molar.

Anesthesia and periodontitis induction

Rats were randomly divided into four groups (n = 4/group), including control (no periodontitis), DEIX, ligature, and ligature/DEIX groups. Rats received a general anesthesia via an intraperitoneal injection of Zoletil (0.4 mL/kg, Virbac Laboratories, Carros, France) mixed with Rompun (10 mg/kg, Bayer Korea Ltd., Seoul, Republic of Korea). In the ligature and ligature/DEIX groups, periodontitis was induced by positioning ligatures on the right maxillary second molar for seven days. The rats were monitored once a day, and any lost or loose ligatures were replaced with new ones. Ligatures were removed from the rats at one week post-induction of periodontitis.

Oral DEIX supplementation and endpoint procedures

Immediately after the removal of ligatures, the ligature/DEIX and DEIX groups were orally supplemented with olive oil (vehicle, 200 µL) containing DEIX at a final concentration of 25 mg/kg body weight once per day for two weeks. Control and ligature groups received vehicle only for the same periods. After the final supplementation with DEIX, all groups of rats were euthanized with CO₂ in a chamber, following the guidelines for the Euthanasia of Animals (2020), the American Veterinary Medical Association (AVMA). Whole blood and periodontal tissues, including the inflamed regions, were collected for further analyses.

Micro-computed tomography (µCT) and bone parameter analyses

The µCT imaging was performed using a desktop scanner (SkyScan 1276, Bruker, Billerica, MA, USA) with software that included NRecon reconstruction, CTAn 1.8, and CTvol. Conditions were set at a maximum voltage of 70 kV and a current of 200 µA, with a 1-mm filter, and a 180° tomographic rotation (0.6° rotation step). Images were obtained at a pixel resolution of 20 μm, and data were analyzed using the SkyScan NRecon reconstruction package (Data Viewer, Bruker-µCT-Analyzer version 1.13, and CT Vol). A global thresholding algorithm was used at a uniform threshold. Based on the constructed 3D images, bone mineral density (BMD; g/cm³) and bone volume/tissue volume (BV/TV; %) were calculated for the newly formed alveolar bones around the second upper molar. In BMD measurements, attenuation data for VOI were converted to Hounsfield units and expressed as a value of BMD using phantoms (SkyScan). These phantoms consisted of rods of calcium hydroxyapatite (CaHA) with a standard density corresponding to that of rat bone. BMD value is expressed in grams per cubic centimeter of CaHA in distilled water. A zero value for BMD corresponds to the density of distilled water alone (no additional CaHA), and a value greater than zero corresponds to non-aerated biological tissue.

Tissue section and histological analyses

The right upper maxilla was isolated and fixed in a 4% paraformaldehyde solution for 48 h and then decalcified in 17% EDTA at 4 °C for four weeks. The decalcified specimens were dehydrated, embedded in paraffin, and sectioned at a thickness of 5.0 μm. For hematoxylin & eosin (H&E) staining, tissue sections were treated with Gill No. 3 hematoxylin after de-paraffinization and rehydration, followed by counterstaining with 0.25% eosin. The distance between the cementoenamel junction (CEJ) and the alveolar bone crest (ABC) in the interproximal region of the first and second molars was measured using a soft imaging system (analySIS^®, Münster, Germany). The average distance measured along the root of molars (M1 and M2) was also plotted. Some tissue sections were subjected to tartrate-resistant acid phosphatase (TRAP) staining using a leukocyte acid phosphatase kit (AK04F, Cosmo Bio Co. Ltd., Tokyo, Japan) followed by counterstaining with hematoxylin. Tissue sections were also processed for immunohistochemistry (IHC) using an IHC accessory kit (PK-6101, Vector Laboratories, Burlingame, CA, USA). In the IHC assay, anti-osterix, -BMP2, -Nrf2, and -HO-1 antibodies were applied at 1:200–400 dilutions. All procedures for TRAP and IHC staining were performed according to the manufacturer’s instructions. The stained tissue samples were photographed using a Motic EasyScan One and Motic DSAssistant (Kowloon, Hong Kong). The intensity of positive staining was analyzed using the ImageJ software program (Version 1.51, NIH, Bethesda, MD, USA). For the immunofluorescence (IF) assay, tissue sections were fixed in 4% PFA solution and washed three times with phosphate-buffered saline (PBS). Sections were incubated with 0.25% Triton X-100 in PBS for 10 min and washed with PBS containing 0.05% Tween^® detergent in PBS (PBS-T). Sections were further incubated for 30 min in 50 mL of PBS-T solution containing 0.5 g of bovine serum albumin (BSA) and 0.375 g of glycine. After washing and blocking processes, the tissue samples were exposed to anti-RANKL, -cathepsin K, -RUNX2, or -OPN antibody, followed by incubation with secondary Alexa Fluor 488-conjugated anti-goat IgG, Alexa Fluor 594-conjugated anti-rabbit IgG, or 488-conjugated anti-mouse IgG antibody. All samples were counterstained with aqueous mounting medium (Santa Cruz Biotechnology) containing 4′,6-diamidino-2-phenylindole (DAPI) at room temperature for 10 min. The expression patterns of antibody-specific molecules were evaluated by confocal laser scanning microscopy (CLSM) (LSM 880 with Airyscan, Carl Zeiss, Ostalbkreis, Germany). The ImageJ software also quantified the antibody-specific fluorescence intensities (a.u.).

Enzyme-linked immunosorbent assay (ELISA)

Serum was obtained from whole blood samples collected from the rat groups 14 days after the induction of periodontitis. Blood samples were centrifuged at 1500×g in serum-separating tubes. The levels of TNF-α and interleukin (IL)-1β in the sera were measured using mouse-anti-TNF-α (#MTA00B, Biotechne R&D systems, Minneapolis, MN, USA) and IL-1β ELISA kits (#MLB00C) in a microplate reader (SPECTROstar^® Nano, BMG LABTECH, Ortenberg, Germany) according to the manufacturers’ instructions.

Cell culture and treatment with LPS, DEIX, or both

To evaluate the direct effects of DEIX on LPS-stimulated cellular responses, we cultured hPDLCs and THP-1 cells. This study obtained hPDLCs from healthy male patients (18–25 years old) who underwent tooth extraction before orthodontic treatment at Jeonbuk National University Hospital of Dentistry (Jeonju, Republic of Korea). All donors provided written informed consent for the use of their periodontal tissue, and the Ethical Committee of Jeonbuk National University Hospital approved the use of that sample. The cultures of hPDLCs and THP-1 cells were maintained in α-Minimum Essential Medium (αMEM) and Roswell Park Memorial Institute (RPMI) 1640 medium, respectively, in the presence of 10% FBS and antibiotics (100 IU/mL penicillin G and 100 µg/mL streptomycin) at 37 °C in a humidified atmosphere of 5% CO₂. Cells were divided onto 96-well, 24-well, 6-well, or 60 mm culture plates at an appropriate number per well. The media of the cultures were replaced with 0.5% FBS-supplemented αMEM or RPMI 1640 medium 16 h before exposure to LPS (2 µg/mL) for the analysis of ROS and inflammatory cytokines. Cells were treated with varied concentrations (0–50 µM) of DEIX dissolved in dimethyl sulfoxide (DMSO) 1 h before the LPS exposure. The amount of DMSO did not exceed 0.02% in the culture medium, and cells treated with 0.02% DMSO alone served as the vehicle control. Cells were processed for further analysis at various time points during treatment with LPS and/or DEIX.

Cell proliferation by CCK-8 and Ki-67 staining

The effect of DEIX on the proliferation of hPDLCs was assessed using a Cell Counting Kit-8 (CCK-8; LOT#DV684, Dojindo Lab, Rockville, MD, USA) according to the manufacturer’s instructions. In brief, the hPDLCs resuspended in αMEM (2.5 × 10⁴ cells/mL) were divided into 96-well culture plates (0.1 mL/well) and grown in the presence of 2% serum and antibiotics. After 12 h of incubation, the cells were exposed to various concentrations (0–50 µM) of DEIX in the presence and absence of 2 µg/mL LPS for 24 h. Cells were treated with 10 µL of CCK-8 solution followed by an additional 2 h of incubation. The optical density specific to the CCK-8 dye was then measured at 450 nm using a microplate reader (SPECTROstar^® Nano). The proliferation rate of hPDLCs was also determined using an IF assay. To this end, the hPDLCs-resuspended αMEM (5 × 10⁴ cells/mL) were spread on coverslips (0.2 mL/coverslip) that were placed on the bottom of 6-well plates. The cells were incubated for 12 h in the presence of 2% serum and antibiotics and then exposed to LPS, DEIX, or both. After an additional 24 h incubation, the cells were fixed with 4% paraformaldehyde, permeabilized with 0.25% Triton X-100 for 10 min, washed with PBS, and incubated with 1% BSA for 1 h. Cells were incubated with Ki-67 primary antibody at 4°C overnight and treated with donkey anti-rabbit secondary antibody (Alexa Fluor^® 488) for 1 h. After being washed with PBS, the cells were counterstained with DAPI for 5 min. Fluorescence images of the cells were captured using CLSM, and the ratio of KI-67-positive cells (%) was calculated.

ROS production evaluated by flow cytometry

The potency of DEIX in reducing intracellular ROS accumulation in LPS-stimulated hPDLCs was assessed by measuring the levels of DCF-positive cells using a flow cytometer. To this end, the hPDLCs-contained αMEM (1 × 10⁶ cells/mL) were divided into 60 mm culture plates (4 mL/plate) and incubated until the cells reached 80% confluence. The culture medium was changed to αMEM supplemented with 0.5% FBS, 2 µg/mL LPS, and/or 20 µM DEIX. After 24 h of incubation, the cultures were treated with 10 µM DCF-DA for 30 min, and DCF-specific green fluorescence intensity at 515 nm (FL-1 H) was recorded from 10,000 events/sample using a FACS Vantage system (Becton-Dickinson Biosciences).

Assays for wound healing and transwell migration

For the in vitro wound healing assay, hPDLCs resuspended in αMEM (2 × 10⁵ cells/mL) were divided onto 6-well culture plates (1.5 mL/well) and incubated in 10% FBS-supplied αMEM. When the cells reached approximately 90% confluence, the bottom of the culture plates was treated with a linear scratch (250 μm in width) using a pipette tip. After washing twice with PBS, cells were incubated in 0.5% FBS-supplemented αMEM containing LPS (2 µg/mL) and/or DEIX (20 µM) for 24 h. The scratched areas were photographed using a light microscope (EL-Einsatz 451888, Carl Zeiss) 24 h after the incubation. Healing of the gap was analyzed in five randomly selected fields in each culture plate, and the results are represented as the area (%) of migrated cells. According to the manufacturer’s protocol, the transwell migration assay was performed using a 24-well transwell chamber with an 8.0 μm pore polycarbonate membrane (Corning Life Sciences, Tewksbury, MA, USA). In brief, 100 µL of serum-free medium containing THP-1 cells (1 × 10⁵ cells) was seeded onto each well of the upper chamber. The same medium (600 µL) containing hPDLCs (1 × 10⁵ cells), LPS (2 µg/mL), and/or DEIX (20 µM) was added into the lower chamber as a chemoattractant. After 24 h of incubation, cells on the upper surface were removed with a cotton swab, and cells on the lower surface were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet solution. Cells were photographed with an inverted light microscope, and the number of migrated THP-1 cells was counted.

Real quantitative polymerase chain reaction (RT-qPCR) assay

The effect of DEIX on the expression of alkaline phosphatase (ALP), type I collagen (COL1A1), COX-2, HO-1, IL-8, Nrf2, OCN, periostin, RUNX2, scleraxis, sclerostin, and vimentin in LPS-stimulated hPDLCs was evaluated by RT-qPCR. To this end, hPDLCs (2 × 10⁵ cells/mL) seeded in 60 mm culture plates (4 mL/plate) were exposed to LPS (2 µg/mL) and/or DEIX (20 µM). For osteogenesis, hPDLCs were incubated in αMEM supplemented with 5% FBS in the presence and absence of DAG (100 nM dexamethasone, 50 µM ascorbic acid, and 10 mM β-glycerophosphate). At various times of incubation, total RNAs were extracted from the cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and the concentration and purity of RNAs were determined using the NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). According to the manufacturer’s instructions, complementary DNA (cDNA) was synthesized from 1 µg of total RNA/sample using an AmpiGene cDNA Synthesis Kit (LOT# 10141515; Enzo Life Sciences). The RT reaction was performed at 25 °C for 10 min, 55 °C for 60 min, and 85 °C for 5 min in a thermal cycler. PCR was performed with Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) and an ABI StepOnePlus RT-PCR System (Applied Biosystems). Each reaction mixture contained 1 µL of cDNA, 10 µM of each primer, and 5 µL of SYBR Green Master Mix. The thermocycling conditions were as follows: pre-denaturation at 95 °C for 10 min and amplification using three-step cycles of denaturation at 95 °C for 15 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s for 40 cycles. The oligonucleotide primers used are provided in Table S1. The level of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was considered as the endogenous reference during the quantification.

Immunoblot analysis

The hPDLCs (2 × 10⁵ cells/mL) seeded in 60 mm culture plates (4 mL/plate) were exposed to LPS (2 µg/mL), DEIX (20 µM), or both in the presence and absence of Nrf2 inhibitor, ML385 (5 µM). After 24 h of incubation, the cells were processed for whole-protein extraction to evaluate the protein levels of COX-2, γ-H₂AX, HO-1, Nrf2, TLR4, and TNF-α by immunoblotting. Whole protein lysates were also extracted from the hPDLCs exposed to LPS (2 µg/mL) and/or DEIX (20 µM) with and without DAG for seven days, after which the levels of BMP2, OCN, OPN, osterix, and RUNX2 proteins were determined by immunoblot assay. In addition, THP-1 cells were incubated with the media collected from the cultures of LPS (2 µg/mL) and/or DEIX (20 µM)-exposed hPDLCs for three days. After an additional four-day incubation, whole protein lysates were extracted, and the levels of cathepsin K, c-Fos, NFATc1, and MMP-9 proteins were analyzed. Protein extracts (20 µg/sample) were separated through sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 8–12% gels and electroblotted onto polyvinylidene difluoride membranes. The blots were washed with a buffer containing 10 mM Tris-HCl (pH 7.6), 150 mM NaCl, and 0.05% Tween-20, followed by blocking in 5% skim milk for 1 h. The blots were treated with primary antibodies at dilutions from 1:250 to 1:2500 before exposure to horseradish peroxidase-conjugated secondary antibodies (1:1000 to 1:5000 dilution). Immunoreactive bands were visualized using an ECL detection kit (Bio-Rad, Hercules, CA, USA) and captured on a chemiluminescence imaging system (FUSION SOLO X, Vilber, Marne-la-Vallée, France). The level of β-actin was used as the loading control.

Osteoclastic differentiation assay

THP-1 cells were seeded into 24-well culture plates (1 × 10⁴ cells/well) in the presence of 20 ng/mL phorbol 12-myristate 13-acetate (PMA) for three days to allow the differentiation into adherent macrophages. Thereafter, cells were treated with 50 ng/mL RANKL, either with or without LPS (2 µg/mL), DEIX (20 µM), or both. After five days of incubation, the cultures were fixed with 4% paraformaldehyde in PBS and stained with a TRAP staining kit (Cosmo Bio Co. Ltd.) according to the manufacturer’s instructions. The TRAP-stained cells were photographed using a light microscope, and multinucleated cells with more than five nuclei were counted as osteoclasts. The mean diameter of the osteoclasts and the average nucleus number per osteoclast were determined from the optical images.

mRNA sequencing and functional enrichment analysis

To further understand the direct effect of DEIX on LPS-stimulated inflammatory responses, mRNA sequence profiling was performed using the hPDLCs exposed to LPS (2 µg/mL), DEIX (20 µM), or both for three days. Total RNAs were isolated from the cells using Trizol reagent. RNA purity was detected using the NanoVue spectrophotometer (GE Healthcare, Piscataway, NJ, USA). The RNA sequence analysis was performed at LAS (Kimpo, Republic of Korea) following the procedures: processing, read mapping, expression quantification, differentially expressed gene (DEG) analysis, functional annotation analysis, and visualization (Figure S1). For preprocessing and genome mapping, potentially existing sequencing adapters and low-quality bases in the raw reads were trimmed using Skewer (ver 0.2.2). After trimming the low-quality bases and sequencing adapters, STAR (version 2.5 software) mapped the cleaned high-quality reads to the reference genome. Since the sequencing libraries were prepared strand-specifically using Illumina’s strand-specific library preparation kit, the strand-specific library option, --library-type = fr-first strand, was applied in the mapping process. Cuffquant in Cufflinks (ver 2.2.1) with the strand-specific library option, -library-type = fr-first strand, and other default options was used to quantify the mapped reads on the reference genome into the gene expression values. The gene annotation of the reference genome from the UCSC genome (https://genome.ucsc.edu) in GTF format was used as gene models, and the expression values were calculated in Fragments Per Kilobase of transcript per Million mapped fragments (FPKM) units. Cuffdiff analyzed the DEGs between the two selected biological conditions in the Cufflinks package with the strand-specific library option. To compare the expression profiles among the samples, the normalized expression values of the selected DEGs were clustered using in-house R scripts. The scatter plots for the gene expression values and the volcano plots for the expression-fold changes and p-values between the two selected samples were also drawn using in-house R scripts. To get the insights on the biological functional role of the differential gene expression between the compared biological conditions, gene set overlapping test between the analyzed DEGs and functional categorized genes, including biological processes of Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and other functional gene sets were performed by g: Profiler2 (ver 0.2.0).

Statistical analyses

All data are represented as the mean ± standard deviation and were analyzed using GraphPad Prism (version 9.5) software (Boston, MA, USA). In relation to the number of samples, unpaired Student’s t-test with Welch’s correction (n ≥ 6) or a non-parametric test (Kolmogorov-Smirnov test, n < 6) was used to determine significant differences between two sets of data. One-way analysis of variance (ANOVA) followed by Tukey’s (n ≥ 6; parametric test) or Dunn’s (n < 6; non-parametric test) multiple comparisons test was used in the same program to compare more than two groups. A value of p < 0.05 was considered statistically significant.

Results

Supplemental DEIX protects rats against periodontitis-induced tissue degradation

Figure 1A illustrates the experimental design for the rat groups, periodontitis induction, DEIX supplementation, and sample collection, along with the chemical structure of DEIX. The results from 2D and 3D µCT analyses indicated ligature-mediated degradation of alveolar bone and its restoration through supplementation with DEIX (Fig. 1B). When bone parameter values were determined based on the 3D images, the ligature group revealed significantly lower values of BV/TV (%, p < 0.05) and BMD (g/cm³, p < 0.05) compared with those of the control group (Fig. 1C and D). Ligature-mediated decreases in BV/TV and BMD values were restored by treatment with DEIX up to levels similar to those of the control group. The results from H&E staining supported the ligature-mediated degradation of gingival and PDL tissues and their restoration by the supplemental DEIX (Fig. 1E). While the control, DEIX, and ligature/DEIX groups presented a normal and clearly defined periodontal structure, the ligature group exhibited a characteristic infiltration and accumulation of inflammatory cells in periodontal soft tissues, along with the destroyed linkage of PDL to the tooth. The distance of CEJ-ABC in the ligature group was also significantly (p < 0.05) longer than that in the control or ligature/DEIX group, indicating DEIX’s potency to protect periodontitis-mediated alveolar bone loss (Fig. 1F). IHC results indicated the involvement of increased inflammatory responses in the ligature-mediated periodontal degradation, as proven by the fact that the PDL of the ligature group showed higher COX-2 and TNF-α intensities than did the control, DEIX, or ligature/DEIX group (Fig. 1G). Similarly, the results from ELISA revealed that the ligature group contained significantly higher levels of TNF-α (Fig. 1H, p < 0.05) and IL-1β (Fig. 1I, p < 0.05) in the sera compared with those of the control group, and these increases were apparently attenuated by supplementation with DEIX. These results indicate that the ligature-stimulated periodontal degradation and its inhibition by DEIX are also associated with the ability of supplemental DEIX to recover a systemic inflammatory condition.

Fig. 1.

Supplementation with DEIX protects against periodontal destruction in a rat model of periodontitis. (A) The experimental designs for periodontitis induction and DEIX supplementation in the rat groups before the collection of blood and tissue samples. (B) The 2D and 3D µCT images show the ligature-mediated degradation of alveolar bones and their restoration through supplementation with DEIX. Arrows in upper panels and rectangles in lower panels indicate the ligature-induced loss of alveolar bones and their recovery via the administration of DEIX. The 3D image-based values of (C) BV/TV (%) and (D) BMD (g/cm³) in the rat groups (n = 4). (E) H&E staining images exhibiting the ligature-mediated degradation of gingival and PDL tissues and the DEIX-mediated restoration (Bar = 100 μm). M, molar; D, dentin; G, gingiva; PDL, periodontal ligament; AB, alveolar bone. (F) The distance (µm) between CEJ and ABC in the periodontium of rat groups (n = 4). (G) IHC assay images showing the ligature-mediated increases in the expression of COX-2 and TNF-α in the PDL (Bar = 100 μm). The levels of (H) COX-2 and (I) TNF-α were determined in the sera of rat groups by ELISA (n = 4). ^*p < 0.05 by unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

Administration of DEIX inhibits osteoclastic activation in the periodontium of periodontitis rats

As osteoclastic activation in the periodontium is one of the main factors responsible for periodontitis-mediated alveolar bone loss, we performed a TRAP staining assay to explore whether periodontitis enhances osteoclast formation, and whether that enhancement is inhibited by supplemental DEIX in the rat model of periodontitis. The ligature group revealed greater TRAP-positive cells in the periodontium compared with the control, DEIX, or ligature/DEIX group (Fig. 2A). When the TRAP-positive area (%) in the magnified area was determined, the ligature group showed approximately 3-fold greater area (p < 0.05) compared with the control or DEIX group, and that increase was suppressed mainly by the supplementation with DEIX (Fig. 2B). Results from IF assay indicated that the ligature-stimulated osteoclast formation and its inhibition by DEIX are also associated with DEIX’s ability to suppress the expression of osteoclastogenic molecules, RANKL and cathepsin K, in the periodontium (Fig. 2C). Indeed, significantly higher intensities specific to RANKL (p < 0.05, Fig. 2D) and cathepsin K (p < 0.05, Fig. 2E) were found in the ligature group compared with the control or ligature/DEIX group.

Fig. 2.

Supplementation with DEIX inhibits osteoclastic activation in the inflamed periodontium of rat groups.

(A) TRAP staining images showing the ligature-mediated osteoclast formation and its suppression by DEIX (Bar = 100 μm), in which the tetragonal regions in the upper panels were 4-fold magnified. (B) The TRAP-positive area (%) in the magnified area of rat groups (n = 4). (C) IF assay images indicating the expression levels of RANKL (green) and cathepsin K (red) in the inflamed periodontium of rat groups (Bar = 50 μm). The intensities (a.u.) of (D) RANKL- and (E) cathepsin K-specific immunofluorescences were determined using the ImageJ program (n = 4). ^*p < 0.05 by unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

Supplemental DEIX recovers the expression of osteogenesis-regulatory molecules in the periodontium of periodontitis rats

Next, we evaluated whether the DEIX-induced restoration of alveolar bone loss is accompanied by the expression of osteogenesis- or antioxidation-related proteins by IHC assay. The periodontium of the ligature group contained osterix- or BMP2-positive periodontal cells less than that of the control, DEIX, or ligature/DEIX group (Fig. 3A). Compared with the control group, the periodontium of the ligature group did not show an apparent change in the expression of Nrf2 and HO-1; instead, the DEIX or ligature/DEIX group exhibited greater numbers of HO-1-positive cells compared with the ligature group (Fig. 3B). The results from the IF assay also indicated that periodontitis-mediated decreases in RUNX2 and OPN levels in the periodontium were recovered by DEIX treatment (Fig. 3C). In parallel with this, the intensities specific to RUNX2 (p < 0.05, Fig. 3D) and OPN (p < 0.05, Fig. 3E) were significantly higher in the control or ligature/DEIX group compared with those in the ligature group.

Fig. 3.

Supplemental DEIX recovers the expression of osteogenic molecules and enhances the induction of HO-1 in the periodontium of rats with periodontitis. IHC images indicating the expression levels of (A) osterix and BMP2 and (B) Nrf2 and HO-1 in the inflamed periodontium of rat groups (Bar = 100 μm). (C) IF assay images exhibiting the expression of RUNX2 (green) and OPN (red) in the periodontium of rat groups (Bar = 50 μm). The intensities of (D) RUNX2- or (E) OPN-specific immunofluorescence are provided (n = 4). ^*p < 0.05 by unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

The direct addition of DEIX inhibits LPS-mediated ROS accumulation and functional loss of hPDLCs

We evaluated the direct effect of DEIX on the proliferation of LPS-exposed hPDLCs using the CCK-8 kit. Proliferation rate of the hPDLCs treated with DEIX at concentrations of 1, 10, or 20 µM for 24 h was similar to that of untreated control cells. In contrast, a significant reduction in the proliferation occurred when the cells were exposed to 50 µM DEIX (Fig. 4A). Regarding this, DEIX at a concentration of 20 µM was selected for all subsequent experiments. Exposure to LPS for 24 h significantly (p < 0.01) decreased the proliferation of hPDLCs, whereas that decrease was mostly recovered by the addition of 20 µM DEIX (Fig. 4B). Results from the IF assay indicated that the LPS-exposed hPDLCs exhibited lower intensity specific to the proliferation-related protein, Ki-67, compared with the control, DEIX-, or LPS/DEIX-treated cells (Fig. 4C). When the Ki-67-positive cells (%) were evaluated based on the CLSM images, significantly lower percentages of Ki-67-positive hPDLCs were also found in LPS-stimulated hPDLCs compared with the control (p < 0.01) or LPS/DEIX-treated cells (p < 0.01) (Fig. 4D). As cellular oxidative stress occurred in response to inflammatory mediators, we determined the mitochondrial ROS levels in the hPDLCs exposed to LPS, DEIX, or both for 24 h. Flow cytometric analysis revealed a shift of DCF-specific signal to the light side after LPS treatment, indicating increased ROS production (Fig. 4E). However, the LPS-stimulated increase of DCF-positive hPDLCs was significantly (p < 0.05) diminished by the combined treatment with 20 µM DEIX (Fig. 4F). LPS treatment also significantly increased (p < 0.05) the expression of a DNA damage marker protein, γ-H₂AX, whereas that increase was inhibited by the addition of DEIX (Fig. 4G). In vitro wound healing assay revealed that the potency of hPDLCs to migrate into the scratched gap was visibly attenuated in the presence of LPS, and that attenuation was recovered mainly by DEIX treatment (Fig. 4H). When the area (%) of migrated cells was calculated, the control and LPS/DEIX-treated hPDLCs showed significantly greater areas (p < 0.05) compared with those of the cells exposed to LPS alone (Fig. 4I). The results from RT-qPCR showed that the levels of vimentin, scleraxis, and periostin were significantly (p < 0.05) downregulated in LPS-exposed hPDLCs compared with those in the untreated control cells. In contrast, that reduction was restored by the direct addition of DEIX at significant levels (p < 0.05) (Fig. 4J). These results indicate that LPS-mediated decrease in cell migration and its recovery by DEIX are correlated with the expression levels of mesenchymal lineage marker genes in the hPDLCs. A significant difference (p < 0.05) of DCF-positive hPDLCs (Fig. 4F) or γ-H₂AX protein level between the control and LPS/DEIX groups (Fig. 4G) may suggest that DEIX at the concentration of 20 µM does not entirely suppress the LPS-stimulated ROS production and DNA damage.

Fig. 4.

The direct addition of DEIX inhibits LPS-mediated oxidative stress in hPDLCs and stimulates in vitro wound healing of the cells. (A) CCK-8 assay results showing the proliferation rate (%) of hPDLCs depending on the indicated concentrations (0–50 µM) of DEIX 24 h after incubation (n = 7). (B) Proliferation rate (%) of hPDLCs exposed to LPS (2 µg/mL) and/or DEIX (20 µM) or not to them for 24 h (n = 7). (C) IF assay-derived CLSM images showing the expression of Ki-67 in hPDLCs in relation to the presence and absence of LPS, DEIX, or both 24 h after the incubation (Bar = 10 μm). (D) Ki-67-positive cells (%)/100 counted cells in the hPDLCs untreated or exposed to LPS, DEIX, or both (n = 5). (E) Flow cytometric histograms showing the DCF-specific signals of hPDLCs 24 h after the incubation with and without LPS, DEIX, or both. (F) LPS-mediated increase of DCF-positive hPDLCs (%) and its inhibition by adding DEIX are shown (n = 4). (G) Western blot image of γ-H₂AX that is expressed in the hPDLCs, untreated or exposed to LPS, DEIX, or both for 24 h, along with its expression ratio to that of the loading control, β-actin (n = 4). Original blots are presented in Figure S4. (H) In vitro wound healing images 24 h after the incubation (Bar = 50 μm). (I)The areas (%) of migrated hPDLCs in the scratched regions (n = 4). (J) The RT-qPCR results showing the expression patterns of vimentin, scleraxis, and periostin in the hPDLCs exposed or not to LPS, DEIX, or both for 24 h (n = 4). The letters in panel A indicate significant differences at p < 0.05 among the cells by one-way ANOVA followed by Tukey’s multiple comparisons test. ^*p < 0.05 and ^**p < 0.01 in panel B were determined by an unpaired Student’s t-test. ^*p < 0.05 and ^**p < 0.01 in panels D, F, G, I, and J were calculated using an unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

Treatment with DEIX inhibits the expression of inflammation-related factors and restores the induction of antioxidant molecules in LPS-exposed hPDLCs

We examined the direct effect of DEIX on the expression of inflammation- or antioxidation-related markers in LPS-stimulated hPDLCs via immunoblot and RT-qPCR analyses. Immunoblot analysis indicated that exposure to LPS increases the immunoreactive bands corresponding to TLR4, COX-2, and TNF-α, but suppresses those of Nrf2 or HO-1 in hPDLCs. In contrast, DEIX apparently attenuated such changes (Fig. 5A). Densitometric analysis supported that exposing LPS significantly (p < 0.05) increased the levels of TLR4, COX-2, and TNF-α, but diminished those of Nrf2 and HO-1 compared with the control hPDLCs (Fig. 5B). The direct addition of DEIX reduced the levels of COX-2 (p < 0.05) and TNF-α (p < 0.05), but elevated those of Nrf2 (p < 0.05) and HO-1 (p < 0.05) in LPS-stimulated hPDLCs compared with the LPS alone-exposed cells. Treatment of hPDLCs with 20 µM DEIX alone did not affect the expression of these proteins at a significant level. Similarly, RT-qPCR results revealed that exposure to LPS upregulated the expression of IL-8 and COX-2, but diminished that of Nrf2 and HO-1 compared with those of untreated control cells (Fig. 5C). The addition of DEIX significantly attenuated the LPS-mediated changes in the expression of IL-8 (p < 0.05), COX-2 (p < 0.05), and Nrf2 (p < 0.05). To explore whether Nrf2 acts as the upstream effector of HO-1 in LPS-stimulated hPDLCs, we incubated the cells in combination with 20 µM DEIX, 5 µM ML385, or both for 24 h. Western blot analysis revealed that adding DEIX restored the LPS-mediated decrease in the expression of Nrf2 and HO-1 proteins; however, this restoration was mainly attenuated in the presence of ML385 (Fig. 5D and E).

Fig. 5.

The addition of DEIX inhibits the expression of inflammation-related molecules while restoring the induction of antioxidant molecules in LPS-exposed hPDLCs. (A) Western blot images showing the immunoreactive bands of TLR4, COX-2, TNF-α, Nrf2, and HO-1 in the control hPDLCs or the cells exposed to LPS, DEIX, or both for 24 h. Original blots are presented in Figure S5. (B) Densitometric analysis results show the protein levels’ ratio to that of β-actin (n = 4). (C) The RT-PCR results showing the relative expression levels of IL-8, COX-2, Nrf2, and HO-1 in hPDLCs exposed to LPS and/or DEIX for 24 h (n = 4). (D) Western blot images exhibiting the immunoreactive bands of Nrf2 and HO-1 in the hPDLCs exposed or not to LPS, DEIX, and/or ML385 for 24 h. Original blots are presented in Figure S6. (E) The densitometric assay results showing the ratio of Nrf2 and HO-1 to that of β-actin in the cells (n = 4). ^*p < 0.05 by an unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

The addition of DEIX restores the LPS-mediated decreases of osteogenesis-related molecules in hPDLCs

We subsequently investigated the direct effect of DEIX on the expression of osteogenic regulatory factors in LPS-exposed hPDLCs. To this end, hPDLCs were incubated in the presence and absence of 2 µg/mL LPS and 20 µM DEIX, with and without DAG. Assays for IF staining, immunoblotting, and RT-qPCR were performed at three, seven, and five days after the incubation, respectively. IF staining images indicated that supplementation with DAG visibly increases the expression of RUNX2 mainly in the nuclei of the hPDLCs, and that expression is not affected by DEIX treatment alone (Fig. 6A). However, LPS treatment apparently reduced the DAG-mediated RUNX2 induction in the cells, and the addition of DEIX mostly recovered that reduction. When the RUNX2-positive hPDLCs (%) were determined based on CLSM images, approximately 50% of LPS-exposed hPDLCs exhibited the RUNX2-specific green fluorescence. In contrast, this percentage was significantly (p < 0.05) increased by DEIX to a level similar to that of control cells or cells treated with DEIX alone (Fig. 6B). Similarly, the control (p < 0.05) and LPS/DEIX-treated hPDLCs (p < 0.05) showed significantly greater intensity (a.u.) specific to the green fluorescence compared with cells exposed to LPS alone (Fig. 6C). The results from the immunoblot assay indicated that DAG-mediated increases in the immunoreactive bands specific to RUNX2, OCN, BMP2, and OPN in hPDLCs are strongly diminished by LPS treatment. In contrast, the LPS-mediated reduction is recovered by the addition of DEIX (Fig. 6D). Densitometric analysis supported the LPS-mediated downregulation of osteogenic molecules and their significant recovery upon addition of DEIX (Fig. 6E). Densitometric analysis also revealed no significant difference in the protein level of OPN between the untreated control and DEIX-only-treated hPDLCs (Fig. 6D and E). The RT-qPCR analysis supported the idea that supplementation with DAG upregulated the expression of RUNX2, COL1A1, OCN, and ALP in hPDLCs. That expression was significantly inhibited by LPS treatment (p < 0.05) (Fig. 6F). In contrast, the sclerostin level in DAG-supplemented hPDLCs was barely detected, but instead augmented in the presence of LPS. In contrast, that augmentation was also significantly (p < 0.05) attenuated in combination with DEIX. These findings suggest that LPS-mediated inflammatory conditions suppress osteogenic activation by downregulating osteogenesis-related molecules and upregulating the BMP antagonist sclerostin. Our results suggest that DEIX can mitigate alveolar bone loss by recovering bone formation-related factors in the inflamed periodontal tissues.

Fig. 6.

The addition of DEIX restores the expression of osteogenic molecules in LPS-exposed hPDLCs. (A) CLSM images exhibiting the expression level of RUNX2 in the hPDLCs incubated with and without LPS, DEIX, or both for three days (Bar = 10 μm). (B) The percentage of RUNX2-positive hPDLCs and (C) the intensity (a.u.) of RUNX2-specific green fluorescence in the cells (n = 4). (D) Immunoblot data showing the immunoreactive bands of RUNX2, osterix, OCN, BMP2, and OPN in the hPDLCs exposed to LPS and/or DEIX in the presence and absence of DAG for seven days. Original blots are presented in Figure S7. (E) Densitometric assay results show the ratio of the evaluated proteins to that of β-actin (n = 4). (F) The RT-qPCR results show the relative expression of RUNX2, COL1A1, OCN, ALP, and sclerostin compared to that of GAPDH in hPDLCs exposed or not to LPS, DEIX, and/or DAG for five days (n = 4). ^*p < 0.05 by an unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

The addition of DEIX suppresses LPS-stimulated transwell migration and osteoclastic activation in THP-1 cells

We examined the direct effect of DEIX on inflammation-stimulated cell migration and osteoclastic activation using THP-1 cells. The captured images from the transwell migration assay exhibit that the migration of THP-1 cells occurred more greatly when the cells were incubated with the medium collected from LPS-exposed hPDLCs compared with that from the untreated control, DEIX-only-, or LPS/DEIX-supplied cells (Fig. 7A). When the number of migrated THP-1 cells per well was counted based on the optical images, the cultures supplied with medium from LPS-stimulated cells revealed a 2.2-fold greater migration compared with untreated control cells (Fig. 7B). However, the LPS-stimulated migration was significantly (p < 0.05) diminished in the presence of DEIX, reaching a level similar to that of untreated control cells, indicating a possible role of DEIX in limiting the recruitment of monocytes under inflammatory conditions. TRAP staining images revealed that LPS treatment effectively augmented osteoclast formation in RANKL-supplied THP-1 cells, whereas that augmentation was directly diminished by the addition of DEIX (Fig. 7C). The number of osteoclasts/well and of the nuclei per osteoclast formed with LPS was approximately 3-fold greater than those formed with RANKL alone (Fig. 7D and E). The combination with LPS increased the mean diameter of osteoclasts (mm²) more significantly than LPS treatment alone, which also augmented the number of osteoclasts and nuclei within the cells (Fig. 7F). However, the cultures of THP-1 cells in combination with DEIX revealed significantly fewer osteoclasts (p < 0.05) and nuclei numbers (p < 0.05) and lower diameters (p < 0.05) of the cells compared with those in the LPS-stimulated cells (Fig. 7D–F). The results from Western blotting indicate that the LPS-mediated augmentation in osteoclast formation of RANKL-supplied THP-1 cells is orchestrated by the enhanced expression of osteoclast-specific proteins (Fig. 7G). Densitometric analysis revealed that the cultures of THP-1 in combination with LPS significantly increased the levels of c-Fos (p < 0.05), cathepsin K (p < 0.05), and MMP-9 (p < 0.05), but not of NFATc1, compared with cells stimulated with RANKL only (Fig. 7H). The levels of c-Fos, NFATc1, and MMP-9 proteins in LPS-stimulated hPDLCs were significantly (p < 0.05) reduced in the presence of DEIX. These results indicate that the levels of these proteins, enhanced by LPS stimulation, are diminished by the addition of DEIX, suggesting the in vitro anti-osteoclastic potency of DEIX through the downregulation of inflammation-stimulated osteoclastic molecules.

Fig. 7.

DEIX directly suppresses the migration and osteoclastic activation in LPS-stimulated THP-1 cells. (A) The transwell migration images exhibiting the LPS-stimulated migration of THP-1 cells and its inhibition by the direct addition of DEIX 24 h after the incubation (Bar = 500 μm). (B) The number of migrated THP-1 cells/well in the cells (n = 4). (C) TRAP staining images revealing the LPS-enhanced osteoclast formation in THP-1 cells and its suppression via the DEIX treatment five days after the incubation (Bar = 300 μm). The number of (D) osteoclasts/well and of (E) the nuclei per osteoclast, along with (F) the mean diameter (mm²) of osteoclasts that were formed by the THP-1 cells exposed to LPS, DEIX, or both in the presence of RANKL (n = 4). (G) Western blot images showing the immunoreactive bands of c-Fos, NFATc1, cathepsin K, and MMP-9 in the THP-1 cells exposed to LPS, DEIX, or both for four days. Original blots are presented in Figure S8. (H) Densitometric results show these proteins’ ratio to that of β-actin (n = 4). ^*p < 0.05 by an unpaired non-parametric test (Kolmogorov-Smirnov test). ns, not significant.

Identification and functional analyses of DEGs via RNA sequence profiling support the potency of DEIX to protect cells against inflammatory damage

To further verify the effect of DEIX on LPS-stimulated cellular responses, RNA sequence profiling on the hPDLCs exposed to LPS and/or DEIX for three days was analyzed. The response of DEGs and their KEGC and GO functional annotations were analyzed using the GEO database with the accession number GSE305324 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE305324). Figure 8A shows the heatmap of the DEGs selected from all the samples in log₁₀ (FPKM + 1) units at the fold change levels ≥ 2 and p value ≤ 0.05. When the numbers of DEGs were compared with the control hPDLCs in the value of p ≤ 0.05, the cells exposed to LPS, LPS/DEIX, or DEIX alone revealed 62, 51, and 40 upregulated genes and 9, 27, and 25 downregulated genes, respectively (Fig. 8B). Compared with the hPDLCs exposed to LPS alone, 54 genes were upregulated, whereas 88 genes were downregulated in the LPS/DEIX-treated cells. Figure 8C represents the heatmap for the DEGs derived from the control and LPS-exposed hPDLCs at the levels of p ≤ 0.05, in which the names of 62 upregulated (red) and 9 downregulated genes (green) were provided. The color key of the heatmap indicated that the DEGs, such as TFP12, MMP1, MMP3, FOS, SLP1, EGR1, IL11, and CXCL8, were highly expressed in the LPS-stimulated cells compared with other DEGs in those cells. We conducted functional enrichment and pathway analysis of significantly enriched modules using g: Profiler to determine GO terms, including molecular function (MF), cellular compartment (CC), and biological process (BP), as well as several biological pathways. Data sources for these pathways included KEGG, Reactome (REAC), Transcription factors (TF), miRTarBase (MIRNA), Human Protein Atlas (HPA), CORUM, Human Phenotype Ontology (HP), and WikiPathways (WP). Figure 8D exhibits the plot summary of various functional analyses, in which the DEGs in each of the indicated pathways are distributed as different colors in variously sized circles in -10 log (adjusted p value). Based on the plot summaries, the hPDLCs exposed to LPS alone revealed greater distributions of DEGs than the cells combined with DEIX. Unlike the LPS-exposed hPDLCs, the LPS/DEIX-treated cells showed the presence of DEGs in the HPA and CORUM databases. To better understand the direct effect of DEIX on the gene expression in LPS-stimulated hPDLCs, the enriched biological terms from the DEGs between the control and experimental groups were divided into the top 10 terms in GO-BP, GO-CC, GO-MF, and KEGG pathways by singular enrichment analysis (SEA) (Figure S2A–D). The top 10 SEA terms of GO-BP between the control and LPS-exposed hPDLCs were shown in the order of neutrophil chemotaxis, chemokine-mediated signaling pathway, response to bacterium, cellular response to chemokine, response to chemokine, inflammatory response, response to cytokine, neutrophil migration, granulocyte chemotaxis, and response to peptide (Figure S2A). The only exception to this term was the response to the bacterium; other SEA terms in LPS-exposed cells were not found in cells exposed to LPS in combination with DEIX. Similar to the LPS/DEIX-treated hPDLCs, the LPS-only-exposed cells revealed extracellular space and region as the main SEA terms of the GO-CC. In contrast, the − 10 log adjusted p values corresponding to these terms were apparently diminished by the addition of DEIX (Figure S2B). The GO-CC terms, specific granule lumen and specific granule, were found only in the LPS-stimulated hPDLCs. The top one SEA term of GO-MF in the LPS-stimulated hPDLCs was CXCR chemokine receptor binding, followed by chemokine and cytokine activities, receptor binding, and signaling receptor activity, and these terms were also mostly found in the cells treated with LPS in combination with DEIX (Figure S2C). However, the − 10 log adjusted p-values of the GO-MF terms shown in LPS-exposed cells were markedly reduced by the addition of DEIX. Figure S2D shows the top 10 SEA terms of the KEGG pathway, in which the IL-17 signaling pathway was the top one in the LPS-exposed hPDLCs, followed by cytokine and chemokine-associated signaling pathways. Similar to the results from GO analyses, the − 10 log adjusted p values of the KEGG pathway-derived SEA terms in LPS-exposed hPDLCs were markedly reduced in the cultures combined with DEIX. The KEGG term’ chemokine signaling pathway’ was also found only in cells exposed to LPS alone, compared with the untreated control or cells combined with DEIX. To clarify the effect of DEIX on the LPS-mediated gene expression alterations, we evaluated how the LPS-mediated up- or down-regulated 71 DEGs were affected by the presence of DEIX. Figure 9A exhibits the magnified heatmap for the mean expression values of the 71 DEGs from all samples at p-value ≤ 0.05. Based on the color key of the heatmap, the up- or down-regulated DEGs in the LPS-stimulated hPDLCs were mostly restored in the presence of DEIX up to the levels similar to those of untreated control cells. Subsequently, we selected 24 DEGs that showed 2-fold less or greater expression in LPS-stimulated hPDLCs compared with control cells and analyzed the statistical differences of the DEGs in relation to the presence or absence of LPS and/or DEIX. Ordinary one-way ANOVA with Tukey’s multiple comparisons test revealed that LPS treatment alone significantly increased the levels of C3, CFB, CXCL1, CXCL2, CXCL3, CXCL6, CXCL8, EGR2, IL11, KRT14, LYG1, MMP3, MMP13, PATL2, PCSK9, SAA1, SFRP4, SLPI, SMIM1, and TNFAIP6, but decreased that of DIO3, HSPB3, LAMP5, and MT1M in hPDLCs, compared with the control or the cells in combination with DEIX (data not shown). However, non-parametric ANOVA results (Kruskal-Wallis test with Dunn’s multiple comparisons test) showed that LPS treatment alone significantly upregulated or downregulated the levels of C3, CFB, CXCL1, CXCL3, CXCL6, CXCL8, KRT14, PATL2, and PCSK9 compared with the control cells (Fig. 9B). The LPS-exposed cells also showed significantly different levels of DIO3, IL11, LYG1, MMP3, SAA1, and TNFAIP6 compared with those from LPS/DEIX- or DEIX-only-treated cells. In addition, no significant differences in the expression of these genes, except LAMP5, were found between the control cells and LPS/DEIX-or DEIX-only-treated cells. These results indicate that adding DEIX directly suppresses the LPS-activated inflammatory damage by regulating the genes that activate chemokine and cytokine-associated signaling pathways in an inflammatory condition.

Fig. 8.

RNA sequence profiling demonstrates the regulatory roles of DEIX on the expression of genes that are up- or down-regulated in LPS-stimulated hPDLCs. (A) The heatmap showing the expression values of DEGs derived from the hPDLCs exposed or not to LPS, DEIX, or both for three days at the levels of fold change ≥ 2 and p value ≤ 0.05. (B) The number of up- or down-regulated DEGs between the indicated samples at p ≤ 0.05 and q ≤ 0.01, respectively (n = 3). (C) The heatmap for the DEGs derived from the control and LPS-exposed hPDLCs at the levels of p ≤ 0.05, along with the indication of 62 upregulated (red) and 9 downregulated genes (green). (D) The plot summary images showing the results from functional enrichment and pathway analysis of significantly enriched modules using g: Profiler, in which DEGs corresponding to GO-MF, GO-CC, and GO-BP, along with the data sources including KEGG (https://www.kegg.jp/kegg/kegg1.html), REAC, TF, MIRNA, HPA, CORUM, HP, and WP are represented as different colors with variously sized circles in -10 log adjusted p value.

Fig. 9.

DEIX effectively downregulates the expression of inflammation-associated chemokines and cytokines in LPS-stimulated hPDLCs. (A) The heatmap showing the expression levels of 71 up- or down-regulated DEGs in LPS-exposed hPDLCs after the comparison with those of untreated control cells at the value of p ≤ 0.05, as well as the color changes corresponding to the expression levels of genes in relation to the presence and absence of DEIX. (B) The graphs illustrate the effects of DEIX on the levels of 24 selected DEGs, which are significantly different from those of control hPDLCs at p < 0.05 (n = 3). The letters in panel B indicate significant differences among these cells, as shown by a non-parametric ANOVA test followed by Dunn’s multiple comparisons test.

The functional analyses of DEGs highlight the nature of DEIX to improve bone metabolism and anti-inflammatory responses

Based on the results from RNA sequencing profiling, we further investigated the nature of DEIX treatment on gene expression in hPDLCs. Figure 10A represents the heatmap for the expression values in log₁₀ (FPKM + 1) units of the DEIX-induced 65 DEGs that differed from the untreated control hPDLCs at the levels of fold change ≥ 2 and p value ≤ 0.05. Compared with the control hPDLCs, 40 DEGs were upregulated, whereas 25 DEGs were downregulated in the DEIX-treated cells. The plot summary of functional analyses revealed that the DEIX-induced DEGs were predominantly distributed in GO, with a partial presence in TF, HPA, and CORUM databases (Fig. 10B). The top one SEA term of GO-BP (Fig. 10C), GO-CC (Fig. 10D), and GO-MF (Fig. 10E) were cell-cell signaling, dendrite membrane, and cytokine activity, respectively. However, given a list of the DEGs from this comparison pair, no significantly enriched terms were found in the KEGG pathway database. Dissimilar to GO-BP and GO-CC, only one significantly enriched term was detected in the GO-MF. From the DEIX-mediated 65 DEGs, we selected 42 DEGs that showed significant differences from those of untreated controls at the level of p < 0.05 by a parametric Student’s t-test, in which the expression levels of 29 up- and 13 down-regulated genes were compared in relation to the presence and absence of LPS and/or DEIX. Differently to the parametric ANOVA results (data not shown), the results from a non-parametric ANOVA test did not show any significant differences in the expression of the selected genes (BMP2, COL13A1, CYGB, FAM43A, LIF, OLFM2, SECTM1, SERPINB2, THBD, INHBE, and TNC) between the DEIX and the control, LPS, or LPS/DEIX groups (Fig. 10F and G). In contrast, the DEIX-treated hPDLCs showed significantly greater TIPARP and lower IL18R1 and TXNIP levels compared with those in the LPS-exposed cells. Moreover, the other 17 up- and 9 down-regulated DEGs at the level of p < 0.05 by a parametric Student’s t-test, except DNAH7 and SHISA9, did not show significant differences between the DEIX and the control or LPS-exposed cells (Figure S3A and B).

Fig. 10.

DEIX treatment alone improves bone metabolism and anti-inflammation-related biological processes in hPDLCs . (A) The heatmap exhibiting the expression values of DEIX-specific DEGs in log₁₀ (FPKM + 1) units at the levels of fold change ≥ 2 and p value ≤ 0.05, along with the comparison with those of the control hPDLCs, in which 40 upregulated DEGs (red) and 25 downregulated DEGs are shown. (B) The plot summary of functional enrichment and pathway analysis of significantly enriched modules derived from the DEGs of DEIX-treated hPDLCs in -10 log adjusted p value. Significantly enriched top 10 terms of (C) GO-BP, (D) GO-CC, and (E) GO-MF in the DEIX-derived DEGs are shown. (F and G) The graphs show the significant differences of the indicated DEGs among the cells untreated or exposed to LPS, DEIX, or both at p < 0.05 (n = 3). The letters in panels F and G indicate significant differences among these cells, as shown by a non-parametric ANOVA test followed by Dunn’s multiple comparisons test.

Discussion

Periodontitis is a chronic infectious disease, in which P. gingivalis, a gram-negative anaerobic bacterium, is one of the key pathogens corresponding to the initiation and development of periodontitis¹⁹. P. gingivalis evades host immune responses, creates a chronic inflammatory environment, and contributes to tissue destruction in the periodontium²⁰. The P. gingivalis-derived inflammatory responses are closely linked to their byproducts, which stimulate immune cells and fibroblasts present in the periodontium, leading to the production of proinflammatory cytokines and excessive ROS^21,22. Regarding this, we used ligatures cultured with P. gingivalis to induce an experimental rat model of periodontitis, as it directly leads to periodontal inflammation similar to that observed in human oral pathogenesis.

As proven by the results from µCT and H&E staining assays, the supplementation with DEIX protected against the degradation of supporting tissues of the tooth, such as gingiva, PDL, and alveolar bone, in the rat model of periodontitis. The results from IHC and ELISA assays indicated that ligature-induced periodontitis involves the host’s inflammatory immune activation in both local and systemic conditions, and supplemental DEIX effectively diminishes this activation. In general, TNF-α and IL-1β are recognized as key cytokines that mediate the inflammatory response to bacterial infection in the gums, promoting the breakdown of connective tissue and bone surrounding the teeth^23,24. COX-2 plays a significant role in the inflammatory response by producing prostaglandins that contribute to tissue destruction and bone resorption in periodontitis^25,26. Together, our findings indicate that DEIX protects against alveolar bone loss and connective tissue degradation by inhibiting the expression of inflammatory mediators in the rat model of periodontitis.

Inflammatory mediators directly and indirectly stimulate alveolar bone loss. Indeed, TNF-α, a highly secreted cytokine in chronic periodontitis, stimulates the activity of osteoclasts and contributes to alveolar bone resorption in periodontitis^27,28. RANKL is a key signaling molecule that triggers the differentiation and maturation of osteoclasts through its interaction with the RANK receptor²⁹. RANKL is produced from the host’s immune response to bacterial infection and promotes excessive bone resorption in periodontitis³⁰. Cathepsin K is a key enzyme in osteoclasts and is produced in proportion to osteoclast activity; therefore, its level indicates the progression of alveolar bone loss^31,32. Our results indicate that ligature-induced alveolar bone loss is closely associated with osteoclastic activation, accompanied by elevated RANKL and cathepsin K levels in the connective tissues of the periodontium. Our findings also suggest that the anti-osteoclastogenic potency of DEIX is correlated with its ability to downregulate the expression of RANKL and cathepsin K in the inflamed periodontium.

While proinflammatory cytokines directly stimulate osteoclast activity in periodontitis, they may also impair the function of osteoblasts³³. The imbalance between osteoclast and osteoblast activity leads to excessive alveolar bone resorption in periodontitis³⁴. Osteoblast activity is tightly regulated by osteogenesis-specific transcription factors, including RUNX2, osterix, and BMP2, as well as their downstream effectors, such as COL1A1, OPN, and OCN³⁵. RUNX2 plays a crucial role in maintaining bone homeostasis, and thus the disruptions in RUNX2 expression or function in periodontitis enhance alveolar bone loss and PDL’s structural and functional destruction³⁶. While osterix is not directly involved in the inflammatory process of periodontitis, it plays important roles in bone formation and remodeling, as well as in the function of PDL cells³⁷. BMP2 plays a pivotal role in bone formation and regeneration³⁵. The Nrf2/HO-1 pathway also plays a crucial role in periodontitis by regulating oxidative stress, inflammation, and bone homeostasis within the periodontium³⁸. In periodontitis, the Nrf2/HO-1 pathway is often suppressed, leading to increased oxidative stress and inflammation in periodontal tissues, which in turn facilitates alveolar bone loss³⁹. Our results from IHC and IF assays indicate that the DEIX-induced protection against alveolar bone loss in the periodontitis animal model is, in part, associated with the efficacy of DEIX to restore the expression of osteogenic transcription factors and the Nrf2/HO-1 pathway in inflamed periodontium.

To explore the cellular mechanisms by which supplemental DEIX protects against alveolar bone loss and connective tissue destruction in the rat model of periodontitis, we investigated the direct effect of DEIX on LPS-stimulated hPDLCs or THP-1 cells. In this study, we utilized P. gingivalis-derived LPS to induce cellular inflammatory stress, which mimics the stress that occurs in the periodontium of the animal model. We also used hPDLCs and THP-1 cells as in vitro experimental cells. This is because the interactions between PDLCs, immune cells, and bacterial pathogens have garnered significant attention in understanding the pathogenesis of periodontitis⁴⁰. THP-1 cells are commonly used as a model for studying macrophage function and immune responses⁴¹. When exposed to oral pathogens, THP-1 cells differentiate into macrophage-like cells, simulating the innate immune response⁴². Moreover, the interaction among PDLCs, THP-1, and P. gingivalis-derived pathogen forms a complex network of cellular and molecular events that drives periodontal inflammation and tissue degradation⁴³. Furthermore, insights gained from these interactions can suggest the possible mechanisms involved in DEIX-mediated protection against periodontitis. They may provide new therapeutic targets for the prevention and treatment of periodontal disease⁴⁴.

As evidenced by the results from CCK-8, IF, flow cytometry, and immunoblot assays, our findings demonstrate that the addition of DEIX directly restores the decreased proliferation, ROS accumulation, and oxidative DNA damage in LPS-stimulated hPDLCs. Our results show that the direct addition of DEIX increases the migration of LPS-exposed hPDLCs. That increase is accompanied by the restored expression of genes encoding the structural and mesenchymal lineage marker proteins, such as vimentin, scleraxis, and periostin. The in vitro study also highlights that LPS-stimulated hPDLCs exhibit TLR4-associated upregulation of COX-2, TNF-α, and IL-8, along with decreased expression of Nrf2 and HO-1. In contrast, the addition of DEIX mainly restores these changes. In addition, this study highlights that DEIX directly restores the expression of osteoblast-specific marker proteins in LPS-stimulated hPDLCs and inhibits LPS-stimulated transwell migration and osteoclast formation in THP-1 cells by suppressing osteoclastogenic factors. Taken as a whole, all these findings support the involvement of anti-inflammatory, antioxidant, and anti-osteoclastic activities in the DEIX-mediated protection against inflammatory periodontal degradation.

While vimentin is the major cytoskeletal component protein⁴⁵, scleraxis, a member of the basic helix-loop-helix superfamily of transcription factors, is involved in mesoderm formation⁴⁶. Periostin is a secreted ECM protein that functions as a ligand to support adhesion and migration of cells and influences ECM restructuring and tissue remodeling⁴⁷. LPS is the primary ligand for TLR4, which activates the innate immune response in periodontitis⁴⁸. Sclerostin, a secreted glycoprotein, is produced primarily by osteocytes and acts as a non-classical BMP antagonist, inhibiting bone formation by osteoblasts⁴⁹. In addition to RANKL and cathepsin K, the c-Fos, NFATc1, and MMP-9 play crucial roles in osteoclast formation and function. For example, a component of activator protein-1, c-Fos, is induced by RANKL and directly regulates the expression of several genes critical for osteoclast development⁵⁰. NFATc1, one of the key target genes regulated by c-Fos, is essential for RANKL-induced osteoclast formation, fusion, and activation, and contributes to both normal bone remodeling and pathological bone loss⁵¹. MMP-9 is expressed in osteoclasts and plays a complex role in osteoclast formation and bone remodeling, as well as in the degradation of bone matrix components⁵². Considering the regulatory roles of these transcription factors or marker proteins, our results from in vitro experiments may support the therapeutic efficacy and the associated mechanism of DEIX on inflammatory periodontal destruction.To further explore the mechanisms involved in DEIX-mediated protection against inflammatory cellular damage, we performed RNA sequence analysis using the hPDLCs exposed to LPS, DEIX, or both. That profiling may indicate that the hPDLCs exposed to LPS highly express the genes encoding the proteins that are involved in chemoattraction (CXCLs)⁵³, inflammatory immune response (C3, CFB, IL11, LYG1, SAA1, and TNFAIP6)^54–56, metabolism (PCSK9, SFRP4, SLPI, and SMIM1)^57–60, and tissue resorption (MMP3, and MMP13)⁶¹, whereas the genes (HSPB3 and MT1M) associated with cellular defense mechanisms^62,63 are downregulated in the cells. Our results also reveal that the direct addition of DEIX may restore changes in gene expression, whereas the DEIX treatment alone does not alter such gene expression. These findings may suggest that DEIX protects cells against LPS-stimulated inflammatory damage mainly by downregulating chemokine and cytokine-associated signaling pathways. Our current findings also show a possibility that DEIX treatment alone upregulates the genes involved in bone remodeling and tissue homeostasis (BMP2), ROS scavenging (CYGB)⁶⁴, stem cell maintenance (LIF)⁶⁵, immune system (LIF, SECTM1, and THBD)^65–67, and various cellular processes, including adhesion, migration, development, differentiation, and inflammation (OLFM2, SERPINB2, and TIPARP)^68–70. However, the genes encoding the proteins involved in inflammatory response (IL18R1)⁷¹, insulin resistance (INHBE)⁷², cellular adhesion inhibition (TNC)⁷³, and oxidative stress (TXNIP)⁷⁴ are unchanged or downregulated in the DEIX-only-treated hPDLCs.

Taken as a whole, this study has several limitations and challenges for clinical translation that warrant future consideration. First, the experiments were conducted exclusively in a murine model of periodontitis, which may limit the generalizability of findings to other species, including humans, due to potential differences in periodontal pathophysiology and immune responses across models. Second, the in vitro assessments relied solely on LPS stimulation, potentially overlooking the contributions of polymicrobial dysbiosis or other pathogen-associated molecular patterns characteristic of clinical periodontitis. Third, this study has used a low number of replicates, specifically in the mRNA sequence profiling. Finally, the absence of pharmacokinetic and bioavailability evaluations for DEIX precludes definitive conclusions regarding its systemic absorption, tissue distribution, and optimal dosing in vivo. Future investigations incorporating diverse animal models, multifaceted in vitro stimuli, and comprehensive pharmacokinetic profiling, along with confirmatory sample size, will be essential to validate and extend these observations.

Conclusion

This study demonstrates that supplemental DEIX protects rats against inflammatory alveolar bone loss and connective tissue degradation by inhibiting inflammatory response, oxidative stress, and the imbalance between osteoclast and osteoblast activity. Our in vitro experiments, combined with RNA sequence profiling, support the potency of DEIX in directly mitigating LPS-stimulated inflammatory damage in hPDLCs or THP-1 cells, as well as the underlying mechanisms. Overall, this study suggests that DEIX may apply to patients with chronic periodontitis, although further detailed experiments are required. Additional experiments are necessary to verify the healing efficacy of DEIX on an experimental model of periodontal tissue defects, as well as to enhance its therapeutic efficacy.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

GB, Y-HA, SS, J-CL and Y-MJ contributed to the study conception and design, methodology, data collection, and investigation. SR, S-MP, and S-HK analyzed the data and participated in validation. The first draft of the manuscript was written by GB, Y-HA and SS and reviewed by J-CL and Y-MJ. S-HK, J-CL and Y-MJ secured project funding. Supervision and project administration were by J-CL and Y-MJ. All the authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the National Research Foundation (NRF) of Korea, funded by the Ministry of Science, Information and Communications Technology, and Future Planning [2021R1A2C2006032, RS-2023-00277774, and RS-2024-00338143].

Data availability

The mRNA sequence data that support the findings of this study have been deposited in the GEO database with the accession number GSE305324 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc= GSE305324). The datasets used and/or analyzed during the current study are also available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.