The role of endothelin B receptor in bone modelling during orthodontic tooth movement: a study on ETB knockout rats

The endothelin system has an important role in bone modelling during orthodontic tooth movement (OTM); however, little is known about the involvement of endothelin B receptors (ETB) in this process. The aim of this study was to evaluate the role of ETB in bone modelling during OTM using ETB knockout rats (ETB-KO). Thirty-two male rats were divided into 4 groups (n = 8 per group): the ETB-KO appliance group, ETB-KO control group, wild type (ETB-WT) appliance group, and ETB-WT control group. The appliance consisted of a super-elastic closed-coil spring placed between the first and second left maxillary molar and the incisors. Tooth movement was measured on days 0 and 35, and maxillary alveolar bone volume, osteoblast, and osteoclast volume were determined histomorphometrically on day 35 of OTM. Next, we determined the serum endothelin 1 (ET-1) level and gene expression levels of the osteoclast activity marker cathepsin K and osteoblast activity markers osteocalcin and dentin matrix acidic phosphoprotein 1 (DMP1) on day 35. The ETB-KO appliance group showed significantly lower osteoblast activity, diminished alveolar bone volume and less OTM than the ETB-WT appliance group. Our results showed that ETB is involved in bone modelling in the late stage of OTM.

. Model of the OTM. Applying of orthodontic force to the tooth compresses the PDL. At the compression side of the tooth, which appears in the direction of the orthodontic force, bone resorption takes place, carried out mainly by osteoclasts. At the tension side, osteoblasts are responsible for the bone formation process.
Scientific RepoRtS | (2020) 10:14226 | https://doi.org/10.1038/s41598-020-71159-8 www.nature.com/scientificreports/ las, Texas, USA). The ET B -KO animals were incorporated with a dopamine beta-hydroxylase (DβH) transgene to enable the development of a normal enteric nervous system. The DβH transgene was also inserted into the control animals 40 . The rats were bred in a homozygous line at the Faculty of Medicine, University of Ljubljana (Slovenia). In this study we used 32 male ET B -KO rats (285 ± 27 g, 13-15 weeks old) and 32 male ET B -WT rats (286 ± 30 g, 13-15 weeks old). The animals were housed as well as procedures were identical to our previous studies 11 . The part of daily intake of rat chow (Teklad 2016 Global rodent diet, Harlan, The Netherlands) was soaked in water to facilitate food intake due to its mild impairment during orthodontic force application.
All the animal procedures and the study protocol were approved by the Ethics Committee for Animal Experiments of the Administration of the Republic of Slovenia for Food Safety, Veterinary Sector and Plant Protection (34401- 62/2008/20), and complied with the guiding principles in ''The Care and Use of Animals'' .
The orthodontic appliance consisted of a superelastic closed coil spring (25 cN; wire diameter, 0.15 mm; GAC Dentsply International, York, PA, USA) which was placed between the first and second left maxillary molars and the incisors by a stainless-steel ligature in the ET B -KO and ET B -WT appliance groups 11 . The coil spring was attached to a steel thread placed around the first and second molar on one side and through a drilled hole into the upper incisors on the other side (Fig. 2). The hole was drilled laterally in the incisors through the area where the vivid tooth structures were unaffected 12 . The orthodontic appliance was placed in each animal under general anaesthesia at the beginning of the study, and replaced to the correct position every 7 days, ensuring its proper activation and the exertion of constant force on the teeth 41 . tooth movement measurements. The distance between the most mesial point of the maxillary first molar and the most distal point of the ipsilateral incisor at the gingival level was measured on the experimental sides in all four groups. The measurements were obtained using a digital calliper (Digitronic Calliper, 144-15 D (Wilson & Wolpert, Utrecht, The Netherlands)) while the animals were under general anaesthesia at days 0 and 35. All the measurements were carried out twice by two investigators independently within a period of few minutes and the reliability of the measurements was assessed by using the intraclass correlation coefficient (ICC) as described in previous studies 11,42 . Tooth movement was calculated by subtracting the distance between the teeth on day 35 from the distance between the teeth on day 0. preparation of histology samples and bone histomorphometry. On day 35, all the animals were sacrificed by intraperitoneal injection of anaesthetics and carbon dioxide. Tissue samples of the maxilla containing all 3 molars were taken. Samples were prepared in vertical section perpendicular to the occlusal plane of the molars. Paraffin blocks were then prepared and stained with haematoxylin and eosin (Fig. 3) as previously published in Plut et al. 42 .
Bone histomorphometry was used to determine alveolar bone, osteoclast and osteoblast volume density in all four groups. Histomorphometry was performed by using a point-counting method. For this purpose, the stereologic cycloid grid system incorporated into the ocular of a light microscope (BX-60, Olympus, Tokyo, Japan) was used. As described by Sprogar et al. 11 , the alveolar bone area was expressed as the percentage of alveolar bone area versus the tissue area consisting of tooth, PDL, connective tissue and bone marrow spaces. In addition, the osteoblast and osteoclast areas were defined as the alveolar bone area covered with osteoblasts or osteoclasts versus alveolar bone area. The cells were counted in the alveolar bone alongside the mesial and distal roots of the second molar 42 . However, because 20 sections from each specimen were examined, alveolar bone area, osteoblast area and osteoclast area were extrapolated to alveolar bone volume, osteoblast volume and osteoclast volume, as we already described in previous studies 11,42 . Briefly, standards and samples were added to the designated wells of a microplate, precoated with an endothelin-1 specific antibody. The plate was incubated at room temperature for 1 h, washed and HRP-conjugated. The endothelin 1 detection antibody was added to each well. After incubation, the unbound detection antibodies were removed and TMB was added to visualise the HRP enzymatic reaction. After incubation, a stop solution was added and the absorbance was read at 450 nm, with correction at 570 nm, using Tecan Safire (Tecan Group Ltd., Switzerland).
RnA isolation and semiquantitative Rt-pcR. Osteocalcin and DMP1 gene expression levels were used to determine osteoblast activity, and the cathepsin K gene expression level was used to determine osteoclast activity in all 4 groups 43,44 . The maxillary bones with all 3 molars and their PDLs were excised and immediately frozen in liquid nitrogen. RNA isolation and semiquantitative RT-PCR was performed as described by Plut et al. 42 . Oligonucleotides for cathepsin K, osteocalcin and DMP1 were chosen from predesigned assays (TaqMan Gene Expression Assays, Applied Biosystems). Thermal cycling, construction of standard curve and cDNA amplification and quantification were performed as we reported in a previous study 43 . In order to exclude variations from different inputs of total mRNA to the reaction, data on cathepsin K, osteocalcin and DMP1 were normalised to an internal housekeeping gene, GAPDH, for which data was obtained by using TaqMan GAPDH predesigned assays (TaqMan Gene Expression Assays, Applied Biosystems). All the reactions for standard samples and for samples from all 4 groups were performed in duplicate. The data were averaged from the values obtained in each reaction 43 . For all the transcripts tested, a time-course-dependent gene expression consensus profile was observed after normalisation to the expression of the housekeeping gene GAPDH.
Statistical analyses. The data were expressed as means ± standard error of the mean (SEM) and calculated for each parameter for all the animal groups. The evaluated parameters were tooth movement, alveolar bone volume density, osteoclast and osteoblast volume densities, serum ET-1 level, and gene expression levels of cathepsin K, DMP1 and osteocalcin. Comparisons within and between the groups were performed using analysis of variance (ANOVA), followed by the Tukey multiple comparison test. A P value less than 0.05 was considered statistically significant.
Interexaminer reliability for the measurements of the distance between the teeth was tested using the intraclass correlation coefficient (ICC).
In the results, not all the groups contained the initial number of rats (n = 8 per group). During the experiment some of the rats had to be excluded. The number of samples in each group is explicitly stated in the Figures.

Results
Orthodontic tooth movement and physiologic distal drift measurements after 35 days. The overall mean value of the ICC for all the measurements of the distances was 0.938. In the ET B -KO appliance group the amount of OTM (1.67 mm ± 0.10 mm) was significantly less pronounced than in the ET B -WT appliance group (2.28 mm ± 0.17 mm) on day 35 (P = 0.0255). The physiologic distal drift did not significantly differ between the ET B -KO and ET B -WT groups (Fig. 4).

Histomorphometric analyses.
The histomorphometric analysis showed that after 35 days of OTM the alveolar bone volume was significantly lower in the ET B -KO appliance group (35.26% ± 1.47%) in comparison to ET B -KO control group (44.45% ± 1.63%) (P = 0.0004). Furthermore, the alveolar bone volume was significantly less in the ET B -KO appliance group than in the ET B -WT appliance group (45.26% ± 1.74%) (P = 0.0001) (Fig. 5a). No significant differences in the osteoblast volume were observed between the groups. Osteoblast volume was non-significantly increased in the ET B -KO control group in comparison to the other three groups   www.nature.com/scientificreports/ (Fig. 5b). The osteoclast volume was significantly increased in the ET B -KO (1.08% ± 0.13%) and ET B -WT appliance (1.30% ± 0.10%) groups compared to their controls (0.47% ± 0.06% and 0.69% ± 0.06%, respectively) (P < 0.0001). The osteoclast volumes in the ET B -KO appliance and the ET B -KO control groups were less than in the ET B -WT groups, but the differences were not significant (Fig. 5c).
Rt-pcR analysis. The gene expression level of osteocalcin on day 35 was significantly downregulated in the ET B -KO appliance group compared to the ET B -WT appliance group (P = 0.0157). A significant difference in the gene expression level of osteocalcin was also determined between the ET B -KO appliance group and the ET B -KO control group (P = 0.0288) (Fig. 7a). Similarly, the gene expression of DMP1 was significantly downregulated in the ET B -KO appliance group in comparison to the WT appliance group (P = 0.0040) (Fig. 7b). The gene expression level of cathepsin K was downregulated in both appliance groups compared to the control groups. Significant differences were determined between the ET B -KO control group compared to the ET B -KO appliance group (P = 0.0314) (Fig. 7c).

Discussion
The results of this study showed that ET B is involved in bone modelling during the late stage of OTM in the rat animal model. The amount of OTM after 35 days of the experiment was significantly less in the ET B -KO appliance group than in the ET B -WT group (P = 0.0255). There was a significant difference in the alveolar bone volume in the ET B -KO appliance group compared to the ET B -WT appliance group (P = 0.0004), probably due to diminished osteoblast activity in the ET B -KO appliance group. Some of the ET B antagonists showed significant inhibition of ET-1 effects in vitro 45,46 ; however, none of the available antagonists have an established pharmacological and toxicological profile. Furthermore, they have to be administered intravenously and daily intravenous application over a longer period of time represents a great deal of stress for the animals. To study the effects of ET-1 in a reduced ET B response, the most suitable model is to use ET B knock-out animals. The ET B -KO strain of rats used in the present study is described as a natural mutation in the progeny of a Wistar rat. Mutations in the ET B gene have been linked to Hirschsprung's disease in humans, a congenital disease characterised by aganglionic megacolon, an absence of enteric ganglia, and a lack of innervations to the lower gastrointestinal tract. The disease is associated with polymorphism and several missense mutations in the ET B gene which lead to decreased expression, changes in cell signalling, and loss of endothelin ET receptor function 47,48 . The animals used in this study were incorporated with a transgene (DβH) to enable the development of a normal enteric nervous system. The resulting transgenic rats are healthy but present with a total absence of ET B in all non-adrenergic tissues 40 . The animal model of OTM used in this study had already been confirmed as appropriate for studying the role of endothelin system in bone modelling [10][11][12] . The advantage of the model was the minimally invasive placement of the coil spring between the molars and incisors  Both appliance groups showed lower alveolar bone volume compared to their control groups. Furthermore, the alveolar bone volume in the ET B -KO appliance group was significantly less than that of the ET B -WT appliance group. Differences in alveolar bone volume during physiological and pathological processes depend on the relationship between bone formation and bone resorption. Similarly to our study, histomorphometric analyses in previous studies have shown that alveolar bone volume was less in appliance groups than in their control groups 11,42,43 . We studied bone formation by determining osteoblast volume and osteoblast activity using osteocalcin and DMP1. There were no significant differences in osteoblast volumes between the four groups, but there was a significant decrease in osteoblast activity as determined by the gene expression levels for osteocalcin and DMP1 in the ET B -KO appliance group compared to the ET B -WT appliance group after 35 days of OTM. It has been shown that bone formation after the application of orthodontic force is increased predominantly by stimulating the differentiation of osteoblasts, and to a lesser extent by an increase in the number of these cells 50 . This is in concordance with our results, which show no considerable difference in osteoblast volume between the groups, but significant changes in osteoblast activity. Osteocalcin plays an important role in the maturation of mineral species 51 , and modulates osteogenic differentiation of MSCs 52 ; DMP1 has a similar role in osteoblast differentiation and matrix mineralisation. Interestingly, it has also been reported that recombinant DMP1 induces the osteogenic differentiation of human periodontal ligament cells 53 , and appears to play an important role in the osteogenic differentiation of dental follicle stem cells 54 .
In a study on Bone Marrow-Derived Stem Cells (BMSCs) it was shown, using specific antagonists, that both receptors ET A and ET B are involved in the differentiation of BMSCs into active osteoblasts, and the osteogenesis of BMSCs was attenuated by blocking ET A and/or ET B receptors. The findings of this study reveal that both ET A and ET B receptors and downstream AKT and ERK signalling are involved in ET-1 primed lineage specification of MSCs 24 . Similarly, a study on ET B -KO mice showed less fibroblast activation and myofibroblast formation in response to bleomycin or ET-1 25 . Therefore, in our study, the lower expression levels of osteoblast activity markers, DMP1 and osteocalcin could be a result of attenuated osteoblast maturation and/or osteogenic differentiation of MSCs, a process that is mediated through both receptors ET A and ET B . In the absence of ET B in the ET B -KO appliance group, osteogenesis is attenuated, resulting in a lower alveolar bone volume and a decreased amount of OTM (Fig. 8).
The bone resorption process was studied by determining osteoclast volume and osteoclast activity using cathepsin K, a marker of bone resorption. A considerable increase in osteoclast volume was determined in both appliance groups compared to the control groups after 35 days of OTM. Similar results were obtained in a previous study, where osteoclast volume increased in all animal appliance groups 11 . In the late stage of OTM, we expected the upregulation of the cathepsin K gene expression level in the ET B -KO and ET B -WT appliance groups, whereas the osteoclast volume increased significantly in both appliance groups 43 . However, a significant downregulation of the gene expression level of cathepsin K appeared in the ET B -KO appliance group compared to ET B -KO control group. Several lines of evidence suggest that changes in osteoclast function and volume/number do not always happen simultaneously 55 . In a previous study it was shown that the absence of ET B in osteoclast specific ET B -KO mice impairs the formation of mature osteoclasts and impairs bone resorption activity with  In the present study, the amount of OTM in the ET B -KO appliance group was significantly less than in the ET B -WT appliance group. Because of the diminished alveolar bone volume and lower osteoblast activity, a higher amount of OTM would initially be expected. However, we observed a lower amount of OTM, and similar results were also reported in Plut et al. 42 . During OTM, due to the lower osteoblast activity, bone formation at the tension site cannot keep up with bone resorption at the compression site, and the consequences are a widening of the PDL space and increased tooth mobility 57 . Tooth movement requires a coupling of bone resorption and bone formation. Due to the delayed bone formation in the absence of ET B there may be a further aggravation of the uncoupling of bone formation and bone resorption, expressing as suppressed bone turnover and resulting in lower amount of OTM. In rats, molars naturally drift distally. Therefore, there is predominately bone resorption on the distal side and bone formation on the mesial side of molar roots 58 . In the present study, the distal drift was smaller in the ET B -KO control group than in the ET B -WT control group, but there was no significant difference. These results suggest a different mechanism of alveolar bone turnover under physiological conditions in comparison to OTM, shown in the ET B -KO groups.
In the ET B -KO control group the serum ET-1 level was significantly higher than in the ET B -WT control group. It was also significantly elevated in the ET B -KO appliance group in comparison to ET B -WT appliance group. This is in concordance with many studies that confirmed the role of ET B as a ET-1 clearance receptor 32,[37][38][39]59 . The lower amount of OTM in this study could be assigned to several processes, among them attenuated osteoblast maturation and increased ET-1 due to a lack of circulatory ET-1 clearance. High ET-1 levels are normally almost exclusively cleared by endothelial ET B, which was lacking in our animal model. It has been shown that chronic exposure of ET receptors to increased plasma ET-1 levels results in a significantly reduced density of ET A in mice aortas, correlating with a reduction in functional response to ET-1 60 . Similarly, ET B -KO mice were found to have a 45% lower ET A density and significantly reduced expression (lower ET A mRNA) in peripheral tissues with no change in receptor affinity. A potential mechanism of reduced ET A density in ET B -KO models was proposed as the compensatory downregulation of the receptor in response to higher circulating ET-1 levels as a result of a lack of ET B clearing receptors or some other role of ET B in the development of cells expressing ET A receptors 60,61 . It is therefore possible that in our study high serum ET-1 in the ET B -KO groups resulted in downregulation of the ET A receptors, which already have an established role in bone modelling, resulting in diminished alveolar bone volume and lower amount of OTM (Fig. 8). www.nature.com/scientificreports/ Many studies report interactions (cross talk) between ET A and ET B receptors, which means that the activation or inhibition of one receptor subtype can alter the function of another receptor subtype. For example, using ET B receptor-deficient rats it was reported that both ET A and ET B are involved in ET-1-induced DNA synthesis in astrocytes, accompanied by MAPK activation 62 . Similarly, it was shown that both ET A and ET B regulate lung myofibroblast proliferation, indicating possible interactions between receptor subtypes 63 . In addition, endothelin receptors (ET A and ET B ) also exist in the form of homo-and heterodimers 64 , and for many receptor heterodimers, co-expression of both receptor subtypes is crucial for functional receptor activity, pharmacological proprieties, maturation, and proper cell-surface trafficking 65,66 . Another example of cooperation between receptor subtypes are HEK 293 cells which, transfected with both receptors, display a considerably prolonged increase in intracellular Ca 2+ concentration in response to ET-1 or the selective ET B receptor agonist Sarafotoxin 6c, in comparison to more transitory responses of cells transfected with either receptor subtype alone 67 . It is therefore possible that similar interactions between the two receptor subtypes are also necessary for normal bone modelling during the late stage of OTM.
Another potential explanation for reduced OTM is anti-apoptotic activity of elevated ET-1 in the two ET B -KO groups. Specifically, apoptosis of osteocytes is critical for osteoclast activation and resorption at the compression site of the tooth during OTM 68,69 . Moreover, in a study of OTM with micro-osteoperforations it was shown that the rate of OTM is increased by increased apoptosis and cell proliferation of PDL cells 70 . Therefore, the lower amount of OTM in the absence of ET B could be a result of several different processes, including the anti-apoptotic activity of ET-1 on numerous cells in periodontal tissues. One limitation of the study, however, are only two time points in the experiment. Thus, no exact mechanism of the action can be elucidated. We rather propose possible mechanisms through which ETB-KO modulated bone modelling during OTM (Fig. 8).

conclusion
Our results showed for the first time that ET B is involved in bone modelling in the late stage of OTM. ET B -KO resulted in lower osteoblast activity and therefore decreased alveolar bone volume and lower amount of OTM. This could be due to the role of ET B in osteoblast maturation and/or the differentiation of mesenchymal precursor cells, the adaptive downregulation of ET A as a response to high levels of circulating ET-1 or anti-apoptotic activity of ET-1. Further research is needed to explain the exact mechanism by which ET B modulates bone modelling in OTM, and which of the proposed mechanisms is predominant.

Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.