Impact of phosphorylation of heat shock protein 27 on the expression profile of periodontal ligament fibroblasts during mechanical strain

Purpose Orthodontic tooth movement is a complex process involving the remodeling of extracellular matrix and bone as well as inflammatory processes. During orthodontic treatment, sterile inflammation and mechanical loading favor the production of receptor activator of NF-κB ligand (RANKL). Simultaneously, expression of osteoprotegerin (OPG) is inhibited. This stimulates bone resorption on the pressure side. Recently, heat shock protein 27 (HSP27) was shown to be expressed in the periodontal ligament after force application and to interfere with inflammatory processes. Methods We investigated the effects of phosphorylated HSP27 on collagen synthesis (COL1A2 mRNA), inflammation (IL1B mRNA, IL6 mRNA, PTGS2 protein) and bone remodeling (RANKL protein, OPG protein) in human periodontal ligament fibroblasts (PDLF) without and with transfection of a plasmid mimicking permanent phosphorylation of HSP27 using real-time quantitative polymerase chain reaction (RT-qPCR), western blot and enzyme-linked immunosorbent assays (ELISAs). Furthermore, we investigated PDLF-induced osteoclastogenesis after compressive strain in a co-culture model with human macrophages. Results In particular, phosphorylated HSP27 increased gene expression of COL1A2 and protein expression of PTGS2, while IL6 mRNA levels were reduced. Furthermore, we observed an increasing effect on the RANKL/OPG ratio and osteoclastogenesis mediated by PDLF. Conclusion Phosphorylation of HSP27 may therefore be involved in the regulation of orthodontic tooth movement by impairment of the sterile inflammation response and osteoclastogenesis. Supplementary Information The online version of this article (10.1007/s00056-022-00391-w) contains supplementary material, which is available to authorized users.


Introduction
Numerous cell types and signal substances play a role in orthodontic tooth movement. Based on the force exerted, tensile and compressive zones can be distinguished from each other in the periodontal ligament. At tensile zones, there is increased bone formation, whereas alveolar bone is resorbed at compression zones [22]. The cells use different mechanisms to detect mechanical strain. Both the cytoskeleton [25] and membrane-bound molecules such as integrins [30,48], connexins [16,44] and ion channels [17,32] can be activated by pressure, triggering downstream effects. This includes the synthesis of two key proteins of bone resorption, receptor activator of NF-kB ligand (RANKL) and osteoprotegerin (OPG) [38]. Osteoclast-dependent bone resorption is the limiting factor of tooth movement [10]. During orthodontic tooth movement, RANKL is produced by periodontal ligament fibroblasts (PDLF), mesenchymal stem cells, lymphocytes, osteoblasts, and osteocytes and promotes the differentiation of osteoclasts from osteoclast progenitor cells by binding to the RANK receptor on osteoclast progenitor cells [14,33,[45][46][47]. OPG is a RANKL decoy receptor, as OPG prevents binding of RANKL to RANK [38]. Accompanying tooth movement, a sterile inflammation takes place in the periodontal ligament [10]. Pressure loading induces the synthesis and release of proinflammatory cytokines such as interleukin-1β (IL-1B) and IL6 in the early phase of tooth movement [14,26,39]. Thus, bone remodeling and a sterile inflammatory response characterize orthodontic tooth movement.
Among others, heat shock proteins (HSPs) can influence these processes. Several HSPs have already been linked with orthodontic tooth movement [3,23,24,37]. During bone formation at tension areas, HSPs act as molecular chaperones, which assist the maturation of bone morphogenetic proteins [24]. HSPs are involved in several processes, including protein transport, protein folding, and the assembly and disassembly of protein structures [50].
They support protein refolding and become active in misfolding, where they either refold the misfolded proteins or release them for degradation [50]. HSPs also play a significant role in immunomodulatory signaling pathways in response to stress and are therefore also considered important biomarkers [50]. Some HSPs are involved in the production of pro-inflammatory cytokines, such as IL1ß and IL6. HSPs are classified according to their molecular size. This results in seven families: HSP110, HSP100, HSP90, HSP70, HSP60, HSP40 and small HSPs, which have an approximate size of 15-30 kDa. HSP27 thus belongs to the small HSPs. Wolf et al. found that HSP70 in PDLF has an effect not only on osteoclast differentiation, but also on proliferation, wound healing, and apoptosis [42]. Recently, HSP27 was shown to inhibit the inflammatory response in intestinal epithelial cells via the NF-κB signaling pathway [49]. HSP27 is associated with protein folding, cell migration, cell metabolism, cell differentiation, cell growth, signal transduction and apoptosis [15]. HSP27 is expressed by most cells, although the amount expressed varies greatly depending on the cell type. Synthesis of HSP27 is induced by heat, stress, estrogens, and nerve injury [15]. Posttranslational modifications can strongly influence the functions of HSP27. It can act as a chaperone, stabilizing denatured or aggregated proteins and folding them back to their original form [12,19]. Phosphorylation, which is triggered by oxidative stress [28], abolishes oligomerization, resulting in increased chaperone activity of monomers [2,11]. The localization of HSP27 is not limited to the cytosol, but has also been detected in the nucleus [1]. In addition, HSP27 also has an anti-apoptotic effect [4,5]. This has a promoting effect on cancer progression and metastasis. In this regard, HSP27 overexpression not only leads to increased invasiveness, but also to resistance to chemotherapeutic treatments [8,27,29]. Despite years of research and numerous findings, the functions of HSP27 have not been fully elucidated.
Increased expression and phosphorylation of HSP27 was detected in the periodontal ligament already 10 min after force application in a mouse model [23,37]. Therefore, our aim was to verify whether HSP27 exhibits an anti-inflammatory effect also in PDLF and thus could be a potential target to influence orthodontic treatments. Furthermore, the functional diversity of heat shock proteins suggests other previously unknown roles in cell homeostasis. Therefore, a possible influence on collagen synthesis as well as mediators of bone formation and degradation was also investigated.

Transfection experiments
The plasmid pHSP27 + (MS16-6, MWG eurofins, Ebersberg, Germany) contains human HSP27 gene with three times aspartate instead of serine. Due to the negative charge of aspartate, permanent phosphorylation occurs. An empty plasmid served as a control plasmid. PDLF were transfected with the corresponding plasmids. For this purpose, 50,000 cells were initially seeded per well of a 24-well plate (662 160, Greiner Bio-One, Frickenhausen, Germany). On the next day, 0.5 μg plasmid DNA (either control or pHSP27 + ) was mixed with 100 μl serum-free RPMI medium, and 0.5 μl Turbo-Fect™ (R0531, Thermo Fisher Scientific, Waltham, MA, USA) and incubated for 20 min at room temperature. Subsequently, transfection of the PDLF was performed by adding 100 μl of the respective transfection mixture to the cells. After an incubation period of another 24 h, the transfected cells were exposed to 6 g/cm 2 for 6 h (supplemental Fig. 1b). Cell number and lactate dehydrogenase (LDH) release were determined and RNA and protein were isolated for further analysis.

Determination of cell number using the crystal violet assay
The supernatant was removed and adherent cells were washed with 500 μl PBS (14190-094, Gibco™, Carlsbad, CA, USA), then 200 μl crystal violet solution consisting of 2.5 g crystal violet (T123.3, Carl Roth, Karlsruhe, Germany); 0.85 g NaCl (3957, Carl Roth, Karlsruhe, Germany); 20 ml 37% formaldehyde (M4003, Sigma-Aldrich, St. Louis, MO, USA), 150 ml ethanol (32205, Sigma-Aldrich, St. Louis, MO, USA) were added in a total volume of 500 ml per well of a 24-well plate. After 15 min at 37°C, the plate was washed three times with tap water and lightly tapped out onto a paper towel. The wells were dried at 37°C for 45 min. After this time, 300 μl of a 33% acetic acid solution (3738.1, Carl Roth, Karlsruhe, Germany) were added per well and measured at a wavelength of 595 nm using an ELISA reader (Multiscan Go, Thermo Fisher Scientific, Waltham, MA, USA).

Determination of cytotoxicity using the lactate dehydrogenase assay
Lactate dehydrogenase (LDH) release was determined with LDH assay (04744926001; Roche, Basel, Switzerland) according to manufacturer's instruction. After an incubation period of 30 min in the dark, stop solution was added to each well. Thereafter, measurements were made at wavelengths of 490 and 690 nm (reference) using Multiscan Go (Thermo Fisher Scientific, Waltham, MA, USA).

RNA isolation and cDNA synthesis
RNA Solv ® Reagent (250 µl, R6830-01, VWR international, Radnor, PA, USA) was added directly to the cell culture plates. After a short incubation on ice to destroy the cells, the supernatant was transferred to microreaction tubes followed by the addition of 100 μl chloroform. Samples were mixed for 30 s and then incubated on ice for 15 min. This was followed by centrifugation at 13,000 rpm at 4°C for 15 min. The clear aqueous phase was transferred to a new tube with 500 μl cold isopropanol (20,842,330; VWR international, Radnor, PA, USA). After inverting, the samples were stored at -80°C overnight. The follow- Table 1 Alphabetical list of reference and target gene primers used for real-time quantitative polymerase chain reaction (RT-qPCR) Tab. 1 Alphabetische Liste der für die RT-qPCR ("real-time quantitative polymerase chain reaction") verwendeten Referenz-und Zielgen-Primer

Gene
Gene Interleukin

Quantitative real-time polymerase chain reaction (RT-qPCR)
The aim of the polymerase chain reaction is to exponentially amplify the synthesized cDNA. For this purpose, 1.5 μl cDNA of each sample was mixed with 8.5 μl of the previously prepared primer mix in duplicate into a 96-well plate (712282, Biozym Scientific, Hessisch Oldendorf, Germany), covered with an adhesive optical film (712350, Biozym Scientific, Hessisch Oldendorf, Germany) and briefly centrifuged. The primer mix consisted of 0.25 μl each of forward and reverse primer (

Statistical analysis
Statistical analysis was performed using the program GraphPadPrism 9.2. Prior to statistical analysis, all absolute data values were divided by the respective arithmetic mean of the control group without mechanical loading to obtain normalized data values relative to these controls. Data are presented as single point diagrams with each symbol representing one data point. In addition, the horizontal lines in the graphs show the mean values and the vertical lines the standard error of the mean. Normal distribution of the data was examined with the Shapiro-Wilk test and homogeneity of the groups was determined with the Brown-Forsythe test. Depending on the normal dis-tribution and homogeneity of the data, the following tests were performed: analysis of variance (ANOVA) followed by Holm-Šídák post hoc tests, Welch-corrected ANOVA followed by Dunnett's T3 post hoc tests, with two groups to be compared: Mann-Whitney U test. Differences were considered statistically significant at P < 0.05.

HSP27 phosphorylation in PDLF after mechanical strain with different force magnitudes
First, we wanted to investigate the optimal timepoints and force magnitudes to study the role of HSP27 phosphorylation on the expression profile of PDLF. With a force size of 2 g/cm 2 , hardly any signal could be obtained for pHSP27 (Fig. 1a). With 4 g/cm 2 , increased HSP27 phosphorylation was evident after 6 h. With 6 g/cm 2 this was already observed after 2 h. Again, a peak of HSP27 phosphorylation was evident after 6 h (Fig. 1a). The densitometric evaluation of the HPS27 phosphorylation after exposure to different force magnitudes for 6 h showed a significant difference only at 6 g/cm 2 compared to the untreated controls (P = 0.013; Fig. 1b).

Effects of high force magnitudes on PDLF
Most experiments dealing with PDLF and compressive force were conducted with 2 g/cm 2 for up to 48 h. After identifying a force magnitude of 6 g/cm 2 for up to 6h as optimal for HSP27 phosphorylation, we first wanted to deter-  Fig. 2b) were affected by the used conditions. Gene expression of Collagen-1-α-2 (COL1A2) was significantly increased after compressive force treatment (P = 0.039; Fig. 2c) indicating increased collagen synthesis. Expression of inflammatory genes like Interleukin-1β (IL1B; P < 0.001; Fig. 2d) and IL6 (P < 0.001; Fig. 2e) was upregulated within 6 h of compression with 6 g/cm 2 . Protein expression of prostaglandin endoperoxide synthase-2 (PTGS2) was also upregulated (P < 0.001; Fig. 2f). We detected no effect on receptor activator of NF-κB ligand (RANKL) protein expression (P = 0.744; Fig. 2g), while osteoprotegerin (OPG) protein secretion was significantly downregulated after 6 h compression with 6 g/cm 2 (P < 0.001; Fig. 2h) resulting in more TRAP + cells in a coculture model using THP1 macrophages (P = 0.002; Fig. 2i). The examined force magnitude and application time is therefore suitable to study effects on inflammatory and bone remodeling mediators.

Effects of phosphorylation of HPS27 on cell vitality of PDLF
To investigate the influence of phosphorylation of HSP27 on the sterile inflammatory response occurring during orthodontic tooth movement in the periodontal ligament, PDLF were transfected with pHSP27 + , which contains the human HSP27 gene with three times aspartate instead of serine, corresponding to permanent phosphorylation. Control cells were transfected with an empty vector. Accordingly, we detected increased HSP27 phosphorylation after compression and transfection with the pHSP27 + (P < 0.001; Fig. 3a). Compression failed to further increase pHSP27 (P = 0.716; Fig. 3a). Surprisingly, transfection with pHSP27 + resulted in reduced cell numbers without (P = 0.011) and with compression (P = 0.003; Fig. 3b).
In line with that, we detected increased LHD release after transfection (without pressure: P = 0.016; pressure: P < 0.001; Fig. 3c) indicating a cytotoxic effect of the plasmid pHSP27 + .

Effects of phosphorylation of HPS27 on the expression of collagen and inflammatory mediators in PDLF
Extracellular matrix remodeling occurs during orthodontic treatment. Accordingly, we detected increased COL1A2 gene expression after transfection with the control and pHSP27 + plasmid after mechanical strain (P < 0.001; Fig. 4a). However, pHSP27 + transfection reduced COL1A2 gene expression without (P < 0.001) and with pressure application (P = 0.003) indicating a regulatory effect of HSP27 phosphorylation on COL1A2 gene expression.
As the sterile inflammatory process is crucial for orthodontic tooth movement, we investigated the effects of phosphorylation of HSP27 on the expression of inflammatory genes and proteins. Mechanical strain increased IL1B gene expression after transfection with an empty vector (control; P = 0.004) and with pHSP27 + (P = 0.039; Fig. 4b). Transfection with the permanently phosphorylated HSP27 plasmid had no effect on IL1B gene expression without (P = 0.836) and with mechanical strain (P = 0.986; Fig. 4b). As expected, IL6 mRNA levels were elevated after compression in cells transfected with the empty vector (P = 0.001; Fig. 4c). Transfection with pHSP27 + reduced IL6 gene expression without and with mechanical strain (P < 0.001; Fig. 4c). PTGS2 protein expression was increased with mechanical strain after transfection with the empty vector (P = 0.009; Fig. 4d). Transfection with pHSP27 + led to an increase of PTGS2 protein expression without pressure application (P < 0.001; Fig. 4d). There was no significant difference due to pHSP27 + transfection after mechanical strain (P = 0.448).

Effects of constitutive phosphorylation of HSP27 on bone remodeling mediated by PDLF
Orthodontic tooth movement depends on bone resorption facilitated by osteoclasts. Osteoclastogenesis itself is crucially regulated by RANKL and its decoy receptor OPG. With the setup of 6 g/cm 2 for 6 h, we detected no increased RANKL protein expression after transfection with an empty control vector and mechanical loading (P = 0.341; Fig. 5a). Transfection with pHSP27 + increased RANKL expression under control (P = 0.015) and pressure conditions (P = 0.021; Fig. 5a). Compressive force of 6 g/cm 2 for 6 h decreased OPG secretion, when cells were transfected with the control plasmid (P = 0.004; Fig. 5b). Treatment with pHSP27 + significantly decreased OPG secretion without mechanical loading (P = 0.013; Fig. 5b). Accordingly, we

Discussion
Muraoka et al. reported increased HSP27 expression and phosphorylation in the periodontal ligament of mice after force application [23]. This effect was already observed 10 min after the load application and reached a peak after nine hours. This was comparable to our results as we observed a maximum of HSP27 phosphorylation after six hours. In contrast to the results of Muraoka et al., we observed no changes in HSP27 expression. Phosphorylation of HSP27 affected collagen, IL6 and PTGS2 expression as well as the RANKL/OPG ratio resulting in the induction of an increased number of TRAP + cells after compressive force treatment.
Remodeling of the extracellular matrix is crucial for orthodontic tooth movement. We observed decreased COL1A2 mRNA levels after transfection with pHSP27 + . Hirano et al. and Deng et al. reported that HSP27 overexpression increased collagen synthesis in fibroblasts [9] or in alveolar type II epithelial cells [6]. Also, HSP70 and HSP47 were associated with collagen secretion in PDLF during mechanical loading [24].
IL6 is the classic proinflammatory cytokine associated with bone resorption [43]. Constitutively phosphorylated HSP27 led to a decrease in IL6 gene expression by PDLF. This observation is consistent with the results of Zhang et al. in intestinal epithelial cells [49]. Sur et al. also identified an involvement of HSP27 in the NF-κB signaling pathway with Dunnett's T3 multiple comparisons tests; *P < 0.05; **P < 0.01; ***P < 0.001 Abb. 5 RANKL("receptor activator of NF-κB ligand")-Proteinexpression (a) und OPG(Osteoprotegerin)-Sekretion (b) in PDLF (parodontale Ligamentfibroblasten) ohne und mit Kompression sowie Bestimmung der TRAP+ Zellen (c) in einem Kokulturmodell mit THP1-Makrophagen. Statistik: Welch-ANOVA ("analysis of variance") mit Dunnett-T3 Post-hoc-Test; *p < 0,05; **p < 0,01; ***p < 0,001 in keratinocytes. They described an increased production of inflammatory mediators as a consequence downregulation of HSP27 [36]. Accordingly, a decrease in IL6 expression could be due to increased levels of phosphorylated HSP27. The increased RANKL/OPG ratio seems to be in apparent contradiction to the lowered expression of IL6. Marciniak et al. (2019) investigated the role of heat shock protein 70 (HSP70) during mechanical loading by inhibiting HSP70 [21]. They used a compressive force of 2 g/cm 2 and observed increased IL6 and IL8 gene expression after compression. Inhibition of HSP70 and simultaneous pressure application resulted in an even higher gene expression of IL6 and IL8, suggesting a protective function of HSP70 [21]. Wolf et al. investigated the effect of heat pretreatment on PDLF [41]. PDLF were subjected to an elevated temperature of 43°C before force application. This treatment significantly reduced protein secretion of IL6 and IL8. Therefore, heat shock proteins are considered a promising therapeutic option to minimize undesirable side effects during orthodontic treatments [41]. For the experiments, PDLF were transfected with HSP27 + , which is characterized by having aspartate instead of the three serine residues. Aspartate, like the phosphate ion, carries a negative charge. Thus, this plasmid corresponds to permanent phosphorylation and acquires phospho-mimetic properties [35]. In its native form, HSP27 is also present in a dephosphorylated state and can only form large oligomers and perform its chaperone function in this way [15]. Phosphorylation of HSP27 causes oligomers to dissociate [34]. These results may imply that there is some protective function against the generation of proinflammatory factors when HSP27 is not phosphorylated and thus is in the form of oligomers. This allows HSP27 to perform its chaperone function.
Compression with 6 g/cm 2 for 6 h led to an increase in the RANKL/OPG ratio, which was due to a reduction in OPG secretion and increased PTSG2 protein expression. PTGS2 is critically involved in prostaglandin synthesis [7,40]. PGE2 has been shown to promote RANKL production and inhibit OPG expression in osteoblasts [18]. Since inflammatory mediators promote RANKL production and inhibit OPG synthesis, one would expect a reduction in this ratio. However, as described at the beginning, RANKL and OPG synthesis depend on numerous cellular factors in addition to inflammatory mediators. These include the cytoskeleton [25], integrins [30,48], connexins [16,44], and ion channels [17,32]. Thus, it could be that HSP27 interferes with another, as yet unknown, signaling pathway, thereby influencing the RANKL/OPG ratio in favor of RANKL. However, further research is needed to test this conjecture. Overall, HSP27 could be a promising target in orthodontic treatments.

Conclusion
Phosphorylation of HSP27 could accelerate orthodontic tooth movement by increasing the RANKL/OPG ratio, which leads to increased differentiation of osteoclasts and thus more bone resorption. In addition, painful inflammatory responses could possibly be attenuated.

Supplementary Information
The online version of this article (https:// doi.org/10.1007/s00056-022-00391-w) contains supplementary material, which is available to authorized users.
Funding This study received no funding.
Funding Open Access funding enabled and organized by Projekt DEAL.

Declarations
Conflict of interest A. Schröder, K. Wagner, F. Cieplik, G. Spanier, P. Proff and C. Kirschneck report no financial or other conflict of interest relevant to this article, which is the intellectual property of the authors. Furthermore, no part of this article has been published before or is considered for publication elsewhere.
Ethical standards All experiments were performed in accordance with the relevant guidelines and regulations. Approval for the collection and use of PDLF was obtained from the ethics committee of the University of Regensburg, Germany (approval number 12-170-0150). We obtained informed consent from all patients or their legal guardians.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/.