Skip to main content
PMC home page
Journal of Dental Research logoLink to Journal of Dental Research
. 2018 Feb 8;97(7):810–819. doi: 10.1177/0022034518755688

Smad6 Methylation Represses NFκB Activation and Periodontal Inflammation

T Zhang 1,2, J Wu 2, N Ungvijanpunya 2, O Jackson-Weaver 2, Y Gou 2, J Feng 2, TV Ho 2, Y Shen 3, J Liu 3, S Richard 4, J Jin 3, G Hajishengallis 5, Y Chai 2, J Xu 2,✉
PMCID: PMC6728583  PMID: 29420098

Abstract

The balance between pro- and anti-inflammatory signals maintains tissue homeostasis and defines the outcome of chronic inflammatory diseases such as periodontitis, a condition that afflicts the tooth-supporting tissues and exerts an impact on systemic health. The induction of tissue inflammation relies heavily on Toll-like receptor (TLR) signaling, which drives a proinflammatory pathway through recruiting myeloid differentiation primary response gene 88 (MyD88) and activating nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB). TLR-induced production of proinflammatory cytokines and chemokines is reined in by anti-inflammatory cytokines, including the transforming growth factor β (TGFβ) family of cytokines. Although Smad6 is a key mediator of TGFβ-induced anti-inflammatory signaling, the exact mechanism by which TGFβ regulates TLR proinflammatory signaling in the periodontal tissue has not been addressed to date. In this study, we demonstrate for the first time that the ability of TGFβ to inhibit TLR-NFκB signaling is mediated by protein arginine methyltransferase 1 (PRMT1)–induced Smad6 methylation. Upon methylation, Smad6 recruited MyD88 and promoted MyD88 degradation, thereby inhibiting NFκB activation. Most important, Smad6 is expressed and methylated in the gingival epithelium, and PRMT1-Smad6 signaling promotes tissue homeostasis by limiting inflammation. Consistent with this, disturbance of Smad6 methylation exacerbates inflammation and bone loss in experimental periodontitis. The dissected mechanism is therapeutically important, as it highlights the manipulation of PRMT1-Smad6 signaling as a novel promising strategy to modulate the host immune response in periodontitis.

Keywords: signal transduction, post-translational modification, periodontal disease, cell signaling, protein-protein interaction, arginine methylation

Introduction

Periodontitis is a chronic inflammatory condition that leads to the destruction of the tooth-supporting hard and soft tissues (Darveau 2010; Hajishengallis 2014, 2015). It is increasingly recognized that the host’s inflammatory responses, controlled by a network of signaling pathways, define the clinical outcome of periodontitis (Darveau 2010; Hajishengallis 2015). Deregulation of the signaling network is the underlying mechanism for pathogenesis during periodontitis (Hajishengallis 2014).

Toll-like receptor (TLR) signaling is the main driving force among proinflammatory pathways (Brown et al. 2011; Hernandez et al. 2011). TLRs, including TLR2 and TLR4, are activated by oral pathogens. They recruit the myeloid differentiation primary response gene 88 (MyD88) adaptor to activate the IL-1 receptor-associated kinase (IRAK), TNF receptor–associated factor 6 (TRAF6), and transforming growth factor β–activated kinase 1 (TAK1) complexes, which then activate nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) to induce the production of proinflammatory cytokines/chemokines and tissue-damaging metalloproteases (Hajishengallis 2015). The TLR4 ligand lipopolysaccharide (LPS) from oral pathogens such as Aggregatibacter actinomycetemcomitans and Porphyromonas gingivalis stimulates alveolar bone resorption in vivo, a hallmark of periodontitis (Nishida et al. 2001). Myd88-deficient mice are resistant to A. actinomycetemcomitans LPS-induced bone loss, suggesting that attenuating the TLR pathway through ablating Myd88 can be beneficial for tissue integrity during chronic periodontitis (Madeira et al. 2013).

To maintain tissue homeostasis, the proinflammatory responses are reined in by anti-inflammatory cytokines, including the transforming growth factor β (TGFβ) family of cytokines, which are produced in the periodontal tissue (Skaleric et al. 1997; Glowacki et al. 2013; Mize et al. 2015). The balance between these pro- and anti-inflammatory signals defines the susceptibility to destructive/progressive periodontitis (Hajishengallis 2015). TGFβ family cytokines bind to type I and type II transmembrane receptor kinases and transmit signals through receptor-regulated Smads, including Smad2 and Smad3, and the co-Smad Smad4 (Feng and Derynck 2005). The TGFβ pathway is also modulated by inhibitory Smad proteins Smad6 and Smad7 (Feng and Derynck 2005). In sepsis and vascular inflammation, TGFβ-elicited inhibition of TLR signaling occurs through the induction of inhibitory Smad6 to prevent excessive inflammation and protect tissues from collateral damage (Lee et al. 2015; Li et al. 2015).

We previously reported that Smad6 methylation by protein arginine methyltransferase 1 (PRMT1) controls the TGFβ family-related bone morphogenetic protein (BMP) signaling activation (Xu et al. 2013). In this study, we demonstrated novel roles of Smad6 methylation in the crosstalk between TGFβ and TLR-NFκB signaling. We discovered that the methylated form of Smad6 sequesters MyD88 to promote MyD88 degradation and inhibit NFκB activation. This PRMT1-Smad6 signaling constrains the TLR-MyD88-NFκB pathway and limits the expression of proinflammatory cytokines to maintain homeostasis. We further showed that Smad6 is expressed and methylated in the gingival epithelium. Disturbance of Smad6 methylation exacerbates inflammation and alveolar bone loss in experimental periodontitis. Our findings highlight the pathophysiologic significance of PRMT1-induced Smad6 methylation in periodontal diseases, which constitutes a novel mechanism in the control of the host inflammatory responses. We thus propose the manipulation of PRMT1 activity as a new strategy for therapeutic intervention in periodontitis.

Materials and Methods

Detailed materials and methods are included in the Appendix.

Results

Methyl-Smad6 Interacts with MyD88 and Recruits Smurf1 to MyD88 for Degradation

The inhibitory Smad6 is a key mediator of TGFβ-elicited anti-inflammatory effects through antagonizing NFκB signaling in macrophages and aortic valvular interstitial cells. Through this mechanism, Smad6 confines the inflammatory responses and protects against adverse outcomes (Lee et al. 2015; Li et al. 2015). We reported that Smad6 methylation by the methyltransferase PRMT1 controls Smad6 activity (Xu et al. 2013). PRMT1-mediated methylation occurs on 2 arginine sites: arginine 74 (R74) and arginine 81 (R81). The 2 sites reside in the same N-terminal domain by which Smad6 interacts with MyD88 to elicit anti-inflammatory effects (Lee et al. 2011; Xu et al. 2013). Since protein methylation often controls protein-protein interactions, we determined whether Smad6 methylation modulates the interaction between Smad6 and MyD88. To examine whether Smad6 methylation is required for interaction with MyD88, we first used Smad6 wild type (WT) and methylation-deficient mutants R74A, R81A, and R74,81A, in which one or both arginine sites (R) were mutated to alanine (A; Xu et al. 2013). We cotransfected hemagglutinin (HA)–tagged MyD88 with flag-tagged Smad6 WT, R74A, R81A, or R74,81A, and assessed their interaction using coimmunoprecipitation. MyD88 interacted with Smad6 WT but not the methylation-deficient mutants of Smad6 (Fig. 1A). Next, we reduced Smad6 methylation by PRMT1 depletion. We used 293T stable cell lines published previously that express either an shRNA to deplete PRMT1 or a scrambled shRNA as control (Xu et al. 2013). Then we cotransfected HA-tagged MyD88 with flag-tagged Smad6 WT in these PRMT1-depleted or control cell lines and assessed Smad6 interaction with MyD88. Smad6 interacted with MyD88 in the control cells, but the interaction was significantly reduced in the PRMT1-depleted cells (Fig. 1B), suggesting that PRMT1-mediated methylation is required for Smad6 interaction with MyD88.

Figure 1.

Figure 1.

Open in a new tab

Methyl-Smad6 interacts with MyD88 and recruits Smurf1 to MyD88 for degradation. (A) Smad6 R74 and R81 methylation is required for Smad6 binding to MyD88. Flag-tagged Smad6 wild type (WT), R74A, R81A, or R74,81A was coexpressed with HA-tagged MyD88 in 293T cells. The interaction between Smad6 and MyD88 was assessed by coimmunoprecipitation (co-IP) with anti-HA antibody, followed by immunoblotting (IB). (B) PRMT1-mediated methylation promotes Smad6 interaction with MyD88. Flag-tagged Smad6 and HA-tagged MyD88 were cotransfected into 293T stable cell lines that express shRNA that targeted PRMT1 or control shRNA. The interaction between Smad6 and MyD88 was assessed by co-IP with anti-HA antibody, followed by IB. (C) A schematic illustration of the in vitro GST binding assays. GST-tagged Smad6 WT or R81A was generated in Escherichia coli and then either methylated by PRMT1 in vitro or left nonmethylated. HA-tagged MyD88 was expressed and purified from 293T and then conjugated to HA affinity beads. Subsequently, recombinant Smad6 and purified MyD88 were incubated in vitro, and the mixture was separated into bead-bound and flow-through fractions by centrifugation to assess their interaction. (D) Methyl-Smad6 had higher binding affinity to MyD88. GST-tagged nonmethylated Smad6 or methyl-Smad6 WT or R81A was incubated with HA-MyD88-bound beads. Bead-bound Smad6 was assessed by IB. (E, F) MyD88 associated with R81-methyl-Smad6 in HaCaT (E) and primary human gingival epithelial cells (F). The interaction between MyD88 and Smad6 was assessed by co-IP with anti-MyD88 antibody, followed by IB. Control IgG served as the control antibody for IP. (G) TGFβ-induced MyD88 degradation was concomitant with the induction of R81 methylation on Smad6 in human gingival epithelial cells. Human gingival epithelial cells were treated with TGFβ for the time indicated. The expression levels of MyD88 and Smad6 and the methylation levels of Smad6 at R74 and R81 were assessed by IB. (H) Smad6 methylation promoted Smurf1 recruitment to MyD88. Flag-tagged Smad6 WT, R74A, R81A, or R74,81A was coexpressed with HA-tagged MyD88 and Myc-tagged Smurf1 in 293T cells. Cells were treated with MG132 for 6 h to block proteasomal degradation and then harvested for co-IP with anti-Myc antibody, followed by IB. All experiments were independently repeated at least 3 times, and representative images are shown. Cont, control; HA, hemagglutinin; IB, immunoblot; IP, immunoprecipitation.

To determine whether methylated Smad6 directly binds to MyD88, we used an in vitro glutathione S-transferase (GST) binding assay, as illustrated in Figure 1C, to examine whether arginine methylation of Smad6 increased the affinity of Smad6 for MyD88. We generated recombinant GST-tagged Smad6 that is either nonmethylated or methylated in vitro by PRMT1 and purified and immobilized MyD88 on agarose beads. Then, methyl-Smad6 was compared with nonmethylated Smad6 or R81A for their binding to MyD88 (Fig. 1D). We found that methylated Smad6 bound to MyD88 efficiently, while nonmethylated Smad6 or Smad6 R81A showed much weaker binding to MyD88 (Fig. 1D, first row). Because R74 methylation did not promote in vitro binding assays (Fig. 1D, first and fourth rows), we conclude that R81 is the only methylation site responsible for Smad6-MyD88 binding. R74 mutation modulated Smad6 interaction with MyD88 in transfected cells (Fig. 1A) mainly because R74 methylation enhances the efficiency of R81 methylation, as demonstrated previously (Xu et al. 2013).

To examine whether MyD88 interacts with methyl-Smad6 in vivo, we purified the MyD88-bound protein complex from both HaCaT cells, a human keratinocyte cell line, and primary human gingival epithelial cells. We then probed for Smad6 expression and methylation using antibodies specific for total, R74-methyl-, or R81-methyl-Smad6. In both cells, we identified R81-methylated Smad6 in complex with MyD88 (Fig. 1E, second row; 1F, second row). In contrast, R74-methylated Smad6 was not in complex with MyD88 (Fig. 1E, fourth row; 1F, fourth row), further supporting that R81 is the only methylation site responsible for Smad6 binding to MyD88.

We found that TGFβ-induced degradation of MyD88 in gingival epithelial cells occurs concomitantly with the induction of R81 methylation on Smad6 (Fig. 1G, first and second rows). It has been documented that Smad6 recruits the E3 ligase Smurf1 to promote the degradation of its binding partners, including MyD88 (Murakami et al. 2003; Horiki et al. 2004; Lee et al. 2011). To examine whether Smad6 R81 methylation controls Smurf recruitment, we cotransfected cells with HA-tagged MyD88, Myc-tagged Smurf1, and flag-tagged Smad6 WT or mutants in the presence of MG132 to block degradation. MyD88 and Smurf1 had a weak interaction in the absence of Smad6 (Fig. 1H, first row, lane 6). Smad6 WT significantly strengthened the interaction (Fig. 1H, first row, lane 1), but the methylation-deficient mutants failed to do so (Fig. 1H, first row, lanes 2 to 4), indicating that Smad6 methylation is required for the recruitment of Smurf1 to MyD88.

Taken together, these data suggest that Smad6 R81 methylation is required for Smad6 interaction with MyD88 and recruitment of Smurf1.

PRMT1-Mediated Smad6 Methylation Is Required for TGFβ-Mediated Repression of NFκB Activation

The interaction between Smad6 and MyD88 controls TGFβ-mediated repression of NFκB activation (Lee et al. 2011). To examine whether Smad6 methylation is required for this function, we first assessed whether reducing Smad6 methylation inhibits TGFβ-induced repression of NFκB translocation. To reduce Smad6 methylation, we silenced Prmt1 in primary mouse gingival cells using siRNA. This approach decreased Prmt1 expression by >95% and diminished Smad6 methylation without changing the total levels of Smad6 (Fig. 2A). We stimulated primary mouse gingival cells with LPS to induce nuclear translocation of NFκB (Fig. 2B, first and second columns; Fig. 2C). TGFβ family ligands TGFβ and BMP4 antagonized LPS-induced translocation in these cells (Fig. 2B, third and fourth columns; Fig. 2C). We found that reducing Smad6 methylation through silencing Prmt1 dramatically inhibited the antagonizing effects of TGFβ and BMP4, as shown by an increase in NFκB translocation (Fig. 2B, fifth and sixth columns; Fig. 2C). A similar effect was apparent in HaCaT cells, whereby reducing Smad6 methylation inhibited the antagonizing effects of TGFβ on NFκB translocation (Fig. 2D, E). To complement this approach, we isolated cytosolic and nuclear fractions for the analysis of NFκB levels and showed that similar effects were evident (Fig. 2F, G). These results suggest that PRMT1 is required for TGFβ- and BMP-mediated repression of NFκB translocation.

Figure 2.

Figure 2.

Open in a new tab

PRMT1-mediated Smad6 methylation is required for TGFβ-mediated repression of NFκB activation. (A) Silencing Prmt1 with a siRNA diminished Prmt1 expression and Smad6 methylation. Mouse primary gingival cells transfected with control or siRNA targeting Prmt1 were assessed with anti-PRMT1, anti-methyl-Smad6R81 (S6R81me2), or anti-Smad6 antibody by immunoblotting (IB). Tubulin served as a loading control. (B) Silencing Prmt1 to reduce Smad6 methylation inhibited TGFβ- and BMP4-mediated repression of NFκB translocation. Mouse primary gingival cells transfected with control or siRNA targeting Prmt1 were pretreated with or without TGFβ or BMP4 for 2 h and then with or without LPS for 4 h. LPS-induced nuclear translocation of NFκB was assessed by immunostaining with an anti-NFκB antibody. DAPI labeled the nuclei. Scale bar = 50 μm. (C) Quantification of nuclear translocation of NFκB in panel B, illustrated as the ratio between nuclear and cytosolic staining of NFκB. In each group, 200 to 300 cells were analyzed with ImageJ. (D) Silencing PRMT1 to reduce Smad6 methylation inhibited TGFβ-mediated repression of NFκB translocation. HaCaT cells transfected with control or siRNA targeting Prmt1 were pretreated with or without TGFβ for 2 h and then with or without LPS for 12 h. LPS-induced nuclear translocation of NFκB was assessed by immunostaining with an anti-NFκB antibody. DAPI labeled the nuclei. Scale bar = 50 μm. (E) Quantification of nuclear translocation of NFκB in panel D, illustrated as the ratio between nuclear and cytosolic staining of NFκB. In each group, 200 to 300 cells were analyzed with ImageJ. (F, G) Cytosolic and nuclear proteins were extracted from human gingival epithelial cells (F) or HaCaT cells (G) for the assessment of NFκB translocation with anti-NFκB antibody by IB. GAPDH served as a marker for the cytosolic fraction, and HDAC1 served as a marker for the nuclear fraction. (H, I) Silencing Prmt1 enhanced Il-8 and Tnf-α expression and blocked TGFβ- and BMP4-mediated repression of Il-8 and Tnf-α expression. Mouse primary gingival cells were transfected with siRNA targeting Prmt1 or control, treated with or without TGFβ or BMP4 for 2 h, and then with or without LPS for 4 h. LPS-induced Il-8 and Tnf-α mRNA expression was assessed by reverse transcription quantitative real-time polymerase chain reaction. #P < 0.05 vs. no treatment group. *P < 0.05 vs. control siRNA-transfected group with the same treatment. Experiments were independently repeated 3 times, and representative data are shown. (J–M) Inhibition of PRMT1 enzymatic activity with MS023 blocked TGFβ- and BMP4-mediated repression of Il-8 and Tnf-α expression. HaCaT cells (J–K) or human gingival epithelial cells (L–M) were treated with PRMT1 inhibitor MS023 or control, treated with or without TGFβ or BMP4 for 2 hours, and then with or without LPS for 4 h. LPS-induced IL-8 and TNF-α mRNA expression was assessed by reverse transcription quantitative real-time polymerase chain reaction. #P < 0.05 vs. no treatment group. *P < 0.05 vs. control siRNA-transfected group with the same treatment. Experiments were independently repeated 3 times, and representative data are shown. All data are presented as mean ± SEM; comparisons between groups were performed with 2 -tailed t tests. IB, immunoblot; LPS, lipopolysaccharide. All experiments were independently repeated 3 times, and representative data are shown.

Nuclear translocation of NFκB results in transcriptional activation of NFκB target genes, many of which encode proinflammatory cytokines and chemokines (Karin 1999). To examine whether PRMT1 and Smad6 methylation are required for TGFβ-mediated repression of NFκB target genes, we depleted PRMT1 in primary mouse gingival cells and assayed for the expression of inflammatory cytokines Il-8 and Tnf-α following LPS treatment. TGFβ and BMP repressed LPS-induced expression of Il-8 and Tnf-α, but silencing Prmt1 abrogated their repression (Fig. 2H, I). These observations suggest that PRMT1 is required for TGFβ- and BMP-mediated repression of NFκB-responsive inflammatory cytokine expression. To further demonstrate that the enzymatic activity of PRMT1 is crucial for TGFβ-mediated repression of NFκB target genes, we inhibited PRMT1 using a small molecular inhibitor, MS023 (Eram et al. 2016). In both primary human gingival epithelial cells and HaCaT cells, PRMT1 inhibition abrogated TGFβ- or BMP-mediated repression of IL-8 and TNF-α expression (Fig. 2J–M). These observations suggest that PRMT1 is required for TGFβ- and BMP-mediated repression of NFκB-responsive inflammatory chemokine expression.

Taken together, these data demonstrate that PRMT1-mediated Smad6 methylation is required for TGFβ- and BMP-mediated repression of NFκB translocation and NFκB-responsive inflammatory chemokine expression and that this role of PRMT1 is conserved between gingival and skin epithelial cells.

Smad6 Is Expressed in the Healthy and Inflamed Gingival Epithelium

Having dissected the mechanism whereby Smad6 regulates inflammatory responses in gingival epithelial cells in vitro, we set out to determine the in vivo functions of this activity, as the role of Smad6 has not been addressed in the context of periodontitis. First, we investigated whether Smad6 is expressed in the murine gingival epithelium using immunohistochemical analysis. We revealed that Smad6 was highly expressed in healthy gingival epithelium, as indicated by colocalization with the epithelial marker E-cadherin (Fig. 3A; above the dashed line). In contrast, Smad6 expression in the underlying connective tissue was very low (Fig. 3A; below the dashed line). The specificity of the staining was confirmed with isotype controls performed simultaneously (Appendix Fig. 1A). To evaluate the expression of Smad6 during inflammation, we stained for Smad6 in inflamed periodontal tissues and showed that Smad6 expression was moderately reduced (Fig. 3B; above the dashed line), whereas in the connective tissue, Smad6 expression was dramatically elevated as compared with its expression in healthy tissue (Fig. 3A; below the dashed line). Therefore, although Smad6 expression is essentially restricted to the gingival epithelium under homeostatic conditions, Smad6 is additionally expressed in the gingival connective tissue under inflammatory conditions.

Figure 3.

Figure 3.

Open in a new tab

Smad6 and NFκB are expressed in healthy and inflamed gingival epithelium. (A) Smad6 was highly expressed in gingival epithelium but not in gingival connective tissues beneath the epithelium. Smad6 expression was detected by immunostaining with an anti-Smad6 antibody in coronal sections of mouse healthy periodontal tissue from molar 1 to 3 (M1 to M3). E-cadherin labeled the gingival epithelium (above the dashed line), and DAPI labeled the nuclei. Scale bar = 50 μm. (B) Smad6 expression in gingival connective tissue was elevated during inflammation. Smad6 expression was detected as in panel A. E-cadherin labeled the gingival epithelium (above the dash line), and DAPI labeled the nuclei. Scale bar = 50 μm. (C, D) NFκB was expressed in the cytoplasm of healthy gingival epithelial cells. NFκB expression was detected by staining with an anti-NFκB antibody in coronal sections of mouse periodontal tissue (M1 to M3). E-cadherin labeled the gingival epithelium (above the dash line), and DAPI labeled the nuclei. Scale bar = 50 μm in panel C. Images inside the frames of panel C are enlarged in panel D to highlight nuclear and cytosolic region. (E, F) NFκB was expressed in the cytosol and nucleus of gingival epithelial cells during inflammation. NFκB expression was detected as in panel C. E-cadherin labeled the gingival epithelium (above the dashed line), and DAPI labeled the nuclei. Scale bar = 50 μm in panel E. Images inside the frames of panel E are enlarged in panel F to highlight nuclear and cytosolic region.

NFκB Is Expressed in Gingival Epithelium and the Expression Elevates during Inflammation

TLR-MyD88-NFκB signaling drives the proinflammatory process in various tissues, including the gingiva (Brown et al. 2011; Hernandez et al. 2011; Hajishengallis 2015). We identified that NFκB p65 subunit was highly expressed in the gingival epithelium (Fig. 3Ca′–d′, above the dashed line; Appendix Fig. 1Ba′–d′) and localized in the cytosol, as illustrated in the enlarged images where NFκB expression and DAPI-stained nuclei display exclusive patterns (Fig. 3Da′–d′). NFκB is normally sequestered by IκB in the cytosol in an inactive complex. When TLRs are activated, TLR-induced signaling cascades lead to IκB degradation and subsequent nuclear translocation of NFκB, which then activates the transcription of proinflammatory cytokines and chemokines (Karin 1999). The cytosolic localization of NFκB in the gingival epithelium suggests that NFκB is maintained in an inactive state and poised to initiate an innate immune response upon stimulation. In inflamed periodontal tissues, NFκB expression was enhanced and the expression pattern changed. NFκB was expressed in the cytosol and nucleus of the inflamed gingival epithelium (Fig. 3E, above the dashed line; Fig. 3F), consistent with NFκB activation during inflammation (Karin 1999).

Figure 5.

Figure 5.

Open in a new tab

Smad6 methylation is required to inhibit nuclear translocation of NFκB and its target gene expression. (A) Smad6 expression is absent in SMAD6 knockout (KO) HaCaT cells. Control and SMAD6 KO HaCaT cells were assessed by immunoblotting (IB) for the expression levels of Smad6 protein. (B–D) Enhancing Smad6 methylation in SMAD6 KO HaCaT cells inhibited lipopolysaccharide (LPS)–induced NFκB translocation by immunostaining. SMAD6 KO HaCaT cells were transfected with the control vector, Smad6 wild type (WT), or R81A and then treated with LPS. Smad6 WT and R81A were expressed at comparable levels as indicated by IB (B). LPS-induced nuclear translocation of NFκB was assessed by immunostaining with an anti-NFκB antibody (C) and quantified (D). Quantification of nuclear translocation of NFκB was illustrated as the ratio between cytosolic and nuclear staining of NFκB. In each group, 200 to 300 cells were analyzed with ImageJ. NFκB localization is further illustrated in the enlarged images (a′′–i′′), where NFκB expression and DAPI-stained nuclei partially overlapped in the control and Smad6 R81A expression groups (a′′–c′′, g′′–i′′) but displayed exclusive patterns in the Smad6 WT expression group (d′′–f′′). All data are presented as mean ± SEM; comparisons between groups were performed with 2-tailed t tests. DAPI labeled the nuclei. (E) Enhancing Smad6 methylation in SMAD6 KO HaCaT cells inhibited LPS-induced NFκB translocation by IB via cytosolic and nuclear protein fractions. GAPDH served as a marker for the cytosolic fraction, and HDAC1 served as a marker for the nuclear fraction. IB, immunoblot. Scale bar = 50 μm. Experiments were independently repeated 3 times, and representative data are shown. (F, G) Methyl-Smad6 inhibited LPS-induced IL-8 and TNF-α expression. SMAD6 KO HaCaT cells were transfected with control vector, Smad6 WT, or R81A and treated with LPS. LPS-induced IL-8 and TNF-α mRNA expression was assessed by reverse transcription quantitative real-time polymerase chain reaction. #P < 0.05 LPS vs. no treatment in the same group. *P < 0.05 vs. vector control group with the same treatment. All data are presented as mean ± SEM; comparisons between groups were performed with 2-tailed t tests. All experiments were independently repeated at least 3 times, and representative data are shown.

Taken together, these data suggest that NFκB is enriched in the healthy gingival epithelium (mostly localized in the cytosol in an inactive state) and becomes activated in the inflamed gingiva.

Depletion of PRMT1 to Reduce Smad6 Methylation Enhances Experimental Periodontitis-Induced Inflammation and Tissue Damage

In inflammatory diseases such as periodontitis, NFκB-induced expression of chemokines results in the recruitment of inflammatory cells and fuels disease progression that leads to tissue damage. We utilized a murine experimental model of ligature-induced periodontitis to study inflammatory responses (Jiao et al. 2013; Appendix Fig. 2).

To examine whether Prmt1 deletion in periodontal epithelium enhances inflammation, we treated Prmt1flox/flox;ROSA-STOP-YFP mice with recombinant Tat-Cre fusion protein to induce local deletion of Prmt1. The ROSA-STOP-YFP transgene allows us to track Cre-mediated recombination with YFP expression. Tat is a short peptide derived from human HIV that facilitates protein transport across the cell membrane. The fusion of Tat peptide with Cre recombinase allows Cre to penetrate into gingival epithelial cells and induce recombination to delete the floxed gene (Becker-Hapak and Dowdy 2003; Han et al. 2013). We inserted the ligature to induce periodontitis, applied Tat-Cre to induce deletion, and then harvested gingival tissue as illustrated (Fig. 4A). Topical application of Tat-Cre protein induced significant reduction of gingival Prmt1 expression (Fig. 4B) and YFP expression in >90% of gingival epithelial cells (Fig. 4Ca′–d′), whereas gingival tissue sections in the phosphate buffered saline–treated group or isotype control group displayed no YFP-positive cells (Fig. 4Ce′–h′, Appendix Fig. 1C).

Figure 4.

Figure 4.

Open in a new tab

Depletion of PRMT1 to reduce Smad6 methylation enhances experimental periodontitis-induced inflammation and tissue damage. (A) The treatment regimen for Tat-Cre application and ligature-induced periodontitis. At day 0, Prmt1flox/flox mice were subjected to ligature insertion on the left side. The contralateral side without ligature insertion served as a control. Then the Tat-Cre or phosphate-buffered saline (PBS) control in glycerol solution was applied to the surface of periodontal regions on the ligature-inserted side and the contralateral control side daily for 3 d from day 0 to day 2. Mice were harvested at day 5 for analysis. (B) Five days after ligature placement, Prmt1flox/flox mice treated with PBS or Tat-Cre were euthanized, and periodontal soft tissues were surgically removed for RNA extraction. Prmt1 mRNA expression was assessed by reverse transcription quantitative real-time polymerase chain reaction. n = 4 per group. All data are presented as mean ± SEM; comparisons between groups were performed with 2-tailed t tests. (C) Tat-Cre-induced deletion caused YFP expression in gingival epithelium of Tat-Cre-treated Prmt1flox/flox; ROSA-STOP-YFP mice but not the PBS-treated group. Maxillae from Prmt1flox/flox; ROSA-STOP-YFP mice treated with PBS or Tat-Cre following ligature treatment were fixed for decalcification and histologic analysis. The expression of YFP was assessed by immunostaining via an anti-YFP antibody. Scale bar = 50 μm. n = 4 per group. (D) Five days after ligature placement, Prmt1flox/flox mice treated with PBS or Tat-Cre were euthanized, and periodontal soft tissues were surgically removed for RNA extraction. Tnf-α mRNA expression was assessed by reverse transcription quantitative real-time polymerase chain reaction. n = 4 per group. All data are presented as mean ± SEM; comparisons between groups were performed with 2-tailed t tests. (E) Maxillae from Prmt1flox/flox mice treated with PBS or Tat-Cre following ligature treatment were fixed for decalcification and histologic analysis. The expression of TNF-α was assessed by DAB staining via an anti-TNF-α antibody. Scale bar = 50 μm. n = 3 per group. (F, G) Prmt1flox/flox mice treated with PBS or Tat-Cre received ligature treatment on the left side (ligature side: F, right panels) and no ligature treatment on the contralateral side as controls (control side: F, left panels). The infiltration of neutrophils, indicated by red arrowheads, was assessed by specific chloroacetate esterase staining and quantified by positive stained cell number per field. Scale bar = 50 μm. n = 3 per group. All data are presented as mean ± SEM; comparisons between groups were performed with 2-tailed t tests. (H–J) Prmt1flox/flox mice treated with PBS (H, upper panels) or Tat-Cre (H, lower panels) received ligature treatment on the left side (ligature side: H, right panels) and no ligature treatment on the contralateral side as controls (control side: H, left panels). n = 7 per group. Three-dimensional images of representative molar areas (H), cementoenamel junction and alveolar bone crest length (CEJ-ABC; mm) (I), and alveolar bone volume fraction (J) were determined 5 d after ligature placement. Scale bar = 500 μm.

Tat-Cre-mediated Prmt1 deletion in Prmt1flox/flox mice significantly increased the expression of proinflammatory cytokine Tnf-α (Fig. 4D, E) and enhanced neutrophil infiltration (Fig. 4F, G). Prmt1 deficiency also augmented tissue damage to the hard tissue, the alveolar bone, as demonstrated by a significant increase in the distance between the cementoenamel junction and alveolar bone crest (Fig. 4I) and a significant decrease in alveolar bone volume (Fig. 4J). These data show that Prmt1 deficiency exacerbates periodontal inflammation and alveolar bone loss.

Smad6 Methylation Is Required to Inhibit Nuclear Translocation of NFκB and Its Target Gene Expression

To obtain an in-depth molecular understanding of the function of Smad6 methylation, we generated SMAD6 knockout HaCaT cells using vectors that express the cas9 nuclease and guide RNAs that targeted the SMAD6 gene (Appendix Fig. 3) and confirmed the loss of Smad6 expression in SMAD6 knockout cells (Fig. 5A). We then expressed Smad6 WT or methylation-deficient mutant R81A at a comparable level in SMAD6 knockout cells (Fig. 5B) to study the function of Smad6 methylation in NFκB translocation. First, NFκB localization was revealed by immuno-staining. Smad6 WT inhibited NFκB translocation in response to LPS treatment, but Smad6 R81A failed to do so (Fig. 5C, D). These findings are further demonstrated by subcellular fractionation that Smad6 WT increased the level of cytosolic NFκB and concomitantly decreased the level of nuclear NFκB, but Smad6 R81A failed to do so (Fig. 5E). These results suggest that methylation of Smad6 is required to repress LPS-induced NFκB translocation.

To examine whether methyl-Smad6 represses NFκB-induced proinflammatory chemokine expression, we assayed for mRNA expression of proinflammatory chemokines IL-8 and TNF-α following LPS treatment. Smad6 WT inhibited IL-8 and TNF-α expression, but Smad6 R81A failed to do so (Fig. 5F, G). Taken together, these data demonstrate that R81 methylation is indispensable for Smad6’s inhibitory effects on NFκB translocation and NFκB-induced chemokine expression.

Discussion

In this study, we identified arginine methylation of Smad6 as a novel mechanism in the regulation of TLR signaling, a discovery with important therapeutic implications in periodontitis. We demonstrated that Smad6 methylation is a prerequisite step for Smad6 binding to MyD88 and recruitment of the E3 ligase Smurf1, thereby promoting MyD88 degradation and inhibiting MyD88-mediated NFκB activation. Methyl-Smad6 inhibited the TLR-MyD88-NFκB signaling and limited the expression of proinflammatory cytokines. We further identified Smad6 expression and methylation in the gingival epithelium. Methyl-Smad6 restrained the inflammatory responses in periodontal tissue and mitigated tissue damage during periodontitis. Disturbance of Smad6 methylation in vivo exacerbated inflammation and alveolar bone loss in periodontitis. Our findings indicate that activation of Smad6 methylation through manipulation of PRMT1 activity constitutes a potential novel strategy to modulate host immune responses in periodontitis.

Protein arginine methylation is a type of posttranslational modification catalyzed by the protein arginine methyltransferase family of proteins (PRMTs). PRMT1 is the major PRMT in mammals and plays pleiotropic roles (Bedford and Clarke 2009). It catalyzes the methylation of histone on H4R3 to modify epigenetic status and the methylation of nonhistone proteins, including receptors, signal mediators, and transcription factors, to control signaling and transcription. The function of PRMT1 has been associated with inflammation via epithelial and immune cell lines, but the findings are incongruous. PRMT1 serves as an epigenetic cofactor and induces the expression of Cox2, CITED2, and NFκB (Hassa et al. 2008; Kleinschmidt et al. 2008), which would predict that PRMT1 promotes inflammation. However, PRMT1 inhibition was also found to increase NFκB-dependent gene expression in immune cells (Tikhanovich et al. 2015), which would indicate anti-inflammatory functions for PRMT1. Now we provide the first in vivo evidence that Prmt1 deficiency significantly promotes inflammation and tissue damage. We integrated independent approaches, including the Prmt1flox/flox mouse genetic model, epithelial cell culture, and biochemical techniques, to define PRMT1-mediated Smad6 methylation and revealed a hitherto unidentified mechanism in the modulation of NFκB activity. We showed that PRMT1 is required for TGFβ-elicited inhibition of TLR signaling. Additionally, PRMT1 and Smad6 methylation control NFκB activity at the baseline, as depletion of either PRMT1 or Smad6 methylation enhanced NFκB translocation and NFκB-induced chemokine expression in the absence of LPS stimulation. These findings suggest that methyl-Smad6 limits NFκB activity in the absence of LPS stimulation.

Smad6 transduces TGFβ-elicited inhibition of TLR signaling through 3 mechanisms: 1) recruiting MyD88 via its N-terminal region to target MyD88 for degradation (Lee et al. 2011), 2) recruiting Pellino1 via its C-terminal region to inhibit the activity of the IRAK-TRAF6 complex (Choi et al. 2006), and 3) recruiting A20 and TRAF6 via its C-terminal region to inhibit TRAF6 activation (Jung et al. 2013). This study focused on the function of Smad6 on MyD88 because Smad6 arginine methylation occurs at the N-terminus, which is also the MyD88-binding region. It is possible that Smad6 R81 methylation could further modulate the C-terminal function of Smad6 to control its association with Pellino1, A20, or TRAF6. These possibilities need to be evaluated in future studies.

We identified gingival epithelium as an arena where TGFβ-induced methyl-Smad6 intercepts TLR-MyD88-activated NFκB. The high levels of immune modulators, including NFκB and Smad6, in gingival epithelium suggest their critical roles in maintaining immune homeostasis and periodontal health. The cytosolic localization of NFκB in the gingival epithelium suggests that NFκB is maintained in an inactive state but capable of promptly initiating inflammatory responses upon stimulation. As expected, a fraction of NFκB localizes to the nucleus in inflamed gingival epithelium to activate chemokine expression. Our findings are consistent with the notion that barrier epithelium produces immune modulatory chemokines and defines inflammatory responses and disease outcomes (Peterson and Artis 2014).

In summary, we investigated molecular functions of Smad6 in gingival epithelial cells. We identified a novel regulatory mechanism involving Smad6 methylation in the regulation of TLR-NFκB signaling. This mechanism defines a new point of intervention with therapeutic potential in periodontitis and perhaps other inflammatory diseases.

Author Contributions

T. Zhang, contributed to conception, design, data acquisition, analysis, and interpretation, drafted and critically revised the manuscript; J. Wu, Y. Shen, J. Liu, S. Richard, J. Jin, contributed to data acquisition, critically revised the manuscript; N. Ungvijanpunya, contributed to data acquisition, analysis, and interpretation, critically revised the manuscript; O. Jackson-Weaver, Y. Gou, T.V. Ho, contributed to data acquisition and analysis, critically revised the manuscript; J. Feng, Y. Chai, contributed to design and data analysis, critically revised the manuscript; G. Hajishengallis, contributed to data analysis and interpretation, critically revised the manuscript; J. Xu, contributed to conception, design, data acquisition, analysis, and interpretation, critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work.

Supplemental Material

DS_10.1177_0022034518755688 – Supplemental material for Smad6 Methylation Represses NFκB Activation and Periodontal Inflammation
Click here for additional data file. (797.3KB, pdf)

Supplemental material, DS_10.1177_0022034518755688 for Smad6 Methylation Represses NFκB Activation and Periodontal Inflammation by T. Zhang, J. Wu, N. Ungvijanpunya, O. Jackson-Weaver, Y. Gou, J. Feng, T.V. Ho, Y. Shen, J. Liu, S. Richard, J. Jin, G. Hajishengallis, Y. Chai, and J. Xu in Journal of Dental Research

Acknowledgments

We thank Drs. Pinghui Feng, Yang Chai, Michael Stallcup, and Malcolm Snead for their critical reading of the manuscript and thoughtful input. We also thank Dr. Bridget Samuels for review and editing of the manuscript.

Footnotes

A supplemental appendix to this article is available online.

This research was supported by a Research Seed Grant from the University of Southern California School of Dentistry (to J.X.), a start-up fund from the provost at the University of Southern California (to J.X.), and R01DE026468 (to J.X.) from the U.S. National Institutes of Health. This research was also supported by grants R01CA218600 (to J.J.), R01GM- 122749 (to J.J.), and R01HD088626 (to J.J.) and R37DE026152 (to G.H.) from the National Institutes of Health.

The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.

References

  1. Becker-Hapak M, Dowdy SF. 2003. Protein transduction: generation of full-length transducible proteins using the tat system. Curr Protoc Cell Biol. 20:202. [DOI] [PubMed] [Google Scholar]
  2. Bedford MT, Clarke SG. 2009. Protein arginine methylation in mammals: who, what, and why. Mol Cell. 33(1):1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Brown J, Wang H, Hajishengallis GN, Martin M. 2011. TLR-signaling networks: an integration of adaptor molecules, kinases, and cross-talk. J Dent Res. 90(4):417–427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Choi KC, Lee YS, Lim S, Choi HK, Lee CH, Lee EK, Hong S, Kim IH, Kim SJ, Park SH. 2006. Smad6 negatively regulates interleukin 1-receptor-Toll-like receptor signaling through direct interaction with the adaptor Pellino-1. Nat Immunol. 7(10):1057–1065. [DOI] [PubMed] [Google Scholar]
  5. Darveau RP. 2010. Periodontitis: a polymicrobial disruption of host homeostasis. Nat Rev Microbiol. 8(7):481–490. [DOI] [PubMed] [Google Scholar]
  6. Eram MS, Shen Y, Szewczyk M, Wu H, Senisterra G, Li F, Butler KV, Kaniskan HU, Speed BA, Dela Sena C, et al. 2016. A potent, selective, and cell-active inhibitor of human type I protein arginine methyltransferases. ACS Chem Biol. 11(3):772–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Feng XH, Derynck R. 2005. Specificity and versatility in TGF-beta signaling through Smads. Annu Rev Cell Dev Biol. 21:659–693. [DOI] [PubMed] [Google Scholar]
  8. Glowacki AJ, Yoshizawa S, Jhunjhunwala S, Vieira AE, Garlet GP, Sfeir C, Little SR. 2013. Prevention of inflammation-mediated bone loss in murine and canine periodontal disease via recruitment of regulatory lymphocytes. Proc Natl Acad Sci U S A. 110(46):18525–18530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Hajishengallis G. 2014. Immunomicrobial pathogenesis of periodontitis: keystones, pathobionts, and host response. Trends Immunol. 35(1):3–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hajishengallis G. 2015. Periodontitis: from microbial immune subversion to systemic inflammation. Nat Rev Immunol. 15(1):30–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Han G, Bian L, Li F, Cotrim A, Wang D, Lu J, Deng Y, Bird G, Sowers A, Mitchell JB, et al. 2013. Preventive and therapeutic effects of Smad7 on radiation-induced oral mucositis. Nat Med. 19(4):421–428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hassa PO, Covic M, Bedford MT, Hottiger MO. 2008. Protein arginine methyltransferase 1 coactivates NF-κB-dependent gene expression synergistically with CARM1 and PARP1. J Mol Biol. 377(3):668–678. [DOI] [PubMed] [Google Scholar]
  13. Hernandez M, Dutzan N, García-Sesnich J, Abusleme L, Dezerega A, Silva N, González FE, Vernal R, Sorsa T, Gamonal J. 2011. Host-pathogen interactions in progressive chronic periodontitis. J Dent Res. 90(10):1164–1170. [DOI] [PubMed] [Google Scholar]
  14. Horiki M, Imamura T, Okamoto M, Hayashi M, Murai J, Myoui A, Ochi T, Miyazono K, Yoshikawa H, Tsumaki N. 2004. Smad6/Smurf1 overexpression in cartilage delays chondrocyte hypertrophy and causes dwarfism with osteopenia. J Cell Biol. 165(3):433–445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jiao Y, Darzi Y, Tawaratsumida K, Marchesan JT, Hasegawa M, Moon H, Chen GY, Nunez G, Giannobile WV, Raes J, et al. 2013. Induction of bone loss by pathobiont-mediated nod1 signaling in the oral cavity. Cell Host Microbe. 13(5):595–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jung SM, Lee JH, Park J, Oh YS, Lee SK, Park JS, Lee YS, Kim JH, Lee JY, Bae YS, et al. 2013. Smad6 inhibits non-canonical TGF-β1 signalling by recruiting the deubiquitinase A20 to TRAF6. Nat Commun. 4:2562. [DOI] [PubMed] [Google Scholar]
  17. Karin M. 1999. How NF-kappaB is activated: the role of the IkappaB kinase (IKK) complex. Oncogene. 18(49):6867–6874. [DOI] [PubMed] [Google Scholar]
  18. Kleinschmidt MA, Streubel G, Samans B, Krause M, Bauer UM. 2008. The protein arginine methyltransferases CARM1 and PRMT1 cooperate in gene regulation. Nucleic Acids Res. 36(10):3202–3213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lee YS, Park JS, Jung SM, Kim SD, Kim JH, Lee JY, Jung KC, Mamura M, Lee S, Kim SJ, et al. 2015. Inhibition of lethal inflammatory responses through the targeting of membrane-associated Toll-like receptor 4 signaling complexes with a Smad6-derived peptide. EMBO Mol Med. 7(5):577–592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lee YS, Park JS, Kim JH, Jung SM, Lee JY, Kim SJ, Park SH. 2011. Smad6-specific recruitment of Smurf E3 ligases mediates TGF-β1-induced degradation of MyD88 in TLR4 signalling. Nat Commun. 2:460. [DOI] [PubMed] [Google Scholar]
  21. Li X, Lim J, Lu J, Pedego TM, Demer L, Tintut Y. 2015. Protective role of Smad6 in inflammation-induced valvular cell calcification. J Cell Biochem. 116(10):2354–2364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Madeira MF, Queiroz-Junior CM, Cisalpino D, Werneck SM, Kikuchi H, Fujise O, Ryffel B, Silva TA, Teixeira MM, Souza DG. 2013. MyD88 is essential for alveolar bone loss induced by Aggregatibacter actinomycetemcomitans lipopolysaccharide in mice. Mol Oral Microbiol. 28(6):415–424. [DOI] [PubMed] [Google Scholar]
  23. Mize TW, Sundararaj KP, Leite RS, Huang Y. 2015. Increased and correlated expression of connective tissue growth factor and transforming growth factor beta 1 in surgically removed periodontal tissues with chronic periodontitis. J Periodontal Res. 50(3):315–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Murakami G, Watabe T, Takaoka K, Miyazono K, Imamura T. 2003. Cooperative inhibition of bone morphogenetic protein signaling by Smurf1 and inhibitory Smads. Mol Biol Cell. 14(7):2809–2817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nishida E, Hara Y, Kaneko T, Ikeda Y, Ukai T, Kato I. 2001. Bone resorption and local interleukin-1alpha and interleukin-1beta synthesis induced by Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis lipopolysaccharide. J Periodontal Res. 36(1):1–8. [DOI] [PubMed] [Google Scholar]
  26. Peterson LW, Artis D. 2014. Intestinal epithelial cells: regulators of barrier function and immune homeostasis. Nat Rev Immunol. 14(3): 141–153. [DOI] [PubMed] [Google Scholar]
  27. Skaleric U, Kramar B, Petelin M, Pavlica Z, Wahl SM. 1997. Changes in TGF-beta 1 levels in gingiva, crevicular fluid and serum associated with periodontal inflammation in humans and dogs. Eur J Oral Sci. 105(2):136–142. [DOI] [PubMed] [Google Scholar]
  28. Tikhanovich I, Kuravi S, Artigues A, Villar MT, Dorko K, Nawabi A, Roberts B, Weinman SA. 2015. Dynamic arginine methylation of tumor necrosis factor (TNF) receptor-associated factor 6 regulates Toll-like receptor signaling. J Biol Chem. 290(36):22236–22249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Xu J, Wang AH, Oses-Prieto J, Makhijani K, Katsuno Y, Pei M, Yan L, Zheng YG, Burlingame A, Bruckner K, et al. 2013. Arginine methylation initiates BMP-induced Smad signaling. Mol Cell. 51(1):5–19. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

DS_10.1177_0022034518755688 – Supplemental material for Smad6 Methylation Represses NFκB Activation and Periodontal Inflammation
Click here for additional data file. (797.3KB, pdf)

Supplemental material, DS_10.1177_0022034518755688 for Smad6 Methylation Represses NFκB Activation and Periodontal Inflammation by T. Zhang, J. Wu, N. Ungvijanpunya, O. Jackson-Weaver, Y. Gou, J. Feng, T.V. Ho, Y. Shen, J. Liu, S. Richard, J. Jin, G. Hajishengallis, Y. Chai, and J. Xu in Journal of Dental Research

Articles from Journal of Dental Research are provided here courtesy of International and American Associations for Dental Research

RESOURCES