Palmitate-Triggered COX2/PGE2-Related Hyperinflammation in Dual-Stressed PdL Fibroblasts Is Mediated by Repressive H3K27 Trimethylation

Abstract

The interrelationships between periodontal disease, obesity-related hyperlipidemia and mechanical forces and their modulating effects on the epigenetic profile of periodontal ligament (PdL) cells are assumed to be remarkably complex. The PdL serves as a connective tissue between teeth and alveolar bone and is involved in pathogen defense and the inflammatory responses to mechanical stimuli occurring during tooth movement. Altered inflammatory signaling could promote root resorption and tooth loss. Hyperinflammatory COX2/PGE2 signaling was reported for human PdL fibroblasts (HPdLFs) concomitantly stressed with Porphyromonas gingivalis lipopolysaccharides and compressive force after exposure to palmitic acid (PA). The aim of this study was to investigate the extent to which this was modulated by global and gene-specific changes in histone modifications. The expression of key epigenetic players and global H3Kac and H3K27me3 levels were quantitatively evaluated in dual-stressed HPdLFs exposed to PA, revealing a minor force-related reduction in repressive H3K27me3. UNC1999-induced H3K27me3 inhibition reversed the hyperinflammatory responses of dual-stressed PA cultures characterized by increased COX2 expression, PGE2 secretion and THP1 adhesion. The reduced expression of the gene encoding the anti-inflammatory cytokine IL-10 and the increased presence of H3K27me3 at its promoter-associated sites were reversed by inhibitor treatment. Thus, the data highlight an important epigenetic interplay between the different stimuli to which the PdL is exposed.

Keywords: COX2/PGE2; IL-10; histone modification; inflammation; obesity; palmitic acid; periodontitis; tooth movement.

Figure 1

Palmitic acid did not impact H3Kac in compressed HPdLFs stimulated with P. gingivalis LPS. (a,b) Quantitative expression analysis of genes encoding histone acetyl transferases (CREBBP, EP300, KAT8, NCOA3) in (a) and genes encoding proteins relevant for histone deacetylation (HDAC1, HDAC2, HDAC3, SIN3A) in (b) in human periodontal ligament fibroblasts (HPdLFs) exposed to palmitic acid (PA) and simultaneously stimulated with compressive force (CF) and P. gingivalis LPS and compared to BSA+LPS controls. (c,d) Representative micrographs of global H3K9/14/18/23/27 acetylation (H3Kac) of HPdLFs under previous conditions. H3Kac staining intensity is shown in thermal LUT (th. LUT) analyzed in relation to BSA+LPS control (c). Cell nuclei were visualized with DAPI. *** p < 0.001 in relation to BSA+LPS; ### p < 0.001 in relation to PA+LPS; n.s., no significant difference between the conditions. One-way ANOVA and post hoc test (Tukey) were used for analysis. Scale bars: 10 μm in (c).

Figure 2

K27 trimethylation in response to dual stimulation was minor in PA-exposed HPdLFs. (a) Quantitative expression analysis of genes encoding components of the polycomb repressive complex (EZH1, EZH2, SUZ12 and EED) in human periodontal ligament fibroblasts (HPdLFs) exposed to palmitic acid (PA) and simultaneously stimulated with compressive force (CF) and P. gingivalis LPS and compared to the BSA+LPS controls. (b,c) Representative microphotographs of global H3K27me3 of HPdLFs under previous conditions. H3K27me3 staining intensity is shown as thermal LUT (th. LUT) in (c) in relation to BSA+LPS control. DAPI indicates the nuclei. *** p < 0.001 in relation to BSA+LPS; ### p < 0.001 in relation to PA+LPS; §§§ p < 0.001 in relation to BSA+LPS+CF; n.s., no significant difference between the conditions. One-way ANOVA and post hoc test (Tukey) were used for analysis. Scale bars: 10 μm in (b).

Figure 3

Inhibition of H3K27 trimethylation abrogates the PA-induced excessive inflammatory response of dual-stimulated HPdLFs. (a) Analysis of the activity of histone methyltransferases (HMTs) specifically methylating H3K27 in HPdLFs treated with palmitic acid (PA) for six days prior to treatment with P. gingivalis LPS and compressive force (CF) for six hours. (b) Metabolic activity analyzed in HPdLFs treated with different concentrations of UNC1999 for six hours normalized to DMSO control. (c) Analysis of H3K27-specific HMT activity in HPdLFs treated with 1.00 µM UNC1999 in relation to DMSO control. (d) Metabolic activity in dual-stressed HPdLFs treated with 1.00 µM UNC1999 in relation to DMSO-treated BSA+LPS+CF control. (e) Representative images of TUNEL-positive dual-stressed HPdLFs (white) treated with 1.00 µM UNC1999, with (f) indicating the number of apoptotic cells analyzed. (g) H3K27-specific HMT activity measured in dual-stressed PA cultures treated with UNC1999. (h) Microphotographs of adherent THP1 monocytic cells (green) on compressed and P. gingivalis LPS-stimulated HPdLFs after treatment with 1.00 µM UNC1999 in relation to DMSO-treated controls. Cell nuclei were labeled with DAPI (blue). The number of THP1 cells per 10² HPdLFs is displayed in (i). (j,k) Analysis of COX2 expression levels (j) and PGE2 secretion (k) in dual-stimulated PA cultures after treatment with 1.00 µM UNC1999 in relation to DMSO-treated BSA+LPS+CF controls. */#/§ p < 0.05; **/##/§§ p < 0.01; ***/###/§§§ p < 0.001; */**/*** in relation to BSA+LPS (a), DMSO (b,c) and BSA+LPS+CF+DMSO (d,f,g,i–k); #/##/### in relation to PA+LPS (a), 0.25 µM UNC1999 (c) and PA+LPS+CF+DMSO (d,f,g,i–k); §/§§/§§§ in relation to BSA+LPS+CF (a), 0.5 µM UNC1999 (c) and BSA+LPS+CF+UNC1999 (d,f,j). One-way ANOVA and post hoc test (Tukey) were used for analysis. Scale bars: 10 μm in (h) and 20 µm in (e).

Figure 4

Palmitic acid exposure of dual-stressed HPdLFs resulted in decreased H3K27 trimethylation near IL10 promoter regions associated with decreased IL10 expression. (a) Quantitative expression analysis of IL10 in human periodontal ligament fibroblasts (HPdLFs) exposed to palmitic acid (PA) and simultaneously stimulated with compressive force (CF) and P. gingivalis LPS after treatment with 1.00 µM UNC1999 compared to DMSO-treated BSA+LPS+CF control. (b) DNA primer locations in the IL10 gene locus indicating promoter-associated pairs (#1, #2 and #3) and a non-promoter-associated pair (#4). (c) Quantitative analysis of the associations of H3K27me3 with specific IL10 gene regions shown in (b) for dual-stimulated HPdLFs exposed to PA after UNC1999 treatment in comparison to the respective controls. Data were normalized to the sample input and IgG controls. **/## p < 0.01; ***/### p < 0.001; **/*** in relation to BSA+LPS+CF+DMSO; ##/### in relation to PA+LPS+CF+DMSO. One-way ANOVA and post hoc test (Tukey) were used for analysis.