Introduction

Bone homeostasis is a complex process controlled by both intrinsic and extrinsic mechanisms1. A striking example of the importance of extrinsic mechanism can be seen in astronauts who develop progressive osteopenia under conditions of weightlessness2. These mechanical stimuli (or lack thereof) influence the differentiation and activity of bone-resorbing osteoclasts and bone-forming osteoblasts3,4. In the past, it has been shown that osteocytes, as terminally differentiated osteoblasts embedded in the bone matrix, are crucial for the perception of mechanical stimuli. The osteocytes can sense the mechanical stimuli and generate signals that induce osteoblast formation5,6,7.

Osteocytes also appear to play a crucial role in bone remodeling processes during orthodontic tooth movement (OTM), which is an illustrative example of mechanically induced bone remodeling8. Here, extrinsic forces act on the tooth, forming pressure and tension zones in the cells of the periodontal ligament (PDL) and the surrounding bone. In the pressure zones, the compression of the periodontium results in inflammatory-mediated osteoclast formation and consequent resorption of the adjacent alveolar bone. In the tension zone, the mechanical stimulus results in increased osteoblast formation, which builds up the adjacent bone.

It was stated that the remodeling of the alveolar bone during orthodontic tooth movement is highly dependent on the RANKL expression of osteocytes9. Additionally, recent studies have shown that orthodontic tooth movement leads to apoptosis of osteocytes as a first step and subsequent activation of osteoclasts10. Another in vivo study showed how the ablation of osteocytes led to a significant decrease in orthodontic tooth movement in mice11. However, to what extent osteocytes sense the mechanical stimuli during orthodontic tooth movement and if this stimulus is necessary for bone remodeling is still not fully understood.

The group of mechanically activated ion channels, named Piezo channels, was first discovered in mammals by Coste et al.12. Previous studies have shown that these mechanosensitive ion channels Piezo1 and Piezo2 are expressed in different bone cells, such as osteoblasts, osteocytes, and osteoclasts, and are crucial for sensing mechanical stimuli and regulating the differentiation and activity of these cells13,14,15,16,17,18. Despite this research, the exact mechanisms by which Piezo1 and Piezo2 regulate bone remodeling are still not entirely known.

For example, in mouse models, cell-specific deletion of Piezo1 in osteoblasts and osteocytes led to decreased bone density and increased bone porosity16,17. In addition to reduced bone formation, these mice also showed increased osteoclast differentiation factor RANKL and increased bone resorption. On the other hand, Piezo2 seems to play a minor role in skeletal remodeling since mice with a cell-specific deficiency of Piezo2 in osteoblast progenitor cells show only a negligible skeletal phenotype17.

Recently, the extent to which the expression of Piezo1 and Piezo2 in specific cell types regulates orthodontic tooth movement and, thus, bone remodeling has been questioned19. Indeed, a study showed that inhibition of Piezo1 using the mechanosensitive ion channel inhibitor GsMTx4 reduced osteoclastogenesis and subsequently tooth movement in rats. However, due to the systemic application, no specific cell type that was affected by the inhibitor could be pointed out20. Furthermore, GsMTX4 is not specific for PIEZO proteins. It also inhibits other mechanosensitive ion channel such as transient receptor potential canonical (TRPC1) and TRPC621. Therefore, in this study, we were particularly interested in the effect of a cell-specific deletion of Piezo1 in osteocytes, which are essential for mechanically induced bone remodeling, using a tooth movement model in mice.

Results

Mechanosensing Ion Channels PIEZO1 and PIEZO2 are expressed in different dentoalveolar hard tissues

To investigate how the mechanosensitive ion channels Piezo1 and Piezo2 can impact dentoalveolar hard tissues during orthodontic tooth movement, we first explored the specific expression pattern of PIEZO1 and PIEZO2 using antibody staining on paraffin sections of control animals (Fig. 1a). Here the staining shows the expression of PIEZO1 in cells like odontoblasts, cementocytes, osteocytes, and osteoblasts, whereas the expression of PIEZO2 was mainly observed in odontoblasts and cementoblasts (Fig. 1b–e). Furthermore, knockout of Piezo2 in Dmp1 expressing cells does not lead to a significant effect on general skull morphology and bone volume17 shown in Supplementary Fig. 1.

Figure 1
figure 1

The expression of mechanosensitive ion channels PIEZO1 and PIEZO2 in dentoalveolar hard tissues. (a) Overview of an antibody staining for PIEZO1 and PIEZO 2 on a paraffine section of a control mouse. Scale bar = 500 µm. Zoom ins (b–e) show the expression of PIEZO1 and PIEZO2 in cells of the dentoalveolar tissues like odontoblasts (OD) and cementoblasts (CB), and cementocytes (CC) (b,c). Osteocytes (OC) and Osteoblasts (OB) show a higher expression of PIEZO1, whereas the expression of PIEZO2 is not evident (d,e). Scale bars = 20 µm.

Full size image

We, therefore, focused our further investigations on Piezo1 and used a Cre/lox Piezo1 mouse model and the Dmp1 promoter-driven expression of Cre recombinase to inactivate Piezo1 expression in mature osteoblasts and osteocytes (Fig. 2a). We confirmed this specific expression of Dmp1 using a Dmp1Cherry;DsppCerulean mouse line showing the expression of Dmp1Cherry transgenes in these particular dentoalveolar cell types. The transgenes were highly expressed not only in osteocytes but also osteoblasts within the alveolar bone, odontoblasts and cementocytes (Fig. 2b–f).

Figure 2
figure 2

Dmp1 expressing cells can be found in dental hard tissues. (a) Piezo1floxed/floxed mice and Dmp1cre mice were mated using the loxP system to generate a tissue specific lack of Piezo1 in Dmp1 expressing cells like osteocytes, osteoblasts, odontoblasts, cementoblasts and cementocytes (Created with BioRender.com) (b) shows a sagittal section through the M1 of a DSPPCerulean;Dmp1Cherry Reporter Mouse using fluorescence imaging. scale bar = 250 µm. M1 = first Molar, dr distal root, mr mesial root, AB alveolar bone, PDL periodontal ligament. Details (c-f) show the expression of Dmp1 in osteocytes (OC), osteoblasts (OB), odontoblasts (OD), cementoblasts (CB), and cementocytes (CC). d dentin, p pulp, AB alveolar bone, PDL periodontal ligament. scale bars = 50 µm.

Full size image

Deletion of Piezo1 in Dmp1 expressing cells affects the alveolar bone volume

Since mechanosensation is an important factor in bone homeostasis, we assumed that a loss of Piezo1 would lead to a change in the skeletal morphology of the skull. To analyze this, we performed micro-computed tomography. In the 3-dimensional reconstructions of these scans, we observed that the inactivation of Piezo1 in osteoblasts and osteocytes did not affect the morphology of the skull and the jaws (Fig. 3). Measurements of the skull length and the length of the mandible did not show differences between Piezo1 and control littermates (Fig. 3a,b,d,e). However, we observed significant bone loss in the craniofacial skeleton of Piezo1floxed/floxed;Dmp1cre mice. The bone volume of the mandible and maxilla of Piezo1floxed/floxed;Dmp1cre animals was significantly reduced compared to control littermates (Fig. 3a,b,f,h). In addition, we observed a loss of the vertical bone height between the tooth-bearing alveolar bones, resulting in a smaller root area covered by bone (Fig. 3c,g).

Figure 3
figure 3

Deletion of Piezo1 in Dmp1 expressing cells leads to decreased alveolar bone volume in mice. (a,b) 3-dimensional reconstruction of Micro-Computed Tomography scans of Control (upper panels) and Piezo1floxed/floxed;Dmp1cre (lower panels) mice. The Wall Thickness analysis of the skulls and mandibles shows, based on relative values, that the Bone thickness is decreased in Piezo1floxed/floxed;Dmp1cre animals. This effect is also visible in the quantification of the Mandible Bone Volume (f) (n = 7). The skull and the mandible length do not differ between Control and Piezo1floxed/floxed;Dmp1cre animals (d,e) (n = 4). (c) shows the bone loss in the maxillary bone. The vertical bone height (g) and also the bone volume in the maxilla (h) are significantly decreased compared to the values of control littermates (n = 4). Values are means ± SD. (*P < 0.05, ***P < 0.005). (Scale bars a = 5 mm, b = 2,5 mm, c = 1 mm).

Full size image

To further investigate these bone changes, we next performed histological analysis of the alveolar bone (Fig. 4). Morphometric analysis of osteocyte numbers on toluidine-blue stained sections showed no significant changes between Piezo1floxed/floxed;Dmp1cre mice and control littermates. In contrast, TRAP staining revealed a significantly increased number of osteoclasts in Piezo1floxed/floxed;Dmp1cre mice (Fig. 4a,b,d,e). The number of osteoblasts per bone surface, on the other hand, was not affected (Fig. 4c,f).

Figure 4
figure 4

Deletion of Piezo1 in Dmp1 expressing cells leads to an increased osteoclast activity in the interradicular bone. (a) shows with the help of a toluidine-blue staining the osteocytes (OT) of the maxillary interradicular bone. Scale bars = 20 µm. They do not differ in number, nor in morphology, comparing the sections of control and Piezo1floxed/floxed;Dmp1cre animals. N.Ot/B.Ar = Number of Osteocytes per Bone Area (d). The visualization of osteoclasts (OC) in the interradicular bone of control (left panel) and Piezo1floxed/floxed;Dmp1cre animals (right panel) (b) was performed with a TRAP-staining. Scale bars = 50 µm. The black arrows show the osteoclasts, lining the bone surface. Quantification of them shows a basal increased osteoclast activity in Piezo1floxed/floxed;Dmp1cre animals. N.Oc/BS = Number of Osteoclasts per Bone Surface (e). (c) shows the osteoblasts that are lining the bone surface of the interradicular bone. The number of osteoblasts in Piezo1floxed/floxed;Dmp1cre animals is not affected compared to their control littermates. N.Ob/BS = Number of Osteoblasts per Bone Surface (f). Scale bars = 20 µm. Values are means ± SD. (*P < 0.05).

Full size image

The bone phenotype in Piezo1 floxed/floxed ;Dmp1 cre mice does not lead to altered tooth movement

Having demonstrated the bone phenotype of Piezo1floxed/floxed;Dmp1cre animals in the absence of an external force, we next asked how orthodontic tooth movement is affected by the loss of Piezo1 in osteocytes and late osteoblasts. To perform orthodontic tooth movement in Piezo1floxed/floxed;Dmp1cre mice and control littermates, NiTi-tension springs with a mesial directed force of 0.3N were placed between the upper M1 and the incisors with dental resin for 12 days (Fig. 5a). Subsequently, the extent of tooth movement was determined from the intercoronal distance between the upper M1 and M2 in cross sections of micro computed tomography (µCT) scans (Fig. 5b). It was surprising to find that when the distance was quantified, there were no significant differences between the two groups in terms of the amount of tooth movement (Fig. 5c). To ensure that the tooth movement is not influenced by migration of the second and third molar, we performed additional analysis on these teeth with no significant difference between the groups (Supplementary Fig. 2).

Figure 5
figure 5

The amount of tooth movement in Piezo1floxed/floxed;Dmp1cre mice is not affected. (a) Illustration of the OTM model in mice created with BioRender.com. A NiTi tension spring was attached with dental composite for 12 days between the M1 and the incisors with a force of 0.3N. The black arrow indicates the mesial direction of the force. The mice were sacrified after 12d of tooth movement. (b) µ-CT cross sections through M1-M3 showing the approximal gap between M1 and M2 after orthodontic tooth movement. (c) Quantification of the intercoronal distance between M1 and M2 after 12d of OTM showing no significant difference in control and Piezo1floxed/floxed;Dmp1cre animals. (n = 6) *P < 0.05. (d) TRAP (upper panels) and toluidine blue staining (lower panels), showing the osteoclasts and the PDL fibroblasts in the pressure zone of the distal root of the M1. In the OTM- group, there is no osteoclast activity visible and the fibroblasts show a straight arrangement, as well as an intact insertion in the alveolar bone (AB). In the OTM + group, the pressure leads to a significant increase in osteoclast activity. dr distal root (f). Furthermore, it leads to a compression of the PDL fibroblasts with subsequent hyalinization of the PDL (black asterisk). (e) is showing the TRAP staining and toluidine blue staining for sections of Piezo1floxed/floxed;Dmp1cre mice. In the OTM- group the osteoclast activity in the PDL is slightly increased, but there is no significance compared to the WT group (f) (Quantification of TRAP staining, n = 4–5). In addition, the force does not lead to a significant increase in osteoclast activity, as in the control group. Regarding the arrangement of the PDL and the formation of areas of hyalinization there is no difference visible between both groups (g) (n = 3). Scale bars = 100 µm. Values are means ± SD. (***P < 0.001).

Full size image

To investigate how the force applied to the molar affected the bone remodeling processes in both groups, we analyzed the number of osteoclasts using TRAP staining in the pressure zones of the mesial and distal root with and without tooth movement (Fig. 5d,e). We counted the osteoclasts covering the bone surface near the root and the osteoclasts directly in the periodontal ligament (PDL). In the pressure zones, tooth movement leads to a significant increase in osteoclast activity in the control group, whereas the number of osteoclasts was not increased in the Piezo1floxed/floxed;Dmp1cre group (Fig. 5f).

Next, we investigated the extent of hyalinization (a common side effect in dentoalveolar soft tissue as a reaction to heavy forces) using toluidine blue staining, focusing mainly on the area around the distal root (Fig. 5d,e and Supplementary Fig. 3). Hyalinization was evident in both control and Piezo1floxed/floxed;Dmp1cre animals, and the quantification of the area shows that it does not differ between the two groups (Fig. 5g). Taken together, these results suggest that Piezo1-mediated mechanotransduction in osteocytes plays a minor role in orthodontic tooth movement.

Discussion

Recently, the mechanosensitive ion channels Piezo1 and Piezo2 have received much attention in the scientific world, which is also due, among other things, to the awarding of the Nobel Prize in physiology and medicine in the year 2021 to David Julius and Ardem Patapoutian “for their discoveries of receptors for temperature and touch”12,22. Both ion channels can be found in a variety of different cell types, but despite nearly identical structures, they seem to have different main functions. For instance, it has been shown that the ion channel Piezo1 plays an essential role in the mechanosensing of osteoblasts at various stages of differentiation and in the communication between osteoblasts and osteoclasts. Piezo2 on the other hand, seems to play an important role in mechanosensation when it comes to itch, touch and mechanical pain23,24,25.

In this manuscript, we used an orthodontic tooth movement model in mice lacking Piezo1 in Dmp1 expressing cells to investigate whether Piezo1 plays a role in the mechanosensation of osteocytes during orthodontic tooth movement.

Here we were able to confirm the specific expression of PIEZO1 and PIEZO2 in the dentoalveolar cell types of untreated control animals. PIEZO1 expression was evident in odontoblasts, osteoblasts, and osteocytes. Interestingly, the expression of PIEZO2 was mainly observed in odontoblasts and cementoblasts but not in osteoblasts and osteocytes. The expression pattern of PIEZO1 as well as the low expression of PIEZO2 in bone cells is consistent with previous studies showing that Piezo2 plays a minor role in mechanically induced bone remodeling in vivo17,26.

Interestingly, when we knocked out Piezo1 in these Dmp1-expressing cells, we observed a decrease in bone volume and bone thickness, while the size and shape of the bones did not differ between both groups. Additionally, we also did not observe a difference in osteoblast and osteocyte numbers. Interestingly, mice lacking Piezo2 in Dmp1-expressing cells showed no effect on overall morphology and bone volume (Supplementary Fig. 1). These results are consistent with previously published data investigating the cell-specific deletion of Piezo1 and Piezo2 in bone-forming cells to study the bone phenotypes of the mice17. The studies indicated that in mice, the elimination of Piezo1 in cells of the osteoblast lineage led to a reduction in bone volume and impaired bone formation. Cell-specific knockout of Piezo1 in osteoblast progenitors resulted in severe osteoporotic phenotypes13,17,18, whereas cell-specific knockout of Piezo2 in osteoblast lineage cells did not result in an osteoporotic phenotype17. Piezo2 appears to affect skeletal integrity via expression in proprioceptive neurons and subsequent regulation of skeletal muscle functions rather than having a direct effect in osteoblast lineage cells27.

The osteopenic phenotype that occurred in the mice lacking Piezo1 in bone-forming cells can be linked to reduced bone formation and increased bone resorption due to an increase in the osteoclast differentiation factor RANKL16. This is consistent with our finding of increased basal osteoclast activity in the alveolar bone of Piezo1floxed/floxed;Dmp1cre animals. Another possible mechanism, leading to the decreased bone formation in Piezo1floxed/floxed;Dmp1cre animals was investigated by Li et al.13. Here, Wnt1 expression (a known bone anabolic ligand28) is significantly decreased in Piezo1floxed/floxed;Dmp1cre mice.

The osteopenic phenotype of mice lacking Piezo1 in osteoblasts and osteocytes further suggests that they have a reduced ability to respond to mechanical stimulation, which has been confirmed in different anabolic tibia and ulna loading models. In a previous study, it was shown that the osteoanabolic response of Piezo1floxed/floxed;Dmp1cre animals was reduced, and bone was less responsive to mechanical signals than controls17. In our tooth movement model, we see a similar effect, as the loss of Piezo1 in osteoblasts and osteocytes did not lead to an increase in tooth movement, despite a basal increased osteoclast activity and the subsequent decrease in bone volume. This effect could be explained by a dysfunctional osteoblast-osteoclast interaction, as was described by Wang et al.16. However, it is of interest, that the increase in osteoclastogenesis was primary detected in the alveolar bone. We did not observe an increase in osteoclastogenesis in the periodontal bone, which is continuously exposed to masticatory forces. This could indicate that continuous force can reduce the osteoclastogenesis phenotype of Piezo1floxed/floxed;Dmp1cre mice. The authors utilized Piezo1floxed/floxed-Prx1cre mice and control mice in a tail suspension model for their research. Their findings indicated that while the suspension treatment resulted in a rise in osteoclast activity in control animals, it did not surpass the basal increase in the number of osteoclasts in the Prx1 mice. By that, they underline the importance of Piezo1 for the modulation of the bone-resorbing activity of osteoclasts16. Further evidence for the negligible role of Piezo1 in osteocytes during OTM could be the areas of hyalinisation, which are the same in both groups. Our initial hypothesis was that a loss of mechanosensing ability in the osteocytes of Piezo1floxed/floxed;Dmp1cre animals would lead to an altered response and force distribution in the PDL. However, our findings showed that this was not the case; the side effects of tooth movement were just as pronounced in Piezo1floxed/floxed;Dmp1cre mice as in control animals.

Our results suggest that mechanically induced bone remodeling using orthodontic forces is not dependent on Piezo1 expression in hard tissue. However, since we performed our tooth movement model for a period of 12 days, it is reasonable to assume that a more extended period could result in a significant difference in bone remodeling since long-term effects of Piezo1 expression on bone remodeling may not be fully captured. Further research with more extended observation periods is needed to fully understand the role of Piezo1 in bone remodeling in hard tissue.

Furthermore, it is important to note that the mechanosensitivity of osteocytes is not solely dependent on mechanosensing ion channels, but also on other structures such as primary cilia, integrins, and G-proteins. These structures play a crucial role in enabling the osteocytes to respond to mechanical forces and thus could influence tooth movement29. Furthermore, targeting osteoblast precursor using a cell-specific knockout of Piezo1 in Runx2 expressing cells might also lead to a more pronounced impact on tooth mobility. However, we did not use this mouse line since a previous study showed that Piezo1floxed/floxed;Runx2cre mice were prone to rib fractures starting from day P5 due to their severe osteoporotic phenotype17.

Additionally, the involvement of mechanosensing ion channels within the periodontal soft tissue and its role during tooth movement might be more critical than previously anticipated30,31,32,33. Here, the role of Piezo1 in fibroblasts of the periodontal ligament has been investigated in previous studies. Jin et al. stated in their in vitro study that pressure on PDL fibroblasts leads to an increased expression of Piezo134. The same effect was later confirmed in vivo using a tooth movement model in control mice35,36. In addition, Jiang et al. confirmed an increase of Piezo1 expression in the tension zone of the PDL as a reaction to mechanical force and furthermore showed that Piezo1 was important for the expression of osteogenic markers20. As part of the research, the investigators employed the mechanosensitive ion channel inhibitor GsMTx4 to explore its impact on tooth movement37. In vitro, periodontal cells treated with GsMTx4 and subsequently exposed to a static pressure showed a reduction in RANKL expression and a concomitant decrease in osteoclastogenesis34,35,36. In vivo a systemical treatment with GsMTx4 led to reduced tooth movement in rats20. However, in this particular setup, all cells of the dentoalveolar apparatus are affected when the inhibitor is administered systemically. Therefore, it is impossible to predict the importance of individual cell types in the overall process.

Overall, our results suggest that despite Piezo1 being crucial for osteoclast function, it may be dispensable for mechanical sensing of bone remodeling.

Materials and methods

Mice

Generation and genotyping Piezo1floxed/floxed and Piezo2floxed/floxed mice have been described previously38,39. For the study, a mouse line with a 14 kb Dmp1 promoter fragment (JAX #023047) was used to drive specific expression of Cre Recombinase. The generation of the mouse line has been previously described by Lu et al.40. To rule out any possible influence of genetic background all analyses were performed with the corresponding Cre-negative littermate controls. These Cre-negative littermates are named control animals in the manuscript.

For this particular study, we utilized a total of 12 mice to that exhibited orthodontic tooth movement, including 6 control mice and 6 Piezo1floxed/floxed;Dmp1cre mice. In addition to evaluating tooth movement, we also analyzed the skull phenotype of 3 Piezo2floxed/floxed;Dmp1cre mice.

Generation of Dspp-Cerulean;Dmp1Cherry reporter mice has been described by Vijaykumar et al.41 and experimental animals were kindly provided by Jan Krivanek (Department of Histology and Embryology, Faculty of Medicine, Masaryk University, Brno, Czech Republic).

Mice were housed in a specific pathogen-free environment with a 12-h light/dark cycle, 45% to 65% relative humidity, and 20 °C to 24 °C ambient temperature in open or individually ventilated cages with wood shavings bedding and nesting material in groups not surpassing six animals. Mice had ad libitum access to tap water and standard rodent chow. Animal care and handling as well as all experimental procedures were performed in accordance with all national and european guidelines. All methods were performed in accordance with the relevant guidelines and regulations. The study is reported in accordance with ARRIVE guidelines.

The animal procedures were approved by the local animal care committees of the University Medical Center Hamburg-Eppendorf and the University of Leipzig Medical Center (TVA 062/20 and TVV 03/21).

Orthodontic appliance

The experiment was performed as a split-mouth model, with the untreated jaw side as an internal control. The experimental procedure was described before21. Mice were anesthetized with a solution of ketamine (40 mg kg−1 bw Ketamin-S), xylazine 2% (16 mg kg−1 bw), and 0.9% NaCl. After etching the left first maxillary molar and the incisors (HS-etch gel 37%, Henry Schein Dental, Langen, Germany), bonding was applied to the teeth, and the distal end of the nitinol open coil spring (Sentalloy Open Coil Spring, Dentsply Sirona, Pennsylvania, USA) was bonded to the surface of the left first maxillary molar with a force of 30 centinewtons using light-cured resin (Estelite Flow Quick, Tokuyama Dental Corp., Tokyo, Japan). After orthodontic treatment, the animals were monitored on a heating mat until their emergence. Animals were euthanized by CO2 inhalation after 12 days of OTM. The heads were fixed in 4% PFA at 4 °C for 24 h and then transferred to 80% ethanol.

Micro-computed tomography

The skulls were scanned with a voxel resolution of 15 µm/5 µm using micro-computed tomography (µCT 40, Scanco Medical, Brüttisellen, Switzerland/Skyscan 1172, Bruker, Kontich, Belgium) as described previously42,43. Wall Thickness Analysis and Bone Volume measurements were performed on 3-dimensional reconstructions of the scans with the software Avizo (FEI, USA). Horizontal Alveolar bone loss (CEJ-ABC) was measured on 2-dimensional microCT-pictures with ImageJ as previously described44.

Histology

After performing µCT-analysis, the upper jaws were prepared for histological examination. They were decalcified using 10% EDTA for 14 days, while the reagent was changed once a week. Afterward, the upper jaws were cut in the parasagittal plane. The halves were dehydrated in ascending alcohol concentrations and embedded in paraffin. Sections of 5 µm thickness were prepared using a microtome. Toluidine Blue staining was performed as previously described, and TRAP staining was performed using the Sigma Aldrich TRAP Kit (387A, Sigma Aldrich, Missouri, USA) according to the manufacturer's instructions.

All histomorphometric analyses were performed with QuPath 0.4.2.45 and ImageJ 1.52 (National Institutes of Health, Betheseda, MD, USA) and its implication BoneJ 1.4.346. Trap Staining was quantified as the Number of Osteoclasts per bone area in the pressure zone of the mesial and distal root and as well in the interrradicular bone with QuPath.

Fluorescence immunohistology

Immunohistology was carried out on paraffin sections. For this, the sections were deparaffined in xylene baths. The samples were rehydrated in descending alcohol series for 2 min each, followed by a short wash in distilled water. Afterward, antigen retrieval was performed with Proteinase K (0,1ul/ml in Tris–HCl) for 20 min. After rinsing in PBS the staining was performed with PIEZO1 (MA5-32876 1:50 in Dako Solution (# s1699)) for 24 h at RT and PIEZO2 (PA5-72975 1:100 in Dako Solution (# s1699)) for 24 h at RT. After rinsing in PBST (0,1% Tween) the secondary antibodies M555 (Invitrogen A28180) and Rb647 (Invitrogen A32733) were diluted in Dako solution 1:1000 and put on the slides for 1 h at RT. DAPI staining (1:1000 in PBST) was perfomed for 5 min at RT. After rinsing, the slides were mounted in Mowiol 4–88 (Kuraray Europe GmbH, Germany) Mounting Medium. Sections were examined and imaged using the Leica Thunder Imager 3D Assay (Leica Microsystems, Wetzlar, Germany) with filter cubes optimized for the detection of GFP (HQ470/40ex, HQ525/50em, HQ495lp beam splitter), TXR (HQ560/40ex, HQ630/75em, HQ585lp beam splitter), Y5 (HQ620/60ex, HQ700/75em, HQ660lp beam splitter) and DAPI (HQ405/60ex, HQ470/40em, HQ455lp beam splitter).

Statistical analysis

The statistical analysis was performed with GraphPad PRISM9 (GraphPad Software, San Diego, USA). Results were expressed as the mean ± SD. Two-sided t-test and one-way ANOVA with Bonferroni post hoc test were used for multiple group comparisons. P values below 0.05 were considered statistically significant.