Cyclic compression emerged dual effects on the bone homeostasis of LPS-induced inflammatory human periodontal ligament cells according to loading force

Background: Appropriate mechanical stimulation is essential for bone homeostasis in healthy periodontal tissues. While the bone homeostasis of inflammatory periodontal tissues under different dynamic loading has not been yet clear. The aim of this study is to clarify the inflammatory, osteogenic and pro-osteoclastic effects of different cyclic stress loading on the inflammatory human periodontal ligament cells (hPDLCs). Methods: hPDLCs were isolated from healthy premolars and cultured in alpha minimum Eagle’s medium (α-MEM). Lipopolysaccharides (LPS) were used to induce the inflammation state of hPDLCs in vitro. Determination of LPS concentration for the model of inflammatory periodontium was based on MTT and genes expression analysis. Then the cyclic stress of 0, 0-50, 0-90 and 0-150 kPa was applied to the inflammatory hPDLCs for 5 days respectively. mRNA and protein levels of osteogenic, osteoclastic and inflammation-related markers were examined after the treatment. Results: MTT and RT-PCR results showed that 10 μg/ml LPS up-regulated TNF-α, IL-1β, IL-6, IL-8 and MCP-1 mRNA levels (P<0.05) and did not affect the cell viability (P>0.05). The excessive loading of stress (150 kPa) with or without LPS strongly increased the expression of inflammatory-related markers TNF-α, IL-1β, IL-6, IL-8, MCP-1 (P<0.05) and osteoclastic markers RANKL, PTHLH and CTSK compared with other groups (P<0.05), but had no significant effect on osteogenic genes. While 0-90 kPa cyclic pressure could up-regulate the expression of osteogenic genes ALP, COL-1 and RUNX2 in the healthy hPDLSCs. Conclusions: Collectively, it could be concluded that 0-150 kPa was an excessive stress loading which accelerated both inflammatory and osteoclastic effects, while 0-90 kPa may be a positive factor for the bone homeostasis of hPDLCs in vitro Abstract Background: Appropriate mechanical stimulation is essential for bone homeostasis in healthy periodontal tissues. While the bone homeostasis of inflammatory periodontal tissues under different dynamic loading has not been yet clear. The aim of this study is to clarify the inflammatory, osteogenic and pro-osteoclastic effects of different cyclic stress loading on the inflammatory human periodontal ligament cells (hPDLCs). Methods: hPDLCs were isolated from healthy premolars and cultured in alpha minimum Eagle’s medium (α-MEM). Lipopolysaccharides (LPS) were used to induce inflammatory periodontium was based on MTT and genes expression analysis. Then the cyclic stress of 0, 0-50, 0-90 and 0-150 kPa was applied to the inflammatory hPDLCs for 5 days respectively. mRNA and protein levels of osteogenic, osteoclastic and inflammation-related markers were examined after the treatment. Results: MTT and RT-PCR results showed that 10 μg/ml LPS up-regulated TNF-α, IL- 1β, IL-6, IL-8 and MCP-1 mRNA levels ( P <0.05) and did not affect the cell viability ( P >0.05). The excessive loading of stress (150 kPa) with or without LPS strongly increased the expression of inflammatory-related markers TNF-α , IL-1β , IL-6 , IL-8 , MCP-1 ( P <0.05) and osteoclastic markers RANKL , PTHLH and CTSK compared with other groups ( P <0.05), but had no significant effect on osteogenic genes. While 0-90 kPa cyclic pressure could up-regulate the expression of osteogenic genes ALP, COL-1 and RUNX2 in the healthy hPDLSCs. Conclusions: Collectively, it could be concluded that 0-150 kPa was an excessive stress loading which accelerated both inflammatory and osteoclastic effects, while 0-90 kPa may be a positive factor for the bone homeostasis of hPDLCs in vitro.

Conclusions: Collectively, it could be concluded that 0-150 kPa was an excessive stress loading which accelerated both inflammatory and osteoclastic effects, while 0-90 kPa may be a positive factor for the bone homeostasis of hPDLCs in vitro.
Keywords: periodontitis; hPDLCs; LPS; dynamic loading; bone homeostasis Background Periodontitis is a chronic infective disease of the periodontium caused by bacteria. It especially occurs among the elderly, and may develop into the defect of dentition, which is one of the severest consequences. To repair the missing teeth for these patients, the control of inflammation state and occlusal force on the involved teeth is the key point which should be well considered. However, there is no conclusion about the differences between periodontitis and healthy abutments under the dynamic mechanical stress, and the range of occlusal force that periodontitis teeth can bear. What's more, the bone homeostasis of the inflammatory periodontal tissue under different dynamic loading has also not been clear yet.
Endotoxin is an important toxic component in the occurrence and development of periodontitis. When human periodontal ligament cells (hPDLCs) were exposed to Lipopolysaccharides (LPS), the major active component of endotoxins, the expressions of pro-inflammatory cytokines was increased [1].
TNF-α, IL-1β, IL-6, IL-8, IL-10, IL-11 etc are the pro-inflammatory cytokines secreted to cause inflammatory response, loss of periodontium and alveolar bone, which would eventually lead to irreversible teeth loosening and falling off [2][3][4] . Kato reported that 1 and 10 μg/ml LPS could affect osteoblastic differentiation and up-regulate IL-1β, IL-6, and IL-8 production in human periodontal ligament stem cells (hPDLSCs) [5]. Besides, in the research of Liu, applying LPS on hPDLCs could trigger the inflammation reaction [6] . However, the concentration and duration of LPS treatment for modeling the periodontitis in hPDLCs in vitro preferably was remained to be clear and definite.
The periodontal ligament (PDL) which mainly contains fibroblasts is connecting the root and alveolar bone and responsible for the formation of collagen fiber networks. Meanwhile, a few osteoblast-like fibroblasts in PDL has the capacity to give rise to bone cells and cementoblasts [7]. Because of the components, PDLCs are able to bear physiological mastication. In recent studies, the mechanical loading within the physiological range has been found to stimulate the differentiation of PDLCs in vitro [8,9]. Compression is the way that scholars simulate the stress state of periodontal ligament cells under normal occlusion. In PDLCs, 65g/cm2 (245 kPa) static compression given by weight could take part in the initiation of osteoclastogenesis. It could up-regulate the expression of pro-osteoclastogenic cytokine, like receptor activator for nuclear factor-κ B ligand (RANKL) and parathyroid hormonerelated protein(PTHrP), and the pro-inflammation cytokines including IL-8 and IL-11 [10,11]. It has been demonstrated that 150 psi (1034 kPa) static compression by air pressure could up-regulate the expressions of MMPs-1/7/9, which are the cytokines regulating the degradation of extracellular matrix, and inflammation-related genes in inflammatory hPDLCs [12]. However, there is no defined physiological pressure range for hPDLCs because of the different ways of pressure loading. As we known, in the process of mastication, the occlusal force borne by periodontium is discontinuous rather than unchangeable. Thus, it would be better to study the inflammation status and the bone homeostasis of the hPDLCs using a dynamic loading way to mimic the functional status. Therefore, in this present study, by applying cyclic air compression, we compared the expression differences of pro-inflammation, pro-osteoclastogenic and osteoblast-related cytokines between healthy and LPS-induced inflammatory hPDLCs under different dynamic loadings. This work may provide the foundation for clearing the reasonable force range of hPDLCs and give the reference for stress designing of the periodontitis abutment teeth in clinic.

Methods
Cell Culture and Periodontitis Induction Model hPDLCs were isolated from healthy premolars for orthodontic reasons among patients at age 18-30.
The primary cells were collected by scrapping the middle third of the roots, minced into pieces at about 1 mm 3, and subjected to 0.3% collagenase type I (Sigma, USA) at 37 °C with gently shaking for 30 min. After centrifugation, the precipitate was transferred to culture flasks (Corning, USA) with α-minimum essential medium (a-MEM) (Hyclone, USA) supplemented with 10% fetal bovine serum Application of dynamic cyclic stress hPDLCs were plated at a density of 5×105 cells/well in 6-well plates, and cultured in basal medium.
After treated with or without LPS, all the plates were applied to cyclic hydrostatic pressure of 0, 0-50, 0-90 and 0-150 kPa (0.1 Hz) using the dynamic hydrostatic pressure booster [13][14][15] Expression data were normalized to the amount of β-actin mRNA using the -ΔΔCt method. The primers for all the genes are listed in Table 1. Each reaction was performed in triplicate.

Western Blotting Analysis
The total protein of hPDLCs in each group was extracted by Ripa lysate (Boster, China) according to the manufacturer's instructions. And the protein concentration was determined using the bicinchoninic acid (BCA) protein assay kit (Thermo scientific, USA). About 20 μg of protein was separated with 10% and 15% SDS-PAGE gels. The isolated protein was transferred onto polyvinylidenedifluoride membrane, and then blocked with 5% non-fat dry milk in TBS containing 0.1% Tween for 1 h at room temperature. Then the membranes were trimmed to narrow strips based on the molecular weight of the target proteins according to the markers. These strips were probed with an antibody to a single target protein. The membranes were incubated with the appropriate primary antibodies overnight at 4 °C . After washing, the membranes were incubated with a secondary horseradish peroxidase (HRP)-coupled antibody and processed for an enhanced chemiluminescence detection using Immobilon HRP substrate (Millipore, USA). Signals were visualized and analyzed on a UVP Vision Works LS BioSpectrum (Aplegen, USA). The intensity of bands was quantified using IMAGEJ software (National Institute of Health, USA). The ratio of the intensity of the target protein to that of β-actin loading control was calculated to represent the expression level of the protein. Antibodies were as follows: anti-IL-1β (1:1000) (ab2105), anti-TNF-α

Establishment of periodontitis induction model in vitro
According to the results of the MTT assay, the proliferation of hPDLCs showed a significant reduction

Discussion
As known, the process of periodontitis begins with the endotoxin released by bacteria. Then the proinflammatory factors such as TNF-α and IL-1β are secreted by local periodontal ligament cells, which invades periodontal tissues and finally leads to the absorption and destruction of parodontium [16]. TNF-α and IL-1β were demonstrated to be the key factors in periodontits [17]. From another perspective, LPS, by which endotoxin exists on the walls of some bacteria, can lead to inflammatory reactions in multiple tissues such as genitourinary inflammation [18] and chronic fatigue syndrome [19]. It was reported that LPS can also stimulate the defensive cells in PDL to produce pro-inflammatory factors such as TNF-α and IL-1β, and then cause the destruction of PDL and alveolar bone [1,20]. However, the concentration of LPS applied on hPDLCs to mimic the periodontitis in vitro is still controversial. 0.1-10 μg/ml LPS were used on hPDLCs to induce the inflammatory state to observe the effects of IL-6 and MCP-1 in the previous study [21]. And it was also reported that 1.0 μg/ml of LPS could be used to establish the model of periodontitis and could contribute to the secretion of inflammatory cytokines in hPDLCs [22]. The miRNA expression patterns were investigated in the inflammatory hPDLCs induced by 0, 0.5, 1.0, 1.5 and 2.0 μg/ml of LPS [23]. What's more, the scholars also applied 0, 10, 20, 50, and 100 μg/ml of LPS on hPDLCs to make an inflammatory environment, and aimed to investigate the anti-inflammatory effect of a certain therapy [24]. So, we selected the concentrations of 0.1, 1.0, 10, 100 and 500 μg/ml to make clear the appropriate working dosage of LPS in the periodontitis induction of hPDLCs in vitro.
According to both MTT assay and real-time PCR analysis, it could be concluded that 10 μg/ml of LPS showed no effect on the proliferation of cells and promoted the inflammatory response of hPDLCs, which could be used to induce the model of periodontitis in vitro in the following study.
Mechanical stress is essential for the physiological function of a healthy periodontium. But excessive occlusal stress could cause the damage of periodontal tissue in vivo. Cyclic hydrostatic pressure has been applied on hPDLCs in vitro to mimic the physiological state of periodontium [25]. Previous studies had suggested that cyclic pressure higher than 150 kPa could significantly affect the morphology and function of hPDLCs and also inhibit the proliferation and differentiation of these cells [26]. What's more, it had been reported that the feasible pressure condition for hPDLCs should be 90 kPa for 60 min, under which the ALP activity of the cells would be promoted without affecting cell proliferation rates [25]. Therefore, in our study, 0-50 kPa, 0-90 kPa and 0-150 kPa were selected as the loading range to simulate the state of normal occlusion, critical occlusion and over occlusion in vivo. And the conclusions of this study were accompanied with the previous finding that over loading of pressure would significantly enhance the inflammation and osteoclast effects on LPS-induced inflammatory hPDLCs.
It is well known that PDLCs can be induced to differentiate into osteoblasts and pre-osteoclasts, and participate in the bone remodeling according to different mechanical stimulations. In some orthodontic studies, new bone formation was always found on the tension side while the osteoclasts on the pressure side were extremely active [27]. Studies showed PDLCs under cyclic stretch stress could express higher OCN, ALP and Runx2 [28]. While there were few researches about the compression-relative osteoblastic differentiation. It was reported that the expression of osteogenic factors in MC3T3 E1 cells increased significantly after the cyclic pressure applied within a certain range [29]. Therefore, we speculated that appropriate cyclic pressure may promote the osteogenic differentiation in hPDLCs. In the present study, we found an obvious up-regulated expression of ALP, COL-1 and RUNX2 in healthy hPDLCs under the dynamic cyclic pressure of 0-90 kPa, which suggested the osteogenic differentiated trend of hPDLCs in a certain range of stress. But there was no similar observation in the Western blotting analysis among these groups. The reason for this inconsistency may be related to the short experimental period and the lack of osteogenic induction medium during the experiment, which still warrants further investigation. In addition, both mRNA and protein level of the osteogenic markers showed no obvious difference among the corresponding LPS(+) groups. This result may indicate that the role of appropriate mechanical stimulation to induce osteogenic differentiation of hPDLCs would be interfered by inflammation.
Otherwise, we found that the expression pattern of pro-osteoclastic cytokines was similar to the proinflammatory cytokines among the groups in this present study. After dynamic cyclic pressure and LPS treatment, the osteoclastic and inflammatory effects on the hPDLCs were both aggravated. In our present study, both mRNA and protein expression level of RANKL increased significantly after the LPS plus 0-90 kPa or 0-150 kPa dynamic cyclic pressure treatment. RANKL is an important preosteoclastic marker, appearing to be both necessary and sufficient for the complete differentiation of osteoclast pre-cursor cells into mature osteoclasts. Previous studies showed that the expression of RANKL can be up-regulated under compressive force in PDLCs, which is an essential factor for osteoclastogenesis [30]. Hence the results of our study have defined the conditions of osteoclastogenesis, which are over loading on healthy hPDLCs and the critical loading on LPS-induced inflammatory hPDLCs in vitro.

Conclusion
Therefore, we conclude that dynamic cyclic pressure can promote the osteogenic differentiation of

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.