A20 Orchestrates Inflammatory Response in the Oral Mucosa through Restraining NF-κB Activity

Abstract

Deregulated immune response to a dysbiotic resident microflora within the oral cavity leads to chronic periodontal disease, local tissue destruction, and various systemic complications. To preserve tissue homeostasis, inflammatory signaling pathways involved in the progression of periodontitis must be tightly regulated. A20 (TNFAIP3), a ubiquitin-editing enzyme, has emerged as one of the key regulators of inflammation. Yet, the function of A20 in the oral mucosa and the biological pathways in which A20 mitigates periodontal inflammation remain elusive. Using a combination of in vivo and ex vivo disease models, we report in this study that A20 regulates inflammatory responses to a keystone oral bacterium, Porphyromonas gingivalis, and restrains periodontal inflammation through its effect on NF-κB signaling and cytokine production. Depletion of A20 using gene editing in human macrophage-like cells (THP-1) significantly increased cytokine secretion, whereas A20 overexpression using lentivirus infection dampened the cytokine production following bacterial challenge through modulating NF-κB activity. Similar to human cells, bone marrow–derived macrophages from A20-deficient mice infected with P. gingivalis displayed increased NF-κB activity and cytokine production compared with the cells isolated from A20-competent mice. Subsequent experiments using a murine ligature-induced periodontitis model showed that even a partial loss of A20 promotes an increased inflammatory phenotype and more severe bone loss, further verifying the critical function of A20 in the oral mucosa. Collectively, to our knowledge, these findings reveal the first systematic evidence of a physiological role for A20 in the maintenance of oral tissue homeostasis as a negative regulator of inflammation.

Introduction

The oral cavity is home to a diverse group of microbiota, including more than 700 bacterial species, fungi, and viruses and serves as a gateway between the environment and the body (1). The gingival tissues exploit an elaborate immune system to protect the host against continuous microbial insult and external stress and keep the subtle balance between the host immune response and the resident oral microbiota to maintain tissue homeostasis (2). Similar to various other tissues, initiation of inflammation in the oral mucosa is mainly driven by TLRs and NF-κB signaling cascade. Specifically, engagement of TLR2, TLR4, and TLR9 with the oral microbe-associated molecular patterns triggers production of inflammatory cytokines/chemokines and tissue-damaging metalloproteases through recruiting the MyD88 adaptor to activate the IL-1 receptor-associated kinase (IRAK), TNF receptor–associated factor 6 (TRAF6), and TGF β–activated kinase 1 (TAK1) complexes, which subsequently lead to degradation of IκB and nuclear translocation of NF-κB (2, 3). Initiation of inflammation through the NF-κB signaling and subsequent release of proinflammatory cytokines are certainly protective against continuous microbial challenge, whereas failure to attain timely termination of the immune response leads to unsustainable inflammation, disruption of tissue homeostasis, and consequent periodontal disease (1, 2).

Periodontal disease is one of the most prevalent chronic inflammatory diseases affecting almost half of the adult population and is characterized by alveolar bone destruction and tooth loss (4). Persistent forms of the disease are also associated with several systemic conditions, including diabetes, cardiovascular disease, cancer, adverse pregnancy outcomes, and rheumatoid arthritis (5–8). Prevention of adverse clinical outcomes at local and distant tissues requires thorough understanding of the interactions between different components of the host immune system with each other and the resident oral microbiome and identification of key regulatory pathways involved in the maintenance of the oral mucosal homeostasis. In this regard, although the biological pathways initiating the inflammatory responses in the oral mucosa are relatively well- characterized, there are still gaps in our knowledge about the downstream regulatory pathways that function to terminate periodontal inflammation.

The NF-κB signaling pathway is prominently regulated by ubiquitination, which is a reversible posttranslational modification that can terminate cell signaling through proteasome-mediated degradation and/or interfering with protein trafficking through activation of kinases and phosphatases (9). Although ubiquitination and ubiquitin- related molecules receive attention as regulators of multiple biological pathways in the pathophysiology of numerous chronic inflammatory conditions, we still do not know much about their function in the oral mucosa (10). A20, also known as TNF-α inducible protein 3 (TNFAIP3), is a ubiquitin-editing enzyme and has recently emerged as a critical negative regulator of inflammation through termination of NF-κB activation as part of negative feedback loop (11). A20 is comprised of two functional domains, including an N-terminal ovarian tumor domain with deubiquitinating activity, and a C-terminal with seven zinc finger domains that support E3 ubiquitin ligase activity. A20 gets activated downstream of NF-κB and interferes with the ubiquitination of multiple substrates such as TRAF6, NF-κB essential modulator (NEMO), and receptor-interacting serine/threonine-protein kinase 1 (RIP), subsequently limiting inflammation triggered by the activation of numerous immune sensors such as TLRs and nucleotide-binding oligomerization domain–like receptors, TNFR, IL-1R, and IL-17R (11–15). These signaling pathways are also involved in periodontal disease pathophysiology. Substantiating its essential role in restraining inflammation, the A20 knockout (A20^−/−) mouse develops spontaneous multiorgan inflammation and dies within 3–4 wk after birth (16). Impaired A20 function or deficiency is associated with unsustainable inflammation and tissue damage in several disease models, including autoimmune diseases, rheumatoid arthritis, gastrointestinal and hepatic disorders, cystic fibrosis, aging, cancer, and psoriasis (17–25). Remarkably, a majority of the diseases associated with A20 share common pathophysiology with periodontitis, and it is likely that A20 functions as a gatekeeper of inflammation at the host–microbiome interface in the oral mucosa as well. Indeed, our previous study reported A20 expression in the oral mucosa and revealed A20 as a potential regulator of periodontal inflammation downstream of TLR signaling (26). Clinical findings also suggested that A20 protein levels might not be sustained during the course of periodontitis to restrain inflammation (26). Yet, the function of A20 in the oral mucosa and the biological pathways in which A20 mitigates periodontal inflammation remain elusive. Therefore, in this study, we seek to characterize the role of A20 in regulating inflammatory responses in the oral mucosa, particularly focusing on A20-driven cellular function in macrophages, using established preclinical periodontal disease models. Porphyromonas gingivalis, one of the keystone periodontal bacteria, can activate NF-κB signaling in macrophages and induce A20 expression in vivo when injected into the peritoneal cavity of mice and, therefore, it is included as a model oral bacteria in in vitro assays (26). Our results revealed that A20 inhibits proinflammatory cytokine production both in human and mouse macrophages infected with P. gingivalis through its effect on NF-κB signaling. Collaborating with the in vitro data, mice with partial loss of A20 (A20^+/−) displayed an increased inflammatory phenotype as measured by increased bone loss and gingival inflammatory cytokine expression compared with the A20-competent mice (A20^+/+) mice following induction of periodontitis. Taken together, to our knowledge, our findings present first systematic evidence and establish a novel physiological role of A20 in the oral mucosa as a regulator of inflammation.

Materials and Methods

Bacteria

P. gingivalis (strain ATCC33277) was maintained in anaerobic chambers with brain heart infusion broth supplemented with 0.5% (w/v) yeast extract, 5 g/ml hemin, 0.5 g/ml vitamin K, and 0.1% (w/v) cysteine. The bacteria were killed by heating at 80°C for 10 min and confirmed by inoculation on agar plates. (26)

Bone marrow–derived macrophage isolation

Bone marrow–derived macrophages (BMDMs) were isolated from femurs and tibias of mice aged 4–8 wk old following standard protocol with mild modification (27). Briefly, femurs and tibias were isolated, and the bone marrows were washed down with RPMI 1640 medium. RBCs were eliminated with ACK lysis buffer (Quality Biological). The bone marrow cells were then counted and seeded in a 10-cm petri dish with a density of 2 × 10⁶ cells per dish in DMEM with L929 supernatant (glutamine, 4.5 g/L glucose, 100 mg/L sodium pyruvate, 10% FCS, 0.05 mM mercaptoethanol, and 25% supernatant from L929 cells [American Type Culture Collection; CCL-1]). Cells were harvested on day 6 and reseeded in 12-well plates with a density of 2 × 10⁶ cells per well of DMEM without L929 supernatant (glutamine, 4.5 g/L glucose, 100 mg/L sodium pyruvate, 10% FCS, and 0.05 mM mercaptoethanol) and used for experiments.

Cell culture and enzyme-linked immunoabsorbent assay

HEK293T cells were maintained in DMEM (Life Technologies). THP-1 cells were maintained in RPMI 1640 (Life Technologies). THP-1 cells were treated with 25 ng/ml PMA overnight to differentiate into macrophages. The cell culture supernatants from THP-1 or BMDMs were harvested after P. gingivalis infection (100 multiplicity of infection [MOI]) at different time points and analyzed using Human or Mouse IL-6/TNF ELISA Ready-SET-Go! Kit (eBioscience), respectively, following the user manual; each experiment was performed with minimum of three times independently with triplicates or tetraplicates. The plates were read at 450 nm with Multiskan MCC (Thermo Fisher Scientific).

Animal studies

All studies involving mice have been approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University. A20 ^+/− mice of the C57BL/6 background were kindly provided by Averil Ma (University of California, San Francisco). C57BL/6 A20^+/− mice were crossbred to obtain A20^+/+ and A20^+/− mice. To analyze the offspring genotype, genomic DNA samples isolated from mice ear punches were genotyped based on the method previously described (28). Mice with different genotypes (wild type [WT] and HET-A20^+/− heterozygote) were studied at 10–12 wk of age. All experiments included age- and gender-matched animals. The animals used belonged to the same litters and were cohoused.

Plasmid construction

Plasmid construction was performed following standard protocols from molecular cloning. Briefly, pCDH-GFP plasmid DNA were digested with NotI (New England Biolabs) and XbaI (New England Biolabs). The A20 PCR fragments were amplified with forward primer (5′-TGCTCTAGAATGGATTACAAGGATGACGAC-3′) and reverse primer (5′-TTTGCGGCCGCTTAGCCATACATCTGCTTGAAC-3′) with template of pCMV-Tag2-Flag-A20 (a gift from Dr. A. Ma) using Expand Long Template PCR System (Roche) following the user manual. The PCR product plasmid fragments were ligated with T4 DNA ligase (New England Biolabs). The colonies with correct insertion were sequenced, and the insertions were confirmed by the National Center for Biotechnology Information Basic Local Alignment Search Tool. lentiCRISPR v2 was a gift from Feng Zhang (plasmid no. 52961; Addgene). The single-guide RNA (sgRNA) sequence was determined with the guidelines from the Zhang laboratory, and the plasmids vectors were constructed by protocol provided by the Zhang laboratory (29). Briefly, the synthetic DNA oligonucleotides (sgRNA sequence: 5′-TTCCAGTGTGTATCG GTGCA-3′) were ligated to the vector digested with BsmBI (Fermentas Life Sciences) and transformed into One Shot Stbl3 Chemically Competent E. coli (Invitrogen). The sequences were confirmed by the National Center for Biotechnology Information Basic Local Alignment Search Tool. pRSV-Rev (plasmid no. 12253; Addgene), pMDLg/pRRE (plasmid no. 12251; Addgene), and pMD2.G (plasmid no. 12259; Addgene) were gifts from Prof. Didier Trono (Ecole Polytechnique Federal de Lausanne, Lausanne, Switzerland).

Lentivirus preparation and infection of THP-1 cells

HEK293T cells were seeded at density of 1 × 10⁶ cells per 10-cm dish and transfected with pRSV-Rev, pMDLg/pRRE, pMD2.G, and pCDH-GFP or pCDH-A20 plasmids (for overexpression), or lentiCRISPR–sgRNA or empty vector (lentiCRISPR v2) (for A20 knockdown). Forty-eight hours after the transfection, the media were collected and incubated with THP-1 cells (1 × 10⁵) for∼12 h. Twenty-four hours postinfection, the cells were selected with puromycin (1 μg/ml) and pooled together to be used for ELISA and Western blot.

Real-time quantitative PCR

The RNAs were prepared with RNeasy Plus Mini Kit (QIAGEN) with QIAcube (QIAGEN). The RNA concentrations were determined with NanoDrop, and 800 ng of total RNA was used for cDNA synthesis with High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) following the user manual. The quantitative real-time PCR (qRT-PCR) was performed with the Applied Biosystems 7500 using SYBR Green Master Mix (SABiosciences). Primers used for qRT-PCR were as follows: hGAPDH forward: 5′-CAATGACCCCTTCATTGACC-3′, hGAPDH reverse: 5′-TTGATTTTG GAGGGATCTCG-3′; hA20 forward: 5′-TTGTCCTCAGTTTCGGGAG AT-3′, hA20 reverse: 5′-ACTTCTCGACACCAGTTGAGTT-3′; mGAPDH forward: 5′-AGGTCGGTGTGAACGGATTTG-3′, mGAPDH reverse: 5′-GGGGTCGTTGATGGCA ACA-3′; mIL-6 forward: 5′-TCTATACCACTTCACAAGTC GGA-3′, mIL-6 reverse: 5′-GAAT TGCCATTGCACAACTCTTT-3′; mTNF forward: 5′-CTGAACTTCGGGGTGAT CGG -3′, mTNF reverse: 5′-GGCTTGTCACTCGAATTTTGAGA-3′; mIL-17 forward: 5′-GGCCCTCAGACTACCTCAAC-3′, mIL-17 reverse: 5′-TCTCGACCCTGAAAGTGA AGG-3′; and mIL-23 forward: 5′-AATAATGTGCCCCGTATCCAGT-3′, mIL-23 reverse: 5′-GCTCCCCTTTGAAGATGTCAG-3′.

Western blot

Western blot was performed following standard protocol with minor modifications (26). Briefly, the cells were harvested and rinsed with precold PBS followed by lysing with RIPA Buffer (Bio-Rad Laboratories) or 1× SDS loading buffer. The protein concentrations were determined with a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Samples were then loaded to Mini-PROTEAN TGX Precast Protein Gels (Bio-Rad Laboratories), and electrophoresis was performed. The gels were then transferred to PVDF membranes (Millipore) and followed by Ab incubation according to the user manuals. The Abs used were as follows: anti-TNFAIP3 (A20) Ab (D13H3, 5630S, 1:1000; Cell Signaling Technology), anti–β-Actin Ab (C4, sc-47778, 1:1000; Santa Cruz Biotechnology), anti–β-tubulin Ab (D-10, SC-5274, 1:500; Santa Cruz Biotechnology), anti–NF-κB inhibitor α (IκBα) Ab (C21, SC-371, 1:1000; Santa Cruz Biotechnology), and anti-p-NF-κB p65 Ab (27. Ser536, SC-136548, 1:500; Santa Cruz Biotechnology). Scanned blots were processed with an ImageJ program, version 1.4.3.67, and band intensities of bands were measured. Data analysis was performed using Microsoft Excel 2010 and GraphPad Prism 6.

ImageStream quantitation of p65 nuclear translocation

A20-overexpressing cells, A20 knockdown and control THP-1 cells (1 × 10⁷) infected with P. gingivalis or uninfected as control, were harvested and fixed in 4% paraformaldehyde and stored overnight at 4°C. Following washing in permeabilization wash buffer (PWB) consisting of 3% FCS and 0.02% Triton X-100 in PBS, the cells were labeled with p65 Ab and incubated for 30 min at room temperature and washed again in PWB buffer. The cells were then incubated with Alexa Fluor 488 goat anti-mouse IgG Ab and DAPI at room temperature in the dark for 45 min and washed three times with PWB buffer. Following the final wash, the cells were resuspended in 50 μl of PBS prior to imaging by an ImageStream Mark II Flow Cytometer (30). Analyses were performed using IDEAS software (Amnis, Seattle, WA). The Abs used in the assays were as follows: p65 (SC-136548, 1:25; Santa Cruz Biotechnology), anti-mouse IgG–Alexa Fluor 488 (A11001, 1:200; Thermo Fisher Scientific), and DAPI (46190, 1:400; Thermo Fisher Scientific).

Ligature-induced periodontitis model

Periodontal inflammation and bone loss were induced by ligature placement following protocols published previously (26). Briefly, A20 ^+/+ (n = 17) and A20 ^+/− mice (n = 16) (males, littermates) were anesthetized i.p. with a 200 μl mixture of ketamine (10 mg/ml) and xylazine. Black-braided silk 5-0 threads were placed at the left side of the maxilla interdentally between the first and second maxillary molars, and the right sides were left unligated as controls. The mice were euthanized 7 d after ligature placement. Microcomputed tomography (micro-CT) (SkyScan 1173; Bruker, Kontich, Belgium) imaging was used to determine alveolar bone loss (26). Skulls used for micro-CT were stored and fixed in 10% neutral-buffered formalin for at least 24 h prior to imaging. Samples were scanned at a resolution of 1120 × 1120 pixels (image pixel size of 15.82 μm) over 180°, 80 kV voltage, 80 μA current, and 250 ms exposure time. Five x-ray projections were acquired every 0.2° and averaged. A standard Feldkamp reconstruction was done using NRecon software (Bruker) with a beam-hardening correction of 15% and a Gaussian-smoothing kernel of 1.

Immunohistochemistry of mouse gingival tissues

Explanted gingival tissues from A20 ^+/+ (n = 6) and A20 ^+/− (n = 6) mice (males, littermates) were fixed in 4% formaldehyde, sectioned, and stained with H&E following standard protocol. Histology images were acquired using Q-Color 5 Imaging System from Olympus Microscopy with a 10× magnification objective lens. Quantification of numbers of nucleated cells (hematoxylin-positive) in the gingival was performed by a blinded examiner using cellSens software at 20× magnification. Counting was performed using the “Count and Measure” tool and the “Manual Threshold” option to choose an initial nucleated cell for reference. Once the original cell was selected, subsequent cells were automatically selected until all nucleated cells were highlighted. Data were reported as area stained (square micrometers) in each field of view. Nine regions/fields of view were analyzed per ligated side, and six regions/fields of view were analyzed per control side, per mouse, respectively.

Statistical analysis

Data were analyzed by one-way ANOVA and the Tukey multiple-comparison test or unpaired t test with Mann–Whitney correction using the InStat program (GraphPad Software, San Diego, CA). Each experiment was repeated at least three times in triplicate, and a p value <0.05 was considered significant.

Results

A20 expression increased in human macrophages upon P. gingivalis infection

Increased A20 expression, which is mainly triggered by NF-κB activation and subsequent cytokine production, restricts intracellular inflammatory signals through targeting multiple substrates residing downstream of innate sensors, including TLRs. TLRs are expressed by various cells of the periodontium, including macrophages, and respond to oral microbiota, thereby directly contributing to the pathophysiology of periodontitis. A20 is expressed within the gingival connective tissue (26). To determine the function of A20 in the oral mucosa, we used an oral bacteria-induced disease model and assessed A20 expression in human macrophage-like cells (THP-1) following P. gingivalis infection. Briefly, the level of A20 mRNA and protein were determined following P. gingivalis infection using qRT-PCR and Western blot, respectively. THP-1 cells displayed increased A20 mRNA expression as early as 1 h upon P. gingivalis infection, which peaked at 3 h and sustained at high levels up to 12 h (Fig. 1A). A20 protein levels were also significantly increased following bacterial infection, which peaked at 3 h and remained high up to 12 h (Fig. 1B, 1C). These results reveal the pattern of A20 expression in macrophages following an oral microbial infection.

FIGURE 1.

A20 expression is increased in human macrophages upon P. gingivalis infection. THP-1 cells were infected with P. gingivalis (100 MOI) for the indicated time, and cells were harvested and applied to qRT-PCR (A) and Western blot (B), and the protein levels were plotted (C). The blots are derived from the same protein samples. Results are representative of a minimum of three independent experiments with averages and SD shown. All through the manuscript, blots within one panel are derived from the same protein samples.**p ≤ 0.01, ****p ≤ 0.0001.

Loss of A20 instigates increased inflammatory cytokine production in human macrophages following P. gingivalis infection

Macrophages are considered one of the key cell types in the milieu of periodontal inflammation. Activated macrophages produce inflammatory mediators that recruit various immune cells to confine infection. Failure to terminate the inflammatory response in a timely manner results in periodontal tissue destruction. To determine if A20 serves as one of the downstream regulatory molecules in oral pathogen-induced inflammation within the oral mucosa, our next set of experiments employed loss-of-function assays in which an A20 knockdown system was constructed in THP-1 cells followed by oral bacterial infection. THP-1 cells transduced with lentivirus containing the sgRNA-targeting A20 gene displayed diminished A20 protein levels compared with the control cells (Fig. 2A). As expected, A20 depletion triggered significantly increased IL-6 and TNF production following P. gingivalis infection compared with control cells (Fig. 2B, 2C). These results revealed that A20 deficiency instigates elevated proinflammatory cytokine production in human macrophages following oral microbial infection.

FIGURE 2.

Loss of A20 instigates increased inflammatory cytokine production in human macrophages following P. gingivalis infection. THP-1 cells were infected with lentivirus containing sgRNA-targeting A20 (sgRNA-A20) or empty vector (empty), and A20 protein levels were determined with Western blot (A). The cells were infected with P. gingivalis (100 MOI) for 6 h. IL-6 (B) and TNF (C) levels in the supernatants were determined with ELISA. Results are representative of a minimum of three independent experiments analyzed in triplicates with averages and SD shown. ***p ≤ 0.001.

A20 overexpression inhibits cytokine production in human macrophages following P. gingivalis infection

To further verify the role of A20 in inflammatory responses in the oral cavity we performed gain-of-function assays where the A20-overexpressing system was constructed in THP-1 cells through transductions with lentivirus expressing A20 or GFP (as control). Western blot analysis confirmed increased A20 protein levels in cells transduced with lentivirus expressing A20 compared with control cells (Fig. 3A). Specifically, there was a 1.5- to 2-fold increase in basal A20 levels in genetically engineered cells compared with control cells. It is possible that lentiviral transduction itself can activate NF-κB signaling pathway (31). Following bacterial infection, A20 expression was increased 3-fold in A20-overexpressing and control cells and 1.5-fold in GFP-expressing cells compared to basal levels. Although GFP-expressing control cells displayed relatively less A20 induction in response to bacteria (28% less compared with bacteria-infected control cells), this was not related to any significant phenotypic differences among cells. Macrophages with enhanced A20 expression produced significantly less IL-6 and TNF upon P. gingivalis challenge compared with control cells (Fig. 3B, 3C). Corroborating the results obtained with loss-of-function assays, these data demonstrate that A20 overexpression impedes oral pathogen-induced proinflammatory cytokine production in macrophages.

FIGURE 3.

A20 overexpression inhibits cytokine production in human macrophages following P. gingivalis infection. THP-1 cells were infected with lentivirus expressing A20 (A20) or GFP; A20 protein levels were determined with Western blot (A). The cells were infected with P. gingivalis (100 MOI) for 6 h. IL-6 (B) and TNF (C) levels in the supernatants were determined with ELISA. Results are representative of a minimum of three independent experiments analyzed in triplicates with averages and SD shown. ****p ≤ 0.0001.

A20 regulates inflammatory response to P. gingivalis infection through interfering with NF-κB signaling

Through a rigorous experimental approach, which employed genetic manipulation of A20, we established unequivocally that A20 is required to repress proinflammatory cytokine production following oral bacterial infection in human macrophages. The NF-κB signaling pathway is one of the most important pathways mediating inflammation and also plays an essential role in periodontal disease pathogenesis through activating cytokine genes. To determine if A20 exerts its effect on oral bacteria-induced inflammatory response through the NF-κB signaling pathway, we monitored the degradation of IκBα, a hallmark of NF-κB activation and p65 phosphorylation, and nuclear translocation in A20-competent and -altered cells using immunoblotting and ImageStream. Following cellular stimulation by extracellular stimuli, IκBα is phosphorylated and then subsequently degraded in the cytoplasm. The eventual reaccumulation of IκBα drives the inhibition of NF-κB, leading to decreased cytokine production (32–35). Hence, we first measured IκBα protein levels at different times for up to 6 h following P. gingivalis infection. The results showed rapid degradation of IκBα as early as 15 min followed by its recovery at later time points (Figs. 4A, 4B, 5A, 5B). This pattern was consistent with the established models of stimulus-induced degradation of IκBα followed by its reaccumulation and indicated NF-κB activation in response to P. gingivalis (32–35). The analyses of the protein levels at 3 and 6 h postinfection revealed increased IκBα degradation and decreased reaccumulation in the A20-depleted cells compared with the A20-competent cells, which indicated increased NF-κB activity in the A20-deficient cells (Fig. 4A, 4B). To further confirm our findings, we also assessed p65 nuclear translocation with ImageStream technology, which combines high-resolution digital imaging with quantitative flow cytometry technology. The data were analyzed with the IDEAs software “Similarity” feature to determine the overlap (or similarity) of staining of NF-κB (p65) with nuclear staining (DAPI) so that a high correlation of NF-κB/DAPI localization is reflected by a high similarity score (30, 36–39). In addition, the percentage of cells exhibiting p65 nuclear translocation was also determined based on having similarity scores >0.5 (denoted by the orange line in the graphs) (Figs. 4C, 5C). A higher score of p65/DAPI similarity indicates a higher percentage of overlap between p65 and DAPI, suggesting an increased level of p65 nuclear translocation. The analysis revealed negative similarity scores at baseline both for A20-depleted and A20-competent cells, which corresponded to the absence of nuclear NF-κB (p65) translocation (Fig. 4C). Corroborating the results obtained with immunoblotting, a higher percentage of p65 translocation was noted in A20-depleted cells versus A20-competent cells after bacterial challenge (64.7% for A20-depleted and 21.3% for A20-competent at 15 min postinfection; 80.3% for A20-depleted and 67.1% for A20-competent at 30 min postinfection) (Fig. 4D, 4E). To further verify NF-κB activity, we also determined the levels of phosphorylated p65 and observed increased p65 phosphorylation in A20-depleted cells compared with A20- competent control cells at 15 and 30 min following P. gingivalis infection (Fig. 4F). It is also of note that depleted cells displayed slightly increased basal levels of phosphorylated p65, whereas no increase was detected in control cells containing the lentivirus with empty vector. This increase, again, could likely be the effect of lentivirus transduction possessing a slight stimulatory effect on NF-κB signaling (31). Yet, increased p65 phosphorylation, noted only in A20-depleted cells but not in A20-competent control cells, suggests that basal expression of A20 restricts the activation of NF-κB signaling pathway. These results collectively confirmed increased NF-κB activity associated with A20 deficiency. We further assessed whether overexpression of A20 had an effect on NF-κB signaling following oral bacterial infection. As expected, exogenous expression of A20 interfered with IκBα degradation (Fig. 5A, 5B) and p65 nuclear translocation (Fig. 5C−E) following P. gingivalis infection. Our results indicated earlier recovery of IκBα at 30 min in A20-overexpressing cells compared with WT cells, which displayed a delayed reaccumulation at 3 h (Fig. 5A, 5B). Analyses of p65 translocation using ImageStream also showed consistent results, in which all cells exhibited negative similarity scores at baseline (Fig 5C). Further analyses revealed a lower percentage of p65 translocation to the nucleus in the A20- overexpressing cells compared with the control cells (9.43% for A20-overexpressing and 29.6% for control cells at 15 min postinfection; 52.1% for A20-overexpressing and 75.5% for control cells at 30 min postinfection) (Fig. 5C–E). Substantiating the results, decreased levels of phosphorylated p65 were noted in A20-overexpressing cells compared with control cells at 15, 30 min, and 1 h after P. gingivalis infection (Fig. 5F), suggesting A20 inhibition of p65 phosphorylation. Taken together, our results indicate that A20 regulates oral bacteria-induced inflammatory cytokine response through its effect on NF-κB signaling pathway.

FIGURE 4.

Knockdown of A20 increases IκBα degradation and promotes p65 nuclear translocation. THP-1 cells were infected with lentivirus containing sgRNA-targeting A20 (sgRNA) or empty vector (empty) and stimulated with P. gingivalis (100 MOI) for indicated time courses. IκBα protein levels were detected with Western blot (A and B). THP-1 cells were stimulated with P. gingivalis (5 MOI) for 0, 15, or 30 min. Cells were collected and stained with p65 Ab and DAPI. Images were collected on the ImageStream, and graphs reflect the percentage of cells with similarity defined as p65 translocated and colocalizing with the DAPI. Graphs show the percentage of cells with a similarity score >0.5 as represented by the orange line (C). Representative cell images using ImageStream (original magnification ×60) (D). The percentage of p65 nucleus colocalization is plotted (E). Results are representative of a minimum of three independent experiments. A20- depleted and control cells were stimulated with P. gingivalis (100 MOI) for indicated time points, and cells were lysed with 1× SDS loading buffer. The lysates were applied to Western blot, detecting the protein levels of phosphorylated p65 (F).

FIGURE 5.

Exogenous A20 decreases IκBα degradation and p65 nuclear translocation. THP-1 cells infected with lentivirus expressing A20 (A20) or GFP were stimulated with P. gingivalis (100 MOI) for indicated time courses. IκBα protein levels were detected with Western blot (A and B). THP-1 cells (Control) and A20-overexpressing cells (A20) were stimulated with P. gingivalis (5 MOI) for 0, 15, or 30 min. Cells were collected and stained with p65 Ab and DAPI. Images were collected on the ImageStream, and graphs reflect the percentage of cells with similarity defined as p65 translocated and colocalizing with the DAPI. Graphs show the percentage of cells with a similarity score >0.5 as represented by the orange line (C). Representative cell images are shown using ImageStream (original magnification ×60) (D). The percentage of p65 nucleus colocalization is plotted (E). Results are representative of a minimum of three independent experiments. A20-overexpressing and control cells were stimulated with P. gingivalis (100 MOI) for indicated time points. The cells were lysed with 1× SDS loading buffer, and the lysates were applied to Western blot, detecting the protein levels of phosphorylated p65 (F).

A20 deficiency increases oral pathogen-induced cytokine production and NF-κB activity in murine BMDMs

Our next set of experiments included ex vivo periodontal disease models using A20-deficient transgenic mice to assess A20 function in murine cells. BMDMs derived from transgenic mice deficient in A20 (A20^+/−) and A20 WT mice (A20^+/+) were challenged with P. gingivalis for up to 12 h, and the cytokine production in cell culture supernatants was determined with ELISA; the degradation of IκBα in the cell extracts was monitored at different time points with Western blot. Similar to the results obtained with human cells, A20-deficient (A20^+/−) murine macrophages produced significantly increased IL-6 and TNF compared with the cells derived from A20 WT mice (A20^+/+) (Fig. 6A, 6B). We also detected increased degradation of IκBα in BMDMs derived from A20^+/− mice compared with those from the A20^+/+ mice at 30 min to 6 h postinfection, indicating elevated NF-κB activation in A20-deficient cells (Fig. 6C, 6D). Overall, the preclinical studies using human and murine cells confirmed that A20 can restrain inflammatory response to oral bacteria through modulating NF-κB signaling pathway.

FIGURE 6.

A20 deficiency increases cytokine production and NF-κB activity following P. gingivalis infection in murine bone marrow–derived macrophages (BMDMs). BMDMs isolated from A20^+/+ and A20^+/− mice were stimulated with P. gingivalis (100 MOI) for 12 h, and the supernatant was harvested and applied to ELISA for IL-6 (A) and TNF (B) levels. BMDMs isolated from A20^+/+ and A20^+/− mice were infected with P. gingivalis (100 MOI) for indicated time courses, and IκBα protein levels were detected with Western blot (C and D). Results are representative of a minimum of three independent experiments analyzed, with averages and SD shown. ***p ≤ 0.001, ****p ≤ 0.0001.

Partial A20 loss instigates increased gingival inflammation and alveolar bone loss compared with A20-competent mice

Genetic deletion of A20 in the mouse (A20^−/−) results in severe multiorgan inflammation and premature death within 3–4 wk after birth, whereas the A20^+/− mouse is viable and normal in size and displays no gross physiological or behavioral abnormalities (16). We observed increased NF-κB activity and cytokine production in the cells isolated from the A20^+/− mouse upon oral bacterial infection ex vivo compared with those from the A20^+/+ mice (Fig. 6). This observation prompted us to hypothesize that even a partial loss of A20 may lead to an overt inflammatory phenotype within the oral mucosa in vivo. To test this hypothesis, we used a ligature-induced periodontitis model in an A20 haploid-deficient mouse (A20^+/−) and WT mice (A20^+/+). Briefly, silk sutures were placed between maxillary left molars, and right teeth were left unligated as baseline control in two groups of animals (A20^+/+ and A20^+/−). The alveolar bone loss was determined by measuring the distance between the cemento-enamel-junction and alveolar bone crest using micro-CT imaging at day 7 postligature placement by a blinded examiner. As expected, ligated sides of both A20^+/+ and A20^+/− mice displayed significantly increased bone loss compared with the unligated control sides (Fig. 7A, 7B). However, the ligated teeth of A20^+/− mice exhibited more bone loss compared with the ligated teeth in WT mice. We further assessed immune cell infiltration in the dissected gingival tissues of each group of mice using H&E staining. In line with the bone loss data, ligated sites from both WT (A20^+/+) and A20-deficient (A20^+/−) mice exhibited increased inflammatory cell infiltration in which there was slightly more inflammatory cell infiltration observed in the gingival tissues obtained from A20^+/− mice compared with WT counterparts (Fig. 7C, 7D). We subsequently determined cytokine expression in the gingival tissues to further elucidate the effect of A20 loss in inflammatory responses in the oral mucosa in vivo. Whereas the expression of IL-6, TNF, IL-17, and IL-23 were upregulated in both A20^+/+ and A20^+/− ligated sites compared with the unligated sites, consistent with the ex vivo data, the gingival tissues with partial loss of A20 (A20^+/−) exhibited significantly elevated cytokine expression compared with their A20-competent counterparts (Fig. 7E–H). Overall, the results of in vivo studies support the ex vivo data and serve as a proof of concept substantiating the role of A20 as a key regulatory molecule for restraining inflammation and maintaining tissue homeostasis in the oral mucosa (Fig. 8).

FIGURE 7.

Partial A20 loss instigates increased gingival inflammation and alveolar bone loss compared with A20-competent mice. Silk 5-0 threads were placed at the left side of the maxilla interdentally between the first and second maxillary molars, and the right sides were left unligated as controls on A20^+/+ and A20^+/− mice. The bone loss was determined by micro-CT (A), and the cemento-enamel-junction–alveolar bone crest was analyzed and plotted (B) (n = 17 for A20^+/+, n = 16 for A20^+/−). Averages and SD are shown. Gingival tissues derived from ligated or control periodontium were applied to H&E staining (C) and observed under microscope, and the percentage of inflammatory cells in the periodontium were calculated and plotted (D). Data are representative of three independent experiments, with a minimum of four mice analyzed in each group per experiment. The gingival tissues derived from ligated or control periodontium were digested and applied to RNA preparation. The mRNA levels of mIL-6 (E), mTNF (F), mIL-17 (G), and mIL-23 (H) were determined with qRT-PCR, and the relative mRNA expression levels were plotted. Data are representative of three independent experiments analyzed, with averages and SD shown. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

FIGURE 8.

A representative model of NF-κB/A20/cytokine axis in the maintenance of oral mucosal tissue homeostasis. Dysregulated immune responses to a diverse group of resident oral microflora elicits persistent inflammation, disruption of tissue homeostasis, and eventually periodontal disease, which is characterized by the destruction of gingival tissue and bone surrounding the teeth. The maintenance of tissue homeostasis requires the presence of fully functional regulatory pathways in place to control both timely initiation and termination of inflammation. Initiation of inflammation in the oral mucosa is mainly driven by the activation of innate sensors and NF-κB signaling cascade and subsequent release of inflammatory cytokines. Briefly, microbe-induced activation of innate sensors (e.g., TLRs and nucleotide-binding oligomerization domain–like receptors [NLRs]) triggers the recruitment of the adaptors such as MyD88 and TRIF to activate IRAK, TRAF6, and TAK1 complexes. Subsequent ubiquitination of TRAF6 results in the phosphorylation and degradation of IκBα, leading to the nuclear translocation of NF-κB and transcription of genes that encode the expression of proinflammatory cytokines. Activation of NF-κB signaling and initial influx of inflammatory cytokines trigger increased A20 expression as part of negative feedback loop. A20 subsequently facilitates timely termination of NF-κB signaling by deubiquitinating TRAF6, thereby supporting the resolution of inflammation and maintenance of oral tissue homeostasis, whereas A20 deficiency results in sustained inflammatory response and collateral tissue damage because of prolonged release of cytokines and tissue-destructive enzymes.

Discussion

Chronic periodontitis is the clinical manifestation of deregulated immune responses within the oral cavity, which ensues from the disruption of tissue homeostasis because of prolonged release of proinflammatory cytokines and tissue-destructive enzymes in response to a dysbiotic resident microflora (40). It is now widely recognized that therapies targeting the modulation of host immune responses offer promising options to fine-tune unrestrained periodontal inflammation and restore tissue homeostasis (41, 42). The studies conducted over the last decade defined interactions between oral microbe-associated molecular patterns and TLRs and established key inflammatory pathways such as NF-κB, which initiate periodontal inflammation (1–3). Yet, we still do not know much about the downstream signaling pathways and molecules that promote timely termination of the inflammatory response in the oral mucosa. In the current study, through a series of experiments involving well- established preclinical periodontal disease models, we provided several lines of evidence revealing that A20 mitigates inflammatory responses to resident oral bacteria by interfering with NF-κB signaling and cytokine production and acts as a gatekeeper of inflammation in the oral mucosa (Fig. 8). To our knowledge, this is the first investigation that highlights the involvement of a ubiquitin-editing molecule in governing the oral cavity tissue homeostasis. First, A20 depletion significantly increased NF-κB activity and proinflammatory cytokine production in response to P. gingivalis infection both in human and murine cells. Consistently, gain-of-function assays in which A20 was overexpressed using lentivirus infection resulted in diminished NF-κB activity and cytokine production. Subsequent experiments using a murine ligature-induced model of periodontitis showed, to our knowledge, for the first time that even a partial loss of A20 modulates the oral cavity microenvironment and promotes an increased inflammatory phenotype, further verifying the participation of A20 in the biological responses in the periodontal tissues. Ultimately, our findings reveal a unique physiological role for A20 as one of the key regulatory molecules of NF-κB signaling in the oral mucosa. This discovery sheds new light on the pathogenesis of periodontitis and opens up new areas of investigation to assess the potential for A20-targeted therapies to alleviate adverse clinical outcomes, both at local and distant tissues.

Emerging evidence suggests that host sensing of commensal microflora is critical in the maintenance of tissue homeostasis where A20 serves as one of the endogenous regulators mainly functioning downstream of NF-κB signaling. Corroborating this, elimination of MyD88, which is an adaptor molecule of TLR signaling, alleviates spontaneous inflammation and death in A20-deficient mice (A20^−/−), confirming the crucial role of A20 in the host–microbiome interface (16). Hitherto, there was a lack of knowledge about the regulatory role for A20 in the oral mucosa as it relates to specific cell and microbiome interactions. In this study, we filled this gap with a dissected cell-specific function for A20 in human and murine cells using P. gingivalis as a model oral microorganism. Our results indicated that A20 acts as a negative regulator of inflammation through modulating NF-κB and, subsequently, cytokine production in macrophages upon oral microbial infection. These findings are consistent with the previous studies, which assessed the function of A20 in other tissues and cells (43, 44). For example, A20 has been identified as one of the key regulators of intestinal homeostasis, in which it functions to limit spontaneous activation and expansion of myeloid cells while protecting barrier stability of the intestinal epithelium by preventing epithelial cell apoptosis during inflammation (45). In the intestine, severe inflammatory response associated with A20 deficiency was reported to be strongest in the colon, where the bacteria were more abundant. In another study, host inflammatory responses to Mycobacterium fortuitum infection were also shown to be negatively regulated through TLR2-dependent A20 expression in murine macrophages (46). Notably, it was reported that the rough morphotype of the clinical strain triggered increased proinflammatory signaling activation and less A20 induction in BMDMs compared with the responses observed with smooth morphotype, suggesting variations among different bacteria in facilitating A20 induction. In fact, there are also reports of specific bacteria and probiotics promoting A20 expression (47, 48). In this study, we used P. gingivalis, which is one of the key stone periodontal pathogens, as the model microorganism to mimic bacteria-induced oral inflammation. Considering the diversity of the oral microbiome, it will be crucial to determine how certain bacteria or groups of bacteria induce A20 expression and whether they support A20 sustainability in the oral mucosa.

In this study, we used macrophages to assess the function of A20 function in oral bacteria-induced inflammatory responses. Of note is that A20 serves as a key regulatory molecule in several other cell types. For example, both airway and intestinal epithelial cells were reported to display increased A20 expression in a time- and dose-dependent manner following LPS stimulation (22, 49–51). In a sepsis model, dendritic cell–specific A20 was reported to preserve immune homeostasis in steady-state conditions and induce LPS tolerance (52). Moreover, it has been reported that A20 may be essential for immune cell differentiation and maturation and also functions to prevent autoimmunity by restricting B cell survival (53–55). Considering that periodontal tissues include both myeloid and nonmyeloid cells possessing unique functions in facilitating physiological responses and during the course of infectious challenge, it is yet to be determined how A20 functions in other cells of the periodontium besides macrophages.

Clinically, A20 is expressed both in the gingival epithelial and connective tissues (26). Interestingly, despite elevated mRNA levels, A20 protein levels do not increase in periodontitis lesions, suggesting that A20 protein levels may not be sustained in the oral mucosa, likely because of posttranscriptional regulation (26). In fact, patients with cystic fibrosis, a disease in which there is prolonged NF-κB–mediated inflammation in multiple organs, also exhibit similar characteristics. The majority of the deaths in cystic fibrosis are mainly attributed to the development of excessive pulmonary inflammation because of microbial perturbations, such as Pseudomonas aeruginosa infection. It has been shown that, compared with the alveolar epithelial cells from healthy subjects, the cells from cystic fibrosis patients display lower basal A20 expression, implicating impaired A20 functionality with the excessive inflammation in these patients (56). In another clinical study, endobronchial biopsy specimens obtained from the patients with asthma showed that the level of A20 was negatively correlated with the severity of the disease (57). It is, therefore, likely that impaired A20 levels may lead to persistent inflammation in the oral mucosa as well. Indeed, in this study using a murine model of periodontitis and A20 haploinsufficient mice (A20^+/−), we were able to show that even a partial loss of A20 enhances inflammation in the gingival tissues and results in a more severe periodontal disease phenotype as determined by significantly more bone loss and gingival tissue cytokine expression in A20-deficient mice (A20^+/−) versus A20-competent mice (A20^+/+). Similar observations were noted in other conditions as well. For example, A20 haploinsufficiency has been implicated in autoimmune lymphoproliferative syndrome and early onset of autoinflammatory disease, indicating the role of A20 in disease susceptibility (58, 59). In another report, mice lacking A20 in enterocytes were reported to develop an overwhelming systemic inflammation following treatment with TNF (60). Similarly, decreased A20 expression was noted in the bone marrow mesenchymal stem cells from rheumatoid arthritis patients, in which A20 was shown to regulate Th17/regulatory T cell balance (61). IL-17 is one of the pivotal cytokines in the periodontal disease pathologic condition and it was also the most highly expressed cytokine in the gingival tissues of partial A20-deficient mice (A20^+/−) that received ligatures in this study (62). Thus, it’s likely that IL-17 signaling could be one of the key pathways governing the inflammatory responses associated with A20 deficiency in the oral mucosa and needs further investigation.

Collectively, the results of this study serve as a proof of concept and support the notion that A20 is necessary to sustain tissue homeostasis in the oral mucosa. One of the questions that still remains to be answered is how A20 levels are regulated in the oral mucosa (26). This could be related to many factors, as impaired A20 levels because of genetic, epigenetic, or various other environmental factors have been associated with unsustainable inflammation and tissue damage in several conditions (57). For example, polymorphisms in the A20 gene are associated with increased susceptibility to rheumatoid arthritis, lupus, systemic sclerosis, Crohn disease, and psoriasis (63–66). Furthermore, A20 expression is regulated by multiple microRNAs (67–74). Intriguingly, some of these microRNAs are found to be differentially expressed in periodontitis lesions, which can possibly be a mechanism to modulate A20 levels in the oral mucosa, which needs to be further explored (75). Thus, future studies are warranted to characterize the cell-specific regulation of A20 through clinical studies and using a lineage-specific knockout mouse to identify susceptible individuals and develop more targeted therapeutics.

In summary, NF-κB signaling is considered as the central regulator of gene transcription in the immune system and is also involved in numerous biological processes in the milieu of periodontitis, which is one of the most common chronic inflammatory diseases (76, 77). Although A20 function is fairly well-characterized in various other sites (e.g., GI tract and alveolar tissues) as one of the essential regulators of NF-κB signaling pathway in the resolution of inflammation, only recently did it start to receive attention in the pathologies affecting oral mucosal tissues (78, 79). There are notable differences between the oral mucosa and those other tissues, such as the diversity of the oral microbiota, the composition of the cellular network forming the gingival epithelial and connective tissues, and bone structure supporting the dentoalveolar unit. In addition, unlike other sites of the body, there is continuous and direct exposure of oral tissues to physical and chemical stimuli through regular daily activities, which creates a unique environment with an ongoing external stress. Thus, sustaining tissue homeostasis and integrity requires a delicate balance between the host response, microbiome, and outside stress. It is, therefore, important to dissect how key inflammatory pathways are regulated at a cellular and molecular level to maintain oral mucosal tissue homeostasis and affect disease phenotype through using disease-specific models that can mimic this distinct and dynamic network of the oral cavity microenvironment. This will eventually lead to development of targeted therapies with maximum efficacy and minimal side effects. In this study, we systematically characterized the function of A20 in relation to host–microbiome interactions in the oral cavity and revealed A20 as a novel regulator of periodontal inflammation downstream of NF-κB signaling using both in vitro and in vivo periodontal disease models. Our study provides critical evidence in support of the previous clinical observations, which reported insufficient A20 sustainability during the course of periodontitis and serve as a proof of concept for future translational studies to assess the implications of controlling NF-κB signaling through targeting A20 in the oral cavity and possibly in other oral inflammation-related conditions at distant tissues.

Disclosures

The authors have no financial conflicts of interest.