Effects of estrogen deficiency during puberty on maxillary and mandibular growth and associated gene expression – an μCT study on rats

Background Estrogen is a well-known and important hormone involved in skeletal homeostasis, which regulates genes involved in bone biology. Some studies support that estrogen is important for craniofacial growth and development. Therefore this in vivo animal study aimed to investigate, whether and in which way low estrogen levels in the prepubertal period affect craniofacial development in the postpubertal stage and to quantify the gene expression of RANK, RANKL and OPG in cranial growth sites in ovariectomized estrogen-deficient rats during puberty. Methods Control (sham-operated, n = 18) and ovariectomy (OVX, n = 18) surgeries were performed on 21-days-old female Wistar rats. Animals euthanized at an age of 45 days (pubertal stage) were used for gene expression analyses (n = 6 per group) and immunohistochemistry of RANK, RANKL and OPG. Animals euthanized at 63 days of age (post-pubertal stage) were used for craniofacial two-dimensional and three-dimensional craniofacial measurements using μCT imaging (n = 12 per group). Results In the μCT analysis of the mandible and maxilla many statistically significant differences between sham-operated and OVX groups were observed, such as increased maxillary and mandibular bone length in OVX animals (p < 0.05). Condylar volume was also significantly different between groups (p < 0.05). The sham-operated group showed a higher level of RANK expression in the midpalatal suture (p = 0.036) and the RANKL:OPG ratio levels were higher in the OVX group (p = 0.015). Conclusions Our results suggest that estrogen deficiency during the prepubertal period is associated with alterations in the maxillary and mandibular bone length and condylar growth.


Background
Postnatal craniofacial skeletogenesis is a unique process, in which many factors can affect growth and development. Estrogen is an important hormone involved in the skeletal homeostasis that regulates different aspects of bone metabolism, development, modeling and remodeling. Endogenous levels of estrogen change according to age and gender [1]. During puberty, there is an increase in estrogen levels leading to the development of secondary sexual characteristics and a significant increase in growth rate [2]. During the pubertal growth spurt, this hormone plays an important role controlling growth plate acceleration and fusion [3]. Estrogen deficiency has been reported in syndromes and genetic conditions [4][5][6], menstrual disorders [7], primary ovarian insufficiencies [8], underweight [9], excessive exercise [10] and chemotherapy [11]. It is well known that estrogen deficiency can cause osteoporosis [12,13], decrease mineral density in bones and delay the epiphyseal maturation [14].
It is well established that sex steroids, including estrogen, regulate the RANK (receptor activator of nuclear factor-κB), RANKL (receptor activator of nuclear factor-κB ligand) and OPG (osteoprotegerin) axis, members of the tumor-necrosis-factor superfamily. One of the most important downstream mediators of the action of estrogen on bone is RANKL [15], which is crucial for osteoclast differentiation, activation and survival. OPG is a soluble decoy receptor binding RANKL, inhibiting osteoclastogenesis via the RANK receptor on osteoclast precursor cells [16,17]. RANKL, which is expressed membrane-bound by osteoblasts and can also be released in soluble form, acts via its receptor RANK, which is expressed on the cell membrane of osteoclasts and osteoclast precursor cells [18]. RANK, RANKL and OPG are essential, nonredundant factors for osteoclast biology [15].
Some studies using rodent animal models were conducted previously to describe, how estrogen affects growth and development of craniofacial structures [19][20][21][22][23] and a variety of outcomes were observed. Some studies demonstrated a growth inhibition of the craniofacial complex in estrogen-deficient newborn mice [14,20,21], while increased condyle dimensions were observed, when the estrogen deficiency was induced in 8-weeks-old mice [19]. In a study performed on rats with estrogen deficiency induced at an age of 30 days, no alterations in craniofacial growth were observed [1]. On the other hand, in our previous study with estrogen deficiency induced in the prepubertal stage (age: 21 days), estrogen-deficient animals presented increased maxillary and mandibular measurements [23]. In this previous project from our research group, the two-dimensional radiographic analysis performed suggested that estrogen deficiency from the prepubertal stage affects the dimensions of the maxilla and mandible in female rats [23]. Therefore, in the present study, we used high-resolution microcomputed tomography (μCT) to analyze the maxillary and mandibular skeletal dimensions of adult female rats, which were subjected to estrogen deficiency during the prepubertal stage, in three dimensions. We also evaluated, if estrogen deficiency affects craniofacial growth via the RANK, RANKL and OPG axis by investigating their expression at growth sites of both maxilla and mandible during puberty and estrogen deficiency.

Ethical aspects
The present study was performed and reported in accordance with the ARRIVE guidelines [24]. The Ethical Committee in Animal Experimentation from the School of Dentistry of Ribeirão Preto, University of São Paulo, Brazil, approved the protocol of this study (2014.1.721.58.7).

Sample selection
The sample size was calculated for the application of independent-measures t tests based on estimates from a pilot study in radiographic images (n = 5). Several calculations were performed considering the results on each comparison related to the measures evaluated in the morphometric analysis and gene expression analysis. The highest estimations of the outcomes analyzed were considered as the sample size required. The following parameters were considered for the highest sample size estimation for the morphometric analysis: effect size = 1.4, α = 0.05 (5% error), power = 0.8, number of groups = 2. Regarding the gene expression outcome, the effect size considered was 2.2 and the other parameters were the same to the above-mentioned. The calculation estimated a minimum total sample of 20 animals for morphometric analysis and 10 animals for gene expression (total n = 30). Considering the possibility of using nonparametric statistics and possible losses, an increase of 20% was performed, which resulted in the adjustment of the sample size for 36 animals (6 rats for gene expression and 12 rats for morphometric analysis, per group). The sample size calculation was performed in G*Power (version 3.1.9.7). Posteriorly, the animals were randomly assigned into OVX and sham-operated control groups using sealed envelopes to ensure the allocation concealment.
The housing room was temperature and humidity controlled and rats had ad libitum access to food. Briefly, the rats were anesthetized using an intraperitoneal injection of 10% ketamine hydrochloride (55 mg/kg of gross body weight) and 2% xylazine hydrochloride (10 mg/kg of gross body weight). At an age of 21 days (prepubertal stage), bilateral ovariectomy was performed in the OVX group and placebo surgery was performed in the shamoperated group according to the protocol of Omori et al. [23]. At an age of 45 days (pubertal stage) animals were euthanized for gene expression analyses. At an age of 63 days (post-pubertal stage -young adult) the remaining rats were euthanized for morphometric μCT analyses. Ages/developmental stages were established according to Sengupta [25].
As previously described by Chen et al. [26] and Omori et al. [23], body weight and uterus atrophy were significantly higher in the OVX rats than in sham-operated rats at 63 days of age (p < 0.05), confirming the success of ovariectomy. In case that the uterus atrophy was not observed, the animal would be excluded from the analysis.

Morphometric μCT analysis
The entire maxilla and mandible of each rat was retrieved after euthanasia and scanned with the micro-CT system phoenix v|tome|x s 240/180 research edition from GE Sensing & Inspection Technologies GmbH in the Regensburg Center of Biomedical Engineering. Scanning parameters for the rats' upper jaws were as follows: 60 kV voltage, 800 μA current, 333 ms time, 1500 images, voxel size 55 μm, fast scan; while for the rats' lower jaws settings were as follows: 60 kV voltage, 600 μA current, 333 ms time, 1500 images, voxel size 37 μm, fast scan. Reconstructed volumes were processed using the corresponding manufacturer's software phoenix datos|× 2 reconstruction 2.4.0 (GE Sensing & Inspection Technologies GmbH, Wunstorf, Germany).
The 3D data of the maxilla and mandible were both analyzed using VGSTUDIO MAX 3.3 -Voxel Data Analysis and Visualization (Volume Graphics GmbH, Heidelberg, Germany) and Image J software (National Institutes of Health, Bethesda, MD, USA). To perform the maxillary and mandibular morphometric measurements, the 3D jaw images obtained by μCT were aligned in dorsal, lateral and ventral view and standardized scaled 2D images were taken for all rats. Image J software was used to measure linear (mm) and angular (°) dimensions within these 2D images. The used landmarks for both maxilla and mandible are demonstrated in Fig. 1 and described in Table 1 and were selected based on Wei et al. [27], Fujita et al. [20], Corte et .al [28], Wang et al. [29] and Perilo et al. [30]. The linear and angular measurements evaluated here are described in Table 2. One single calibrated examiner blindly performed all morphometric analyses. Each measurement was taken three times and the mean of the three measurements was used to perform statistical analyses. To assess intraexaminer reliability, the Intra-class Correlation Coefficient (ICC) was calculated to assess the concordance of measurements.
Condylar volume was also obtained using VGSTUDIO MAX 3.3. All mandibles were aligned in a lateral view of the right condyle, which was separated from the mandible using a specific software tool to measure its volume (mm 3 ). The amount of bone selected to measure the volume was standardized by positioning a pre-sized square (12 mm × 12 mm) over the condyle, in a way that the superior edge of the square matched with the most superior border of the condyle, and the left edge of the square matched with the most posterior border of the

Gene expression of RANK, RANKL and OPG
Gene expression analysis was performed in growth sites of the maxilla (midpalatal suture) and mandible (condyle, mandibular angle, symphysis/parasymphysis and coronoid process) at the pubertal stage. Bone samples were dissected after euthanasia at the age of 45 days and stored in RNAlater (Life Technologies Corporation -Carlsbad, CA, USA) at − 80°C until processing. Total RNA was extracted using the mirVana™ miRNA Isolation kit (Ambion/Life Technologies™, USA). Complementary DNA (cDNA) was synthesized by reversetranscription with a High Capacity kit (Applied Biosystems, Foster City, CA, USA).
Quantitative real-time polymerase chain reaction (RT-qPCR) was blindly performed using a StepOnePlus™ sequence detection system (Applied Biosystems™, Foster City, CA, USA). The thermal cycling was carried out by Table 1 Description of maxillary and mandibular landmarks used ) and ACTB (Rn01412977-g1) were used as endogenous controls and confirmed to be stably expressed. The relative levels of mRNA expression were determined by the 2 -ΔΔ CT method. GAPDH and ACTB genes were used for sample normalization according to Livak and Schmittgen [31] to calculate relative gene expression. All procedures were performed following the respective manufacturer's instructions and according to established protocols.

Immunohistochemical analysis of RANK, RANKL and OPG
Immunohistochemical analysis was performed as previously described [32].

Results
Intraexaminer reliability of measurements was good with ICC ≥ 0.78. Eight animals died during the experiment. All the remaining animals were included, as follow: For gene expression analysis, samples from 5 OVX rats and samples from 4 sham-operated rats were used. For morphometric analysis, 8 skulls of OVX rats and 11 skulls of sham-operated rats were evaluated. The overall dimensions in millimeter (mm) of the reconstructed μCT images of the rats' maxilla and mandible are presented in Table 3. The maxillary posterior segment length was smaller in the OVX group (p = 0.012), while the maxillary central incisor length was bigger in the OVX group (p = 0.010). In the mandibular sagittal plane, mandibular height (p = 0.006), upper mandibular length (p = 0.037), mandibular plane length (p = 0.019), diagonal mandibular length (p = 0.014), the distance between the condyle to mandibular angle (p = 0.025) and mandibular height (p = 0.004) were bigger in the OVX group. Also, the mandibular angle composed by the lines intersecting the landmarks 17-19 and 20-22 was higher in the OVX group (p = 0.009). In the mandibular transversal plane, the mandibular interdiastemal breadth (p = 0.002) and the thickness of the condylar process (p < 0.0001) were larger in the OVX group.
Gene expression of RANK, RANKL, OPG and the RANKL:OPG ratio are presented in Table 4. The shamoperated group had a higher level of RANK expression in the midpalatal suture (p = 0.036). The RANKL:OPG The immunohistochemical analysis showed a pattern of results similar to that observed in gene expression tests (Fig. 3). Immunostaining for RANK was more intense in the midpalatal suture chondrocytes for the sham-operated group. RANKL staining was more pronounced in the proliferative and hypertrophic layers of the mandibular condyle for the OVX group, while, on the contrary, OPG staining was more intense in the condyle for the sham-operated group.

Discussion
Estrogen deficiency during (pre) pubertal stage may impact on the development of the facial complex. Our study using a rodent model suggested that estrogen is one of the factors involved in the maxillary and mandibular growth and development. Since low estrogen levels in women and teenage girls can be caused by different conditions and their effects depends on the individual's age and general health, clinicians should be aware of the possible potential impact of estrogen deficiency in dental arch development of girls.
In animal models, it is well known that low levels of estrogen can lead to changes in bone microarchitecture in femurs and mandibles [33], osteoporosis [13] and alterations in the alveolar bone [34] as well as alterations in craniofacial development [14,[19][20][21]23]. In our study, we were able to extend upon the radiographic two-dimensional cephalometric linear findings reported by Omori et al. [23], who found that estrogen deficiency during puberty led to alterations in maxillary and mandibular dimensions. The use of images from μCT in the present study allowed us to perform a more reliable and complete analysis, adding more landmarks and different view positions of both arches resulting in a more detailed description of the phenotypes. Radiographic cephalometric linear and angular analysis of murine skulls has been developed long ago [35] and is very similar to that used in humans, which has practical and successful clinical applications [27]. Although some landmarks are difficult to identify on two-dimensional radiographs of mouse and rat skulls, radiograph-based cephalometric has been used successfully to identify morphometric changes in estrogen models of mice [14,20] and rats [23,36]. However, μCT-based craniofacial measurements have the advantage of a high resolution and the ability to determine both morphology and volume.
In our study, the thickness of the condylar process, as well as condylar volume were bigger in the OVX group. Larger measurements of condyle breadth were found in mice, when estrogen deficiency was induced in eightweek-old mice [20], although a normal width, low trabecular bone volume of the condyle and reduction in bone mineral density were also found [12,13,21].
We were able to observe that although the posterior segment length of the maxilla was smaller in OVX group, the others measurements statistically different among the groups -maxillary dimension (maxillary central incisor length) and the mandibular dimensionswere bigger in the OVX group. The phenotype observed in the maxilla of OVX group, in which the posterior length was smaller, could reflect a maxillary retrusion in humans. On the other hand, in mandible, linear and angular measurements related to the mandibular height and mandibular length were bigger in OVX group. These growth patterns in humans could lead to a mandibular protrusion (in the sagittal plane) and also a brachyfacial biotype tendency.
An explanation for maxillary and mandibular differences could be the fact that estrogen performs two opposite functions in two distinct phases of puberty. During the prepubertal period, the activity of estrogen is systemic in accordance with growth hormone (GH) causing bone elongation [2]. After this period of growth comes the period known as postpubertal, when estrogen acts at the local level causing epiphyseal fusion, resulting in bone maturation [2,33]. Estrogen is needed during bone growth and development for proper closure of epiphyseal growth plates both in females and in males. Also, within the young skeleton, estrogen deficiency leads to increased osteoclast formation and enhanced bone resorption [37]. It is important to emphasize that estrogen plays a role in bone via the RANK/RANKL/ OPG triad.
Our gene expression results and immunohistochemical analysis demonstrated that estrogen deficiency during puberty might affect the expression of RANK, RANKL and OPG in the craniofacial growth sites. Mandibular condylar cartilage is known as the center of most pronounced growth in the mandible and the craniofacial complex, and is associated with morphogenesis of the craniofacial skeleton [38]. The RANKL:OPG ratio was higher in the condyle of the OVX group, which may indicate a higher bone remodeling activity in this group. However, this results should be interpreted with caution. Although gene expression analysis was performed in both groups at the same age, at 45 days-old OVX and sham-operated group could be in a different pubertal stages, once higher levels of estrogen accelerates puberty [2]. It is important to mention that the condylar cartilage is a heterogeneous tissue comprising cells at different stages of chondrogenic maturation. Condylar cartilage is designated as secondary cartilage and differs from primary cartilage in histological organization; modes of proliferation, differentiation and calcification [30]. Chondrocytes also express and produce RANK, RANKL and OPG [39,40]. A study in a rodent model concluded that OPG plays a protective role in the postnatal survival of condylar chondrocytes [41]. A possible limitation of our study is that gene expression was evaluated only in one time-period, which not allowed us to evaluate how the levels of RANK, RANKL and OPG behave throughout craniofacial growth during puberty.
The midpalatal suture is an important growth site in the maxilla. RANK in the midpalatal suture was differentially expressed among groups. Mature osteoclasts have the RANK receptor for their activation. Once RANKL binds to RANK, this activates the process of bone resorption. Although we observed that RANK expression was higher in the sham-operated group, RANKL and OPG expression were not statistically different among the groups in the midpalatal suture. In order to start bone remodeling, RANKL should bind RANK [41]. Therefore, it is possible that the statistical difference observed here does not have a biological impact on bone remodeling.
Briefly, our results using an estrogen-deficient animal model support that estrogen is important for maxillary and mandibular development during puberty and could impact the expression of the RANK/RANKL/OPG system in growth sites of the facial complex. However, the role of high of estrogen levels (hyperestrogenism) in maxillary and mandibular growth, as well as the role of estrogen in contraceptives are still unknown. It is possible that an estrogen excess, which could occur when taking estrogen-based contraceptives during puberty, might have the opposite effect as detected in the OVX animals, that is an inhibiting effect on mandibular and maxillary growth as well as craniofacial growth in general. Further studies should elucidate the effect of high levels of estrogen during facial growth. The levels of RANK, RANKL and OPG in maxillary and mandibular growth sites during different periods of puberty should also be investigated.

Conclusions
Estrogen deficiency during puberty could be involved on maxillary and mandibular dimensions and on RANK/ RANKL/OPG expression at important growth sites of the jaws.