Adiponectin Interacts In-Vitro With Cementoblasts Influencing Cell Migration, Proliferation and Cementogenesis Partly Through the MAPK Signaling Pathway

Current clinical evidences suggest that circulating Adipokines such as Adiponectin can influence the ratio of orthodontic tooth movement. We aimed to investigate the effect that Adiponectin has on cementoblasts (OCCM-30) and on the intracellular signaling molecules of Mitogen-activated protein kinase (MAPK). We demonstrated that OCCM-30 cells express AdipoR1 and AdipoR2. Alizarin Red S staining revealed that Adiponectin increases mineralized nodule formation and quantitative AP activity in a dose-dependent manner. Adiponectin up-regulates the mRNA levels of AP, BSP, OCN, OPG, Runx-2 as well as F-Spondin. Adiponectin also increases the migration and proliferation of OCCM-30 cells. Moreover, Adiponectin induces a transient activation of JNK, P38, ERK1/2 and promotes the phosphorylation of STAT1 and STAT3. The activation of Adiponectin-mediated migration and proliferation was attenuated after pharmacological inhibition of P38, ERK1/2 and JNK in different degrees, whereas mineralization was facilitated by MAPK inhibition in varying degrees. Based on our results, Adiponectin favorably affect OCCM-30 cell migration, proliferation as well as cementogenesis. One of the underlying mechanisms is the activation of MAPK signaling pathway.


INTRODUCTION
The periodontium consists of the gingiva, periodontal ligament, root cementum and alveolar bone (Park et al., 2017). Cementum, a heterogeneous mineralized layer covering the entire root dentin surface, anchors fibrous connective tissues on tooth-root surfaces (Caverzasio and Manen, 2007). As a special mineralized tissue, cementum has a similar composition to bone, consisting of approximately 61% mineralized as well as 27% organic matrix and 12% water (Nanci and Bosshardt, 2006).
It has been demonstrated that Adipocytokines, which are mainly secreted by adipose tissue, can influence bone metabolism (Funahashi et al., 1999). Adiponectin, which is produced by adipocytes but also by salivary gland epithelial cells, has been shown to modulate metabolic and immune functions of salivary gland epithelial cell (Katsiougiannis et al., 2006).
Two Adiponectin receptor (AdipoRs) subtypes (Adiponectin receptor 1 and 2 (AdipoR1 and AdipoR2)), expressed on various tissues and cells such as human gingival fibroblasts and PDL cells (Yamauchi et al., 2003a) mediate the biological effects of Adiponectin. Yamaguchi et al. (2012) showed that the AdipoRs were ubiquitously expressed in the oral tissues of mice such as gingiva, tongue, buccal mucosa and labial mucosa, as well as in RAW24 cells (Yamauchi et al., 2012). Both AdipoRs exert similar effects but have individual signaling preferences. Whereas phosphorylation of Extracellular Regulated Kinase 1/2 (ERK1/2) may depend on both receptors, AdipoR1 is more prominent in AMP-activated protein kinase (AMPK) phosphorylation, and AdipoR2 is involved in PPAR-α activation in muscle cells (Buechler et al., 2010). AdipoR1 and AdipoR2 have been identified in the plasma membrane (Yamauchi et al., 2003b) as well as in the cytoplasm (cytoplasmic puncta) (Ding et al., 2009).
In-vivo, Adiponectin secretion is regulated by the degree of obesity (Carbone et al., 2012). Its secretion is inversely proportional to the amount of adipose tissue (Swarbrick and Havel, 2008;Benedix et al., 2011;de Carvalho et al., 2017). It has been shown that patients with type 2 diabetes and obesity have lower serum Adiponectin levels than normal weight individuals (Cao et al., 2015). Furthermore, clinical studies have analyzed the relationship between the body mass index (BMI) and salivary Adiponectin levels (Mamali et al., 2012;Nigro et al., 2015) and found a trend toward slightly higher expression of salivary Adiponectin in obese patients than in controls, however they could not show a statistic significance between groups, whereas they previously reported that Adiponectin serum levels are reduced in obese patients compared to healthy controls. Also, patients with severe periodontitis had lower Adiponectin-and AdipoRs-levels than healthy subjects (Saito et al., 2008;Yamaguchi et al., 2010).
Recent studies revealed that local submucosal injection of Adiponectin prevented experimental orthodontic tooth movement in rats (Haugen et al., 2017). The data suggested that Adiponectin is involved in the homeostasis of periodontal tissues and thus might influence orthodontic treatment.
It has been demonstrated that Adiponectin and its receptors are also expressed in bone marrow-derived osteoblasts and adipocytes (Seo et al., 2004), suggesting that Adiponectin may play a role in bone metabolism (Chen et al., 2013). Considering the multifunctional role of Adiponectin, it seems possible that it may have functional characteristics in cementoblasts similar to that in other mineralized tissue-forming cells such as osteoblasts (Seo et al., 2004).
Since fat mass may impact peripheral bone formation, it may be one of the critical factors for orthodontic tooth movement rate and orthodontically-induced tooth root resorption. Adipocyte-derived hormones such as Adiponectin may contribute to this relationship. To date, the physiological effect of Adiponectin on cementoblasts has not been elucidated. The decreased levels of Adiponectin, as found in obese individuals, might be a critical pathomechanistic link. Therefore, the present study was undertaken to investigate the effect of Adiponectin on OCCM-30 cells in vitro.

Cell Migration Assay
OCCM-30 cells were plated at a density of 8 × 10 3 cells/well in 6well plates (657160, Greiner Bio-one), in α-MEM (11095-080, Gibco) containing 10% FBS (10270-106, Gibco) and 1% Penicillin/Streptomycin (15140-122, Gibco) and cultured until confluence. Cells were preincubated for 12 h in starvation medium and wounded by scratching using a 100 μL tip. Through this, a cell-free area was created in the center of the cell layer. Afterwards, all non-adherent cells were washed with 1× PBS (10010023, Thermo-Fisher). The wounded cell monolayers were incubated in the presence and absence of different concentration of Adiponectin (50636-M08H, Sino Biological Inc.) for 24 h. Wounded-area images were taken immediately after wounding and 24 h after scratching. The wounded cell layers were photographed at ×10 magnification (Leica Microsystems, Wetzlar, Germany) and the percentages of wound closure area between cell layer borders were analyzed and calculated over time using the Image J software (National Institutes of Health and University of Wisconsin, United States).

2H-tetrazolium (MTS) assay (CellTiter 96 ® Aqueous One Solution
Cell Proliferation Assay, Promega) according to manufacturer's instructions. Briefly, OCCM-30 cells at a passage three to five were seeded at a density of 5 × 10 3 cells/well in a 96-well plate (655180, Greiner Bio-one). Cells were cultured in α-MEM containing 5% FBS overnight to allow adherence. Then, cells were washed twice with 1× PBS (10010023, Thermo-Fisher) and treated with various concentrations of Adiponectin (50636-M08H, Sino Biological Inc.) in α-MEM containing 0.5% FBS over a period of 24 h. To assess involvement of the MAP kinase cascade in Adiponectin-induced proliferation, cells were pretreated 1 h with the inhibitors: SB203580 (InvivoGen), FR180204 (Calbiochem) and SP600125 (InvivoGen) at a concentration of 1.0 μg/mL as well as with DMSO (D8418-50ML, Sigma-Aldrich) at 0.1% (v/v), respectively. Thereafter, 20 μL of the MTS reagent was added into each well and the Frontiers in Pharmacology | www.frontiersin.org December 2020 | Volume 11 | Article 585346 cells were incubated during 2 h at 37°C in a 5% CO 2 atmosphere. Plates were read by 490 nm using a 96-well micro-plate reader (BioTek, Winooski, VT, United States) to measure the amount of formazan by cellular reduction of MTS.
To quantify the degree of calcium accumulation in the mineralized extracellular matrix, Alizarin Red S stained cultures were dissolved using 100 mM Cetylpyridinium chloride (6004-24-6, Sigma-Aldrich) for 1 h to release calciumbound dye into the solution. The absorbance of the released dye was measured at 570 nm using a spectrophotometer (xMarkTM, Microplate Absorbance Spectrophotometer, 1681150 BioRad).

Alkaline Phosphatase Enzymatic Activity Assay
After cementogenesis induction during 14 days, OCCM-30 cells were lysed in distilled deionized water and sonicated for 15 s (SONIFIER 150, BRANSON, G. HEIHEMANA). The lysate was incubated at 37°C for 30 min with p-Nitrophenyl phosphate (p-NPP; Alkaline phosphatase Substrate, Sigma) in an alkaline phosphatase buffer solution (1.5 mM). The reaction was stopped by adding NaOH, and absorbance was read at 405 nm (xMarkTM, Microplate Absorbance Spectrophotometer, 1681150 BioRad).

Statistical Analysis
Statistical analyses were performed using GraphPad Prism 6.0 software (GraphPad software). All values are expressed as means ± standard deviation (SD) and analyzed using oneway t-test for unpaired samples to determine the statistically significant differences between groups. Differences were considered statistically significant at a p value of < 0.05. Data distribution was analyzed using the Kolmogorov-Smirnov and the Shapiro-Wilk test and visually using QQ plots. All experiments were repeated successfully at least three times.

Cementoblasts Express Adiponectin Receptor 1 and 2
First, we aimed to verify if OCCM-30 cementoblasts express Adiponectin receptors. By Western blot analysis, we could establish that AdipoR1 as well as AdipoR2 are expressed on this cell line ( Figure 1A). The mRNA expression of Adiponectin receptors was also demonstrated by RT-PCR analysis ( Figure 1B). Immunofluorescence staining show that AdipoR1 are mostly expressed in the cytoplasm, cytomembrane and nucleus, while AdipoR2 are expressed around the nucleus ( Figure 1C).

Adiponectin Promotes In-Vitro Cementoblast Mineralization
Second, we analyzed the possible effect that exogenous Adiponectin exerts during cementogenesis. Alizarin Red S staining was used to visualize and quantify the biological effect of Adiponectin on OCCM-30 cell mineralization. This method revealed that Adiponectin significantly increased mineralized nodule formation in a dose-dependent manner over a period of 14 days (Figure 2A). Colorimetric analysis revealed that Adiponectin-stimulated OCCM-30 cells had higher levels of mineralized matrix production in comparison to unstimulated cells (p < 0.01) (Figures 2B,C).
The analysis of the Alkaline Phosphatase enzymatic activity (AP) over a period of 48 h of cells stimulated with different concentrations of Adiponectin, showed increased AP activity time and dose-dependently, reaching statistical significance (p < 0.01) after 24 h in the group stimulated with 80 ng/ml Adiponectin ( Figure 2D).
Cells cultivated for a period of 7 days in a mineralizationinducing medium, were afterwards stimulated over a period of 3 h with Adiponectin (100 ng/ml). The kinetic analysis of the relative mRNA expression of AP, Runx-2, BSP, OPG, OCN and F-Spondin increased notably, reaching statistical significance after 45 min of stimulation with Adiponectin (100 ng/ml) in comparison to timepoint 0 min. These stimulatory effects were sustained over the entire period of 3 h (p < 0.05) ( Figure 2E).

Elevated Levels of Adiponectin Facilitate Cell Migration and Proliferation
Next, we analyze the effect that Adiponectin exerts on cell proliferation and migration. Cells were grown to 100% confluency and then were scratched using a 100 µL pipet tip. Immediately thereafter, cells were stimulated with different concentrations of Adiponectin (0, 0.1, 0.5, 1, and 2 μg/ml) Frontiers in Pharmacology | www.frontiersin.org December 2020 | Volume 11 | Article 585346 during 24 h. The cell migration ability was visualized and measured using microscopic photography ( Figure 3A). The analysis of the recovered data indicates that Adiponectin at concentrations of 1 and 2 μg/ml significantly promotes wound closure (24.35 ± 2.38 and 30.3 ± 2.68%, respectively) ( Figure 3B). Furthermore, we observed that cells stimulated with Adiponectin over a period of 24 h have increased mitogenic activity. The groups stimulated with 0.4, 0.8 and 1.6 μg/ml adipokine showed a significantly increased proliferation rate (p < 0.05) in comparison to unstimulated cells ( Figure 3C).

Adiponectin Promotes P38, ERK1/2 and JNK Phosphorylation in OCCM-30 Cells
In order to elucidate if Adiponectin can activate the MAPK pathway, we performed a kinetic analysis of P38, ERK1/2 and JNK protein phosphorylation ( Figure 4A). Western blots revealed that P38 phosphorylation occurs 5 min after Adiponectin (20 ng/ml) stimulation. The phosphorylatedstate of P38 was sustained over a period of 4 h, reaching a peak at time point 10 min (p < 0.001). The phosphorylation of ERK1/2 as well as P54/P46 JNK reached a peak after 5 min Adiponectin Frontiers in Pharmacology | www.frontiersin.org December 2020 | Volume 11 | Article 585346 5 addition, being the WB's bands detectable during 30 min in both cases ( Figure 4B). The WB analysis revealed that cementoblasts express STAT1 and STAT3 being these proteins target of Adiponectin stimulation. The phosphorylation of STAT1 and STAT3 occurred likewise after 5 min and was detectable over a period of 30 min ( Figures 4C,D).
The MTS assay indicated that MAPK blockade reduced the cell proliferation in unstimulated and stimulated cells. However, this effect was partially counteracted by Adiponectin (100 ng/ml) (Figures 5A,B).
As shown in the pictures, the scratched areas were measured after 24 h, which is represented by a front-end yellow edge line ( Figure 5C). The migratory capacity of OCCM-30 cells treated with MAPK inhibitors was examined in the presence or absence of Adiponectin (100 ng/ml). As results, the migration rate of cells was significantly attenuated after ERK1/2 as well as JNK blockade despite Adiponectin (100 ng/ml) co-stimulation, whereas this effect was not observed in the group pretreated with the P38 inhibitor ( Figure 5C).

Inhibition of MAP Kinases Alter Adiponectin-Induced Cementogenesis
The colorimetric analysis performed after dilution of Alizarin Red S staining for a period of 7 days mineralization induction, did not show significant differences among groups despite Adiponectin treatment ( Figures 6A,B). Cells treated with P38 and ERK1/2 inhibitors during 14 days in the absence of Adiponectin (100 ng/ml) microscopically exhibited increased mineralization. In the presence of Adiponectin (100 ng/ml), such effect was slightly but significantly decreased ( Figures 6A,B). On the contrary, the sustained blockade of JNK did not alter mineralized matrix production on OCCM-30 cells despite Adiponectin addition.

DISCUSSION
In the present in vitro study, we demonstrated that OCCM-30 cementoblasts express Adiponectin receptor 1 and 2 and that its activation facilitates cell migration, proliferation and mineralization. These effects are partially orchestrated via activation of the MAPK signaling pathway.
Clinical studies have shown that the normal salivary Adiponectin levels are around 10.92 (3.22-28.71) ng/ml and the normal serum levels are around 12.27 (8.15-14.70) μg/ml (Mamali et al., 2012). These values drastically decrease in case of obesity (Carbone et al., 2012). Reduced levels of Adiponectin as well as AdipoRs are also present in patients with severe periodontitis, a fact that suggests impaired Adiponectin function to be associated with disease severity (Saito et al., 2008;Yamaguchi et al., 2010). In our experiment, we focused Previous reports have shown the favorable effects Adiponectin exerts in vitro on the proliferation of human osteoblasts via the MAPK signaling pathway (Luo et al., 2005). In the present study, we observed over a period of 24 h, that the number of viable cementoblasts was significantly increased in presence of higher concentration of Adiponectin compared to unstimulated cells. Similar results were also obtained by Berner et al. (2004) showing that a supplementation of cell culture medium with recombinant Adiponectin enhances the proliferation of murine osteoblasts (Berner et al., 2004). In contrast to the study performed by Kanazawa et al. (2007) reporting that even low concentrations of Adiponectin stimulate the proliferation of the MC3T3-E1 osteoblast cell line (Kanazawa et al., 2007), we observed that . This effect was rescued after Adiponectin (100 ng/ml) stimulation. (C) Images of cell migration were captured prior to stimulation (0 h) and after 24 h. The recovered areas were calculated by ImageJ software comparing the scratched areas. Blockade of ERK1/2 (FR180204) and JNK (SP600125) over a period of 24 h significantly decreases cell migration whereas P38 inhibition did not significantly influence the migration rate. Values are expressed as means ± SD of three independent results: Ns (not significant), *p < 0.05; **p < 0.01; ***p < 0.001.
Frontiers in Pharmacology | www.frontiersin.org December 2020 | Volume 11 | Article 585346 9 FIGURE 6 | Inhibition of MAP kinases alter Adiponectin-induced cementogenesis. (A) Cells were treated with MAP kinase inhibitors in the presence or absence of Adiponectin (100 ng/ml) for 7 and 14 days, respectively. (B) Graphic shows the absorbance (OD 562 ) of Alizarin Red S dilution after induction of mineralization for 7 and 14 days. Mineralization grade does not differ significantly among the groups after 7 days induction. However, after 14 days induction, the sustained blockade of P38 (SB203580) and ERK1/2 (FR180204) facilitate mineralized nodule formation, effect that was partially counteracted by Adiponectin. This effect was not observed in the groups cultivated with the JNK inhibitor (SP600125) in the presence or absence of Adiponectin. Values are expressed as means ± SD: Ns (not significant), *p < 0.05; **p < 0.01; ***p < 0.001.
In our experiment we demonstrated that the expression of Runx-2, an essential transcription factor for osteoblast differentiation, mineralization and migration (Bosshardt, 2005;Liu and Lee, 2013) as well as for cementogenesis (Hakki et al., 2018) was strongly up-regulated in OCCM-30 cells in response to high concentrations of Adiponectin. Adiponectin-induced Runx-2 upregulation in cementoblasts may be interpreted in the circumstances of wound healing and cementum repair and regeneration, because the regulatory role that Runx-2 exerts in cementum formation (Kimura et al., 2018). F-Spondin, a cementoblast specific gene that orchestrate cementoblast differentiation (Kitagawa et al., 2012), was also up-regulated by Adiponectin. Likewise, the OCCM-30 cells presented upregulated mRNA expression of AP, BSP, OCN and OPG after adipokine stimulation.
Alkaline phosphatase (AP) is essential for osteoblast (Koh et al., 1997) and cementoblasts (Hakki et al., 2018) function and is up-regulated during mineralization, thus, it is generally considered as an early marker for osteoblasts as well as for cementoblasts (Carvalho et al., 2012). It is well known that AP regulates the formation of cementum and is involved more in the formation of acellular than cellular cementum . Here we could observe that Adiponectin strongly increases mRNA AP as well as AP enzymatic activity in cementoblasts.
Osteocalcin (OCN) is a non-collagenous protein that plays a regulatory role during the mineralization process (Tokiyasu et al., 2000;Thomson et al., 2003). Its expression is restricted to cells with mineralizing capacity, including osteoblasts, odontoblasts as well as cementoblasts (McKee and Nanci, 1996;Saygin et al., 2000). In our results, the up-regulation of OCN after Adiponectin stimulation clearly indicates that this adipokine favorably induces cementogenesis. Our in vitro results are consistent with previous data supporting the fact that Adiponectin can enhance the osteogenic differentiation of osteoblasts (Lee et al., 2009) as well as promote osteoblastogenesis of HPDL cells (Iwayama et al., 2012).
We could observe that Adiponectin strongly promotes mineralization in OCCM-30 cells. In concordance with these results, Wang et al. (2017) demonstrated the osteogenic capacity of Adiponectin in vitro using rat mesenchymal stem cells (Wang et al., 2017). Another study showed that Adiponectin enhanced the mRNA expression of AP and the mineralization of osteoblasts (Oshima et al., 2006). Iwayama et al. (2012) also observed that Adiponectin promotes AP and Runx-2 mRNA expression as well as up-regulate the AP enzymatic activity of HPDL cells (Iwayama et al., 2012). However, in the latter study, extremely high concentrations (5-10 μg/ml) of Adiponectin were used to observe these effects (Iwayama et al., 2012). In our experiment, we used ranges of Adiponectin concentrations of among 100 ng/ml, demonstrating that murine cementoblasts require lower concentrations of this adipokine to achieve the same results. However, these concentrations still exceed the biologic concentrations of Adiponectin in saliva or in crevicular fluids (Mamali et al., 2012). In the presence of biological concentrations of Adiponectin, we could observe a moderate but not statistically significant increase of AP, BSP, OCN and OPG (Data not shown).
Our results indicate that in OCCM-30 cells, Adiponectin signals act through activation of ERK1/2, JNK as well as P38. Moreover, our present results show that activation of MAPK pathway is essential for Adiponectin-stimulated OCCM-30 proliferation and migration as these effects were partially counteracted when SB203580, FR180204 and SP600125, three highly selective inhibitors of the MAP kinase cascade, were applied. On the contrary, we observe that the sustained inhibition of P38 and ERK1/2 significantly enhances the mineralization rate of cementoblasts, effect that was partially counteracted by Adiponectin addition. These results accord with previous studies showing that long-term inhibition of the MAPK signaling pathway promotes early mineralization as well as the increase of AP activity in preosteoblastic cells (Higuchi et al., 2002).
Several in vitro studies have shown that Adiponectin plays a role as regulator of the MAP kinase pathways in cell homeostasis (Luo et al., 2005;Miyazaki et al., 2005;Junker et al., 2017). (Miyazaki et al., 2005) shown that Adiponectin activates the P38 and JNK pathways in myocytes (Miyazaki et al., 2005). In osteoarthritis, Adiponectin induces P38-MAPK to promote osteophyte formation (Junker et al., 2017). Interesting, Luo et al. (2005) showed that Adiponectin can activate P38 and JNK but not ERK1/2 on primary osteoblasts and that suppression of AdipoR1 abolished Adiponectin induced cell proliferation, suggesting that cell proliferation is regulated by AdipoR/JNK signaling whereas differentiation is mediated via the AdipoR/P38 cascade (Luo et al., 2005).