Transcriptional activity of vitamin D receptor in human periodontal ligament cells is diminished under inflammatory conditions

Abstract Background Although vitamin D3 deficiency is considered as a risk factor for periodontitis, supplementation during periodontal treatment has not been shown to be beneficial to date. Human periodontal ligament cells (hPDLCs) are regulated by vitamin D3 and play a fundamental role in periodontal tissue homeostasis and inflammatory response in periodontitis. The aim of this study is to investigate possible alterations of the vitamin D3 activity in hPDLCs under inflammatory conditions. Methods Cells isolated from six different donors were treated with either 1,25(OH)2D3 (0 to 10 nM) or 25(OH)D3 (0 to 100 nM) in the presence and absence of ultrapure or standard Porphyromonas gingivalis lipopolysaccharide (PgLPS), Pam3CSK4, or interferon‐γ for 48 hours. Additionally, nuclear factor (NF)‐κB inhibition was performed with BAY 11‐7082. The bioactivity of vitamin D in hPDLCs was assessed based on the gene expression levels of vitamin D receptor (VDR)‐regulated genes osteocalcin and osteopontin. Additionally, VDR and CYP27B1 expression levels were measured. Results The vitamin D3‐induced increase of osteocalcin and osteopontin expression was significantly decreased in the presence of standard PgLPS and Pam3CSK4, which was not observed by ultrapure PgLPS. Interferon‐y had diverse effects on the response of hPDLCs to vitamin D3 metabolites. NF‐kB inhibition abolished the effects of standard PgLPS and Pam3CSK4. Standard PgLPS and Pam3CSK4 increased VDR expression in the presence of vitamin D3. CYP27B1 expression was not affected by vitamin D3 and inflammatory conditions. Conclusions This study indicates that the transcriptional activity of VDR is diminished under inflammatory conditions, which might mitigate the effectiveness of vitamin D3 supplementation during periodontal treatment.


INTRODUCTION
The secosteroid hormone vitamin D 3 is either produced by the skin under the influence of ultraviolet light of the sun or obtained as a dietary supplement. 1 Following hydroxylation into 25-hydroxyvitamin D 3 (25(OH)D 3 ) in the liver, it is further converted in the kidney by the 1αhydroxylase (CYP27B1) into the most potent metabolite 1,25-dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ). 2 The functions of CYP27B1 are opposed by 24-hydroxylase (CYP24A1), which inactivates 1,25(OH) 2 D 3 . 3 Although kidney and liver are considered to be the main sites involved in vitamin D 3 metabolism, local conversion, and inactivation in various cell types and tissues has been reported. [4][5][6] Binding of 1,25(OH) 2 D 3 to the vitamin D receptor (VDR) leads to expression of genes responsible for mediating the effects of vitamin D 3 . 7 In particular, 1,25(OH) 2 D 3 plays a significant role for bone metabolism by regulating calcium and phosphate concentration. 8 Further, immunomodulatory, anti-inflammatory, and anti-proliferative effects have been reported. 9,10 Due to its various positive effects, vitamin D 3 supplementation has been targeted as a potential adjuvant therapeutic approach for many different inflammatory diseases, including periodontitis. 9,11 This chronic inflammatory disease leads to destruction of tooth supporting tissues including the alveolar bone. 12 Periodontitis is associated with a dysbiosis within the host-microbial interaction and is influenced by various risk factors. 13,14 To date, data concerning association of vitamin D 3 deficiency with periodontitis and the effects of vitamin D 3 supplementation on the treatment outcome are controversial. 15 While some cross-sectional observational studies showed associations between vitamin D 3 deficiency and periodontitis risk, there are longitudinal studies claiming the opposite. [15][16][17] Furthermore, only few studies observed a positive impact of vitamin D 3 supplementation on the outcome of periodontal surgical procedures or on clinical parameters during the maintenance phase. 18,19 To the best of our knowledge, there is no evidence of beneficial effects of vitamin D 3 supplementation during the initial non-surgical treatment phase.
Human periodontal ligament cells (hPDLCs) play an integral part in maintaining periodontal tissue homeostasis. 20 They possess mesenchymal stem cell (MSC) character, exhibiting characteristic MSC surface markers and the ability to differentiate in vitro into various cell types including osteoblasts. 21,22 Several studies have revealed the effects of vitamin D 3 on hPDLCs. For example, these cells express VDR and are able to convert 25(OH)D 3 into 1,25(OH) 2 D 3 via CYP27B1. Additionally, 1,25(OH) 2 D 3 enhances osteogenic and immunomodulatory activity of hPDLCs. Furthermore, both vitamin D 3 metabolites decrease the inflammatory response of hPDLCs to bacterial components. 5,[23][24][25] Given the fact that vitamin D 3 has such beneficial effects on bone metabolism and resolution of inflammation, there is no reasonable explanation for the divergence in the existing data concerning vitamin D 3 as a risk factor for periodontitis and supplementation during periodontal treatment. As hPDLCs have been shown to exhibit compromised functions under inflammatory conditions, this may be a potential factor altering the activity of vitamin D 3 metabolites, but such a possibility was never studied to date. 26 This study aims to investigate the bioactivity of 1,25(OH) 2 D 3 and 25(OH)D 3 in hPDLCs under inflammatory conditions. The bioactivity of vitamin D 3 metabolites was assessed based on the expression of VDR regulated genes such as bone γ-carboxyglutamic acid-containing protein (BGLAP or osteocalcin) and secreted phosphoprotein 1 (SPP1 or osteopontin) on both gene and protein levels. In addition, the gene expression levels of vitamin D 3 -related molecules VDR and CYP27B1 were assessed under the same treatment modalities. Inflammatory conditions were simulated by stimulation with lipopolysaccharide (LPS) of the periodontopathogenic bacterium Porphyromonas gingivalis (Pg) (standard and ultrapure preparations), synthetic Toll-like receptor (TLR)-2 agonist Pam3CSK4, or proinflammatory cytokine interferon (IFN)-γ. To assess the role of nuclear factor (NF)-κB activation, experiments were additionally performed with NF-κB inhibitor BAY 11-7082. We hypothesized that the response of hPDLCs to vitamin D 3 metabolites and the expression of VDR may be altered under inflammatory conditions.

Cell culture
The protocol for isolation of primary hPDLCs applied in this study was approved by the ethics committee of the Medical University of Vienna (ethical approval number: 1694/2015, revised in 2018). Experiments were performed following the "Good Scientific Practice" guidelines of the Medical University of Vienna and the Declaration of Helsinki. Periodontal ligament tissue was obtained from third molars of six periodontally healthy individuals (three females, three males) undergoing tooth extraction as part of their orthodontic treatment plan. The donors, which were white non-smokers aged 18 and 24 years, had no systemic or oral diseases. They were informed in detail about the study and gave their written consent before the procedure. Primary hPDLCs isolation and expansion was executed as reported in our previous study. 27 Subsequently, hPDLCs were cultivated in Dulbeccos modified Eagleťs medium (DMEM), * supplemented with 10% fetal bovine serum (FBS) † and 1% penicillin and streptomycin (P/S), ‡ under humidified conditions at 37 • C. Confirmation of MSC character was performed with flow cytometry, analyzing MSC surface markers (CD29, CD90, CD105, CD146) § and hematopoietic cell surface markers (CD14, CD31, CD34, CD45). ¶

Quantitative polymerase chain reaction
Gene expression analysis of BGLAP, SPP1, and VDR was performed with quantitative polymerase chain reaction (qPCR). All steps were performed with commercially available kits ##  instructions, namely cell lysis, mRNA extraction, transcription into cDNA, and qPCR. In the following, reverse transcription was performed with a thermocycler. ∥∥ The ABI StepOnePlus device was used for qPCR, using following gene expression assays *** : BGLAP, Hs01587814_g1; SPP1, Hs00959010_m1; VDR, Hs_00172113_m1; CYP27B1, Hs_00168017_m1. The analysis was performed in duplicates with following thermocycler † † † conditions: 95 • C for 10 minutes; 40 cycles, each for 15 seconds at 95 • C; 60 • C for 1 minute. To evaluate the mRNA expression levels of BGLAP, SPP1, VDR, and CYP27B1, the cycle threshold (C t ) for each sample was determined and changes in the target gene expression were calculated with the 2 −∆∆Ct method: Untreated hPDLCs served as control group and GAPDH was used as endogenous control.

Enzyme-linked immunosorbent assay
Osteocalcin and osteopontin protein concentrations in the cell culture supernatants were analyzed with enzymelinked immunosorbent assay (ELISA). For this purpose, commercially available kits ‡ ‡ ‡ with a sensitivity of 156.5 pg/mL for osteocalcin and 31.25 pg/mL for osteopontin were used according to the manufacturerťs instructions. Samples were applied in duplicates and measurement of optical densities was performed with a photometer at 450 nm. Subsequently, they were plotted against a standard curve to determine the protein concentrations.

Statistical analysis
Mean values of six different donors were used for statistical analysis, which was executed with SPSS 24.0. § § § Statistical differences between groups were analyzed by ANOVA for repeated measures, followed by post-hoc paired t-test.  Data are presented as mean ± SEM of six different donors. *Significant difference between groups, P <0.05

Effect of vitamin D 3 metabolites on the gene expression of osteocalcin by hPDLCs under physiological and inflammatory conditions
The gene expression levels of osteocalcin induced by 1,25(OH) 2 D 3 in concentrations of 0 nM, 1 nM, and 10 nM alone or in combination with either standard PgLPS(1 µg/mL; Figure   Y-axes show the n-fold expression of osteocalcin expression compared with untreated cells ( = 1). GAPDH served as endogenous control. Data are presented as mean ± SEM of six different donors. *Significant difference between groups, P <0.05

Effect of NF-κB inhibition on vitamin D 3 -induced gene expression of osteocalcin and osteopontin by hPDLCs under physiological and inflammatory conditions
Resulting gene expression levels of osteocalcin in response to 1,25(OH) 2 D 3 (10 nM; Figure 5A) or 25(OH)D 3 (100 nM; Figure 5B) under physiological and inflammatory (1 µg/mL standard PgLPS or 1 µg/mL Pam3CSK4) conditions in the presence and absence of NF-κB inhibitor BAY 11-7082 (0.3 µg/mL) are presented in Figure 5. The reduction of the vitamin D 3 -induced osteocalcin expression by Pam3CSK4  Figure 6 shows the gene expression levels of VDR resulting from stimulation with 1,25(OH) 2 D 3 (0 to 10 nM) in the presence and absence of standard PgLPS (1 µg/mL; Figure 6A), Pam3CSK4 (1 µg/mL; Figure 6B), or IFN-γ (0.1 µg/mL; Figure 6C). Under physiological conditions, 1,25(OH) 2 D 3 tended to decrease gene expression levels of VDR, but not statistically significant. Conversely, 1,25(OH) 2 D 3 significantly increased VDR expression in the presence of standard PgLPS in a dose-dependent manner. Similar tendencies were observed after co-stimulation with 1,25(OH) 2 D 3 and Pam3CSK4, but these effects were not significant. Treatment with IFN-γ resulted in decreased VDR expression, which was independent of co-stimulation with vitamin D 3 metabolites. However, this reduction was also not statistically significant.

Effect of vitamin D 3 metabolites on the gene expression of vitamin D 3 -related molecules by hPDLCs under physiological and inflammatory conditions
The influence of standard PgLPS (1 µg/mL; Figure 6D), Pam3CSK4 (1 µg/mL; Figure 6E

Protein concentration of osteocalcin and osteopontin in conditioned media of hPDLCs treated with vitamin D 3 metabolites under physiological and inflammatory conditions
Osteocalcin and osteopontin protein concentration in supernatants of hPDLCs stimulated with 0 to 10 nM 1,25(OH) 2 D 3 and 0 to 100 nM 25(OH)D 3 in the presence and absence of standard PgLPS (1 µg/mL), Pam3CSK4 (1 µg/mL) or IFN-γ (0.1 µg/mL) were analyzed by ELISAs with a sensitivity of 156.5pg/mL and 31.25pg/mL, respectively, but could not be detected (data not shown). Notably, our experiments were performed in serum-free DMEM, since media supplemented with 10% FBS contain high amounts of osteocalcin, which is over the range of commercial ELISA kits (10 ng/mL).

DISCUSSION
The aim of this study was to assess the bioactivity of 1,25(OH) 2 D 3 and 25(OH)D 3 in hPDLCs under inflammatory conditions, which were simulated by PgLPS (standard and ultrapure), Pam3CSK4 or IFN-γ. Analysis of VDRregulated genes osteocalcin and osteopontin, as well as VDR and CYP27B1 was performed via qPCR. The bone gamma-carboxyglutamic acid-containing protein osteocalcin is encoded by the BGLAP gene, which is directly transcriptionally regulated by VDR. 33 Osteocalcin is highly expressed during differentiation of osteoblasts. 34 Likewise, the production of osteopontin, a sialic acid-rich glycosylated phosphoprotein, is regulated by 1,25(OH) 2 D 3 and strongly promoted when bone matrix is formed. 35 Similarly to bone marrow-derived MSCs, these two osteogenic markers have been shown to be enhanced by 1,25(OH) 2 D 3 in hPDLCs. 24,36 The present in vitro study is the first to evaluate these effects in hPDLCs while simulating inflammatory conditions. On the one hand, such conditions were attained by stimulating the cells with TLR agonists, which have been shown to strongly enhance the inflammatory response of hPDLCs. 27 On the other hand, inflammation was simulated by treatment with IFN-γ, which is known to increase the immunomodulatory ability of hPDLCs and plays a substantial role in bone loss by mediation of the Th1-type response. 37,38 Evaluating treatment with vitamin D 3 metabolites under physiological conditions, our results show that not only 1,25(OH) 2 D 3 , but surprisingly also 25(OH)D 3 significantly enhances the gene expression of osteocalcin by hPDLCs ( Figures. 1 and 2). The effects of the vitamin D 3 metabolites on osteopontin gene expression levels were clearly less pronounced ( Figures. 3 and 4). Interestingly, such strong effects of 25(OH)D 3 on osteocalcin and osteopontin gene expression have so far only been reported in bone-marrow derived MSCs starting from a concentration five times higher than in the present study. 39 In contrast, in our study this effect was observed at 25(OH)D 3 concentrations similar to serum levels, which are physiologically relevant. This suggests that the alteration of systemic vitamin D 3 levels might influence local periodontal homeostasis.
Transcriptional activation of VDR-regulated genes induced by vitamin D 3 metabolites was drastically affected by proinflammatory stimuli. The most pronounced effect was observed upon Pam3CSK4 treatment, which significantly diminished the expression of osteocalcin and osteopontin induced by 1,25(OH) 2 D 3 and the expression of osteocalcin induced by 25(OH)D 3 ( Figures. 1  through 4). Pam3CSK4 is a synthetic TLR-2 agonist, which elicits an even stronger inflammatory response of periodontal cells than TLR-4 agonists. 27 As shown in TLR-2 deficient mice, activation of TLR-2 is furthermore crucial in mediating bone loss in periodontitis. 40 Our data showed that activation of NF-κB might regulate the activity of VDR in hPDLCs. This effect might be due to the inhibition of Wnt-and bone morphogenetic protein-signaling pathways. 41 Alternatively, this could be due to the inhibitory effect of inflammatory stimuli on the retinoid x receptor, which facilitates VDR activation by 1,25(OH) 2 D 3 . 42,43 Interestingly, VDR is known to inhibit NF-κB transcriptional activation and cytokine production in periodontal ligament cells. 25,44 Thus it seems that VDR and NF-κB inhibit their activation in a reciprocal way.
Such interaction of these transcription factors might play an important role in different inflammatory diseases and particularly periodontitis.
The effect of standard PgLPS on the expression of VDR regulated genes was qualitatively similar to those of Pam3CSK4, but quantitatively less pronounced. These differences might be explained by the fact that standard PgLPS is a less potent activator of NF-κB response in hPDLCs than Pam3CSK4. 27,32,45 Furthermore, ultrapure PgLPS had no significant effect on vitamin D 3 induced response in hPDLCs (see Figure S1 in online Journal of Periodontology). A previous study shows that standard P. gingivalis LPS activates both TLR-2 and TLR-4, whereas ultrapure PgLPS acts exclusively through TLR-4. Moreover, standard PgLPS induces markedly stronger inflammatory response in monocytes and hPDLCs (Behm et al. , not yet published). 46 Thus, inhibition of the vitamin D 3induced response seems to depend on the degree of NF-κB activation. This assumption is supported by our findings that the Pam3CSK4-and standard PgLPS-induced decrease of the vitamin D 3 -triggered response in hPDLCs is recovered under NF-κB inhibition. Furthermore, it could also explain the observation that the response to vitamin D 3 was not inhibited by IFN-γ, which is a rather weak activator of NF-κB signaling. 47 We further found that the gene expression levels of VDR were not affected by 1,25(OH) 2 D 3 and 25(OH)D 3 (Figure 6), which is in contradiction with the study of Tang et al. who found a 3-fold higher VDR expression after treatment with 10 nM 1,25(OH) 2 D 3 . 48 This discrepancy could be explained by the fact that Tang and colleague used osteogenic induction medium supplemented with 10-mM β-glycerophosphate, 50-µg/mL ascorbic acid, 10 −7 M dexamethasone, and 20% FBS and treated the cells for 6 days. 48 These artificial additives of osteogenic medium might influence the responsiveness to vitamin D 3 , but this is a question that needs to be further explored. TLR agonists standard PgLPS and Pam3CSK4 had no influence on the gene expression levels of VDR ( Figure 6). This observation is in accordance with the findings of Nebel et al., who treated hPDLCs with Escherichia coli LPS. 24 However, 1,25(OH) 2 D 3 significantly enhanced VDR expression in the presence of standard PgLPS ( Figure 6). Qualitatively similar effects have been demonstrated by Pramanik et al. in human monocytic THP-1 cells after combined treatment with 1,25(OH) 2 D 3 and lipopolysaccharide. 49 This suggests some potential interactions between TLR and VDR responses, which should be further investigated.
The inhibitory effects of Pam3CSK4 and standard PgLPS on the gene expression of VDR-regulated genes seem to be independent of the expression of proteins involved in vitamin D metabolism. Inhibition of VDR regulated genes was not accompanied by the decrease of VDR gene expression levels in hPDLCs. Therefore, the effects of Pam3CSK4 and standard PgLPS are associated with inhibition of VDR transcriptional activity. As shown in supplementary Figure S2 in online Journal of Periodontology, gene expression levels of CYP27B1, which is involved in local conversion of 25(OH)D 3 into 1,25(OH) 2 D 3 in hPDLCs, are not affected by Pam3CSK4 or standard PgLPS. This finding is underlined by the fact that both TLR agonists diminished the response to 1,25(OH) 2 D 3 and 25(OH)D 3 by a similar extent. The effect of IFN-γ on the expression of VDR regulated genes was less obvious than that of the TLR agonists as can be seen in Figures 1 through 4. For example, IFN-γ led to a slightly increased osteocalcin expression induced by 10 nM 1,25(OH) 2 D 3 , but this effect was not significant. In contrast, osteocalcin expression triggered by 25(OH)D 3 was significantly diminished by IFN-γ. The different effects of IFN-γ on the 1,25(OH) 2 D 3 and 25(OH)D 3 -induced response imply that this cytokine influences the bio-activation of 25(OH)D 3 by hPDLCs. This assumption is supported by previous data showing that CYP27B1 is regulated by IFN-γ in macrophages, vascular endothelial cells, and keratinocytes. 50 The possibility of such a regulation in hPDLCs, as well as its potential physiological importance needs to be elucidated by future studies.
In our experiments, neither osteocalcin nor osteopontin could be detected in the supernatants by ELISA, which could be explained by several factors. First of all, our study was conducted with primary undifferentiated cells to more closely resemble the in vivo situation and therefore may result in low protein contents that are not detectable by ELISA (lower detection limit: 156.5 pg/mL). Furthermore, our treatment time was rather short compared with other studies, which could contribute to the low protein level. Lastly, our treatment with vitamin D 3 was done in the absence of different additives like serum, dexamethasone, and ascorbic acid, which are usually present in osteogenic medium, but are rather artificial and non-relevant physiologically. According to our measurements, medium supplemented with 10% serum contains more than 10 ng/mL osteocalcin, which impedes analysis of this protein with ELISA. Our experimental protocol facilitates to discriminate between the effects of vitamin D 3 metabolites and avoids the influence of other artificial factors, but does not allow detection of protein production. Nevertheless, our data clearly suggest that transcriptional activation of vitamin D 3 regulated genes is clearly affected by inflammatory conditions. Several reviews and original articles about the relationship between vitamin D 3 deficiency and periodontal disease have been conducted so far. However, most of the data were collected in observational cross-sectional studies and casual associations between vitamin D 3 deficiency and periodontitis remain unclear. 15 In particular, there is no clear evidence for beneficial effects of vitamin D 3 supplementation during periodontal therapy, especially in the initial phase. Such observations are rather surprising, considering the potentially positive impact of vitamin D 3 on maintaining periodontal health, such as induction of antibacterial mechanisms, strengthening of the physical barrier, inhibition of inflammatory processes, and support of wound healing. 15,34 Thus, our study may provide an explanation for the lacking benefit of vitamin D 3 during initial periodontitis treatment, in which inflammation is markedly pronounced.

CONCLUSIONS
Summarizing the results of the present study indicates that the bioactivity of vitamin D 3 might be diminished in periodontal tissue of periodontitis patients, which might be due to disruption of the vitamin D 3 metabolism. Therefore, it may be possible that the beneficial effects of vitamin D 3 supplementation are dampened in those individuals.
Revealing the underlying mechanisms of diminished vitamin D 3 bioactivity during inflammation could be a future goal to enhance the effectiveness of vitamin D 3 supplementation as adjunctive periodontal therapy.

A C K N O W L E D G M E N T S
This work was supported by the authors' institution and the Austrian Science Fund (Project P 29440 to OA). Further, the authors would like to thank Mrs. Phuong Quynh Nguyen for excellent technical assistance and Dr. Vivian Hirsch for proofreading the manuscript. The authors report no conflicts of interest related to this study.