Additional Insights Into the Role of Osteocalcin in Osteoblast Differentiation and in the Early Steps of Developing Alveolar Process of Rat Molars

Abstract

This study investigated whether osteocalcin (OCN) is present in osteoblast precursors and its relationship with initial phases of alveolar process formation. Samples of maxillae of 16-, 18-, and 20-day-old rat embryos (E16, E18, and E20, respectively), and 05-, 10-, and 15-day-old postnatal rats (P05, P10, and P15, respectively) were fixed and embedded in paraffin or araldite. Immunohistochemistry for osterix (Osx), alkaline phosphatase (ALP), and OCN detection was performed and the number of immunolabelled cells was computed. Non-decalcified sections were subjected to the von Kossa method combined with immunohistochemistry for Osx or OCN detection. For OCN immunolocalization, samples were fixed in 0.5% glutaraldehyde/2% formaldehyde and embedded in LR White resin. The highest number of ALP- and OCN-immunolabelled cells was observed in dental follicle of E16 specimens, mainly in basal portions of dental alveolus. In corresponding regions, osteoblasts in differentiation adjacent to von Kossa-positive bone matrix exhibited Osx and OCN immunoreactivity. Ultrastructural analysis revealed OCN immunoreactive particles inside osteoblast in differentiation, and in bone matrix associated with collagen fibrils and within matrix vesicles, at early stages of alveolar process formation. Our results indicate that OCN plays a role in osteoblast differentiation and may regulate calcium/phosphate precipitation during early mineralization of the alveolar process:

Introduction

The neural crest-derived stem cells migrate to the maxillary primordia where they interact with local cells and give rise to ectomesenchymal cells. ¹ A subpopulation of ectomesenchymal cells proliferates and forms a blastema of round-shaped cells, which differentiate into osteoblasts that produce the bone matrix of the alveolar process around the tooth germs. ² Thus, rodent embryo heads containing developing maxillary primordia constitute a suitable in vivo model to investigate the cellular and molecular mechanisms involved in the osteoblast differentiation and in the early stages of bone formation.^3–5

Differentiation of ectomesenchymal cells to mature osteoblasts is a complex process coordinated by several factors, which interact with each other in a spatiotemporal manner.^6,7 Initially, the ectomesenchymal cells expressing Runt-related transcription factor 2 (Runx2) are differentiated into osteoprogenitor cells which, in the next stage of differentiation, express Osterix (Osx) and become preosteoblasts. ⁶ During differentiation of ectomesenchymal cells into osteoblasts, Runx2 stimulates the expression of several factors, including alkaline phosphatase (ALP), type I collagen, bone sialoprotein, and fibronectin. With the progression of differentiation, Runx2 and Osx stimulate the expression of osteocalcin (OCN), osteopontin, and osteonectin.^6,8 Thus, Runx2, Osx, ALP, type I collagen, and bone sialoprotein are considered early markers of osteoblast maturation, while OCN, osteopontin, and osteonectin are considered late markers of osteoblast differentiation.^9,10 In the final stages of maturation, osteoblasts become polarized and cuboid-shaped cells and the bone matrix begins to mineralize. ¹⁰

The process of bone matrix mineralization is regulated by several factors secreted by osteoblasts.^11,12 Some studies have shown that matrix vesicles, which sprout from the surface of osteoblasts,^13,14 are required to initiate the calcium phosphate precipitation in the bone matrix.^13,15,16 Matrix vesicles contain annexin A5 calcium channel, calcium, phosphate, and ALP in their membrane.^17–19 The ALP has a role in the supersaturation of phosphate and calcium ions inside matrix vesicles, resulting in the precipitation of calcium phosphate that give rise to hydroxyapatite crystals. ²⁰ Concurrently, OCN, a glycoprotein also secreted by osteoblasts, plays a crucial role in the maturation of hydroxyapatite crystals and bone matrix mineralization. ²¹ OCN is the most abundant non-collagenous protein in bone tissue, which has traditionally been considered a marker of mature osteoblasts because it is only expressed in differentiated osteoblasts.^9,22,23 Otherwise, OCN expression has been demonstrated in culture of mesenchymal cells as an important early marker of osteoblastic differentiation.^24,25 However, in vivo studies evaluating the OCN immunoexpression during osteoblast differentiation and its relationship with initial phases of bone mineralization have been poorly reported.^26,27

Thus, the purpose of this study was to evaluate whether osteoblasts in differentiation express OCN in the initial phases of formation of the alveolar process surrounding the first molar germs. Furthermore, ultrastructural immunolocalization of OCN in the alveolar process during early stages of development was also carried out. For this purpose, the primordia of the maxillae from 16-, 18-, and 20-day-old rat fetuses and maxillary processes of 5-, 10-, and 15-day-old rats were used to clarify some cellular events of the complex cascade involved in the early bone formation.

Materials and Methods

Experimental Design

All experimental procedures were carried out in accordance with national and international guidelines for laboratory animal care. This study was approved by the Ethical Committee for Animal Research of São Paulo Federal University (UNIFESP—protocol number 1022/11). Virgin female and male Wistar rats (Rattus norvegicus albinus) with 3 months of age and body weight of ±200 g were purchased from the Center for the Development of Animal Models for Biology and Medicine (CEDEME—UNIFESP). The rats were housed in propylene cages filled with pine shavings and maintained in a room with a standard 12-hr light/dark cycle, under controlled temperature (23 ± 2C) and humidity (55 ± 10%), and with water and food provided ad libitum.

After 1 week of adaptation, the rats were left to mate between 7:00 pm to 7:00 am. A vaginal collection was performed every morning (7:00 am) to observe the presence of sperm, which indicates the day 0 of pregnancy. ²⁸ Female rats were then euthanized on days 16 (E16), 18 (E18), and 20 (E20) of pregnancy by anesthetic overdose with a single intraperitoneal injection of a solution containing ketamine hydrochloride (300 mg/kg of body weight; Virbac do Brasil Indústria e Comércio Ltda) and xylazine hydrochloride (30 mg/kg of body weight, União Química Farmacêutica Nacional S/A).

Immediately after euthanasia, the abdominal cavity and the uterine horns of rats were carefully opened, and the embryos were immediately removed and euthanized by anesthetic overdose, as previously described. An average of 12 embryos was recorded in each gestation. Other pregnant rats conceived their offspring, and the postnatal male rats at 5 (P05), 10 (P10), and 15 (P15) days of age were also euthanized by anesthetic overdose. The head samples of embryos and postnatal rats were immersed in fixative solutions, as described ahead.

Sample Processing

Six samples for each period (E16, E18, E20, P05, P10, and P15) were fixed for 48 hr in a solution of 4% formaldehyde (freshly prepared from paraformaldehyde) buffered with 0.1 M sodium phosphate at pH 7.4 at room temperature (RT). Afterward, the samples were decalcified in a solution of 7% ethylenediaminetetraacetic acid (EDTA) containing 0.5% formaldehyde, buffered at pH 7.2 with 0.1-M sodium phosphate. Thereafter, the samples were dehydrated in graded concentrations of ethanol, cleared in xylene, and embedded in paraffin. Nine samples of embryo heads of E16, E18, and E20 (three specimens per period) were fixed but they were not decalcified; these samples were equally processed for paraffin embedding. Frontal 5-µm-thick sections were obtained from each head specimen with the aid of a rotary microtome (Microm, HM 2035) and collected on glass slides.

Frontal semi-serial sections from each sample were carefully selected using a light microscope (Axio Lab.A1; Carl Zeiss, Germany) at ×100 and ×400 magnification. Sections containing the developing alveolar process surrounding the teeth germs at different phases of odontogenesis—bud, cap, bell, and formation of the crown and root, from specimens of E16, E18, E20, P05, P10, and P15, respectively, were selected and determined as the regions of interest (ROIs) (Fig. 1A to F). For morphological analysis of the ROI, the sections were stained with hematoxylin and eosin (H&E).

Figure 1.

(A–F) Light micrographs of H&E-stained sections of embryo heads at 16 (E16), 18 (E18), and 20 (E20) days and heads of 5-(P05), 10-(P10), and 15-day-old (P15) rats. Developing alveolar process (AP) is observed surrounding the first molar germs (G). Scale bars: 75 µm. Abbreviations: BM, bone marrow; DF, dental follicle.

Histochemical Methods

Von Kossa Method

To evaluate the initial mineral precipitation on the forming bone matrix, sections of undecalcified samples (E16, E18, and E20 phases) were submitted to the von Kossa method. ²⁹ After deparaffinization and hydration, the sections were immersed in a 5% silver nitrate solution for 1 hr under a 100-watt incandescent lamp. Subsequently, the sections were washed in distilled water (dH₂O) and incubated in 5% sodium thiosulfate solution. Sections were then washed in dH₂O and counterstained with Carazzi’s hematoxylin. Afterward, the sections were dehydrated in ethanol, cleared in xylene, and mounted with Entellan mounting medium.

Gomori Histochemical Method for ALP Detection

Undecalcified sections of embryo head samples were subjected to the Gomori histochemical method to evaluate the enzymatic activity of ALP at the early steps of developing the alveolar process. ³⁰ After deparaffinization and hydration, the sections were washed in dH₂O and incubated for 2 hr at 37°C in a substrate solution (10 ml of 3% sodium β-glycerophosphate, 25 ml of 2% calcium chloride, 10 drops of magnesium chloride at 10%, and 1 g of sodium barbiturate, at pH 9.0). Thereafter, the sections were washed in dH₂O and incubated in 2% cobalt nitrate solution for 5 min. After washes in dH₂O, the reaction was revealed incubating the sections in 0.2% ammonium sulfide for 2 min and then mounted in glycerin. As negative control, some sections were incubated in the substrate solution without sodium β-glycerophosphate.

Combined Histochemical and Immunohistochemical Reactions

Some undecalcified sections of embryo heads (E16, E18, and E20 phases) were subjected to the von Kossa method followed by the immunohistochemical detection of Osx or OCN. Briefly, deparaffinized and hydrated sections were submitted to the von Kossa method as described above (Histochemical Methods – von Kossa Method). After incubation in the sodium thiosulfate solution, the sections were washed in dH₂O and incubated in 3% H₂O₂ to block endogenous peroxidase. Subsequently, the sections were incubated for 40 min in 0.001 M sodium citrate buffer solution (pH 6.0) at 90–94°C in a vapor cooker, for antigen retrieval. After washings in phosphate-buffered saline (PBS) solution at pH 7.4 and incubation for 20 min with 2% bovine serum albumin (BSA), the sections were incubated in a humidified chamber at 4°C overnight in rabbit polyclonal anti-Osx antibody (ab22552; Abcam, Cambridge Science, UK, diluted 1:600) or mouse monoclonal anti-OCN antibody (MAB1419; R&D Systems, Minneapolis, MN, diluted 1:600). After washes in PBS, immunoreactivity was detected by the Labeled Streptavidin–Biotin system (LSAB + System-HRP Kit; Dako, Carpinteria, CA). Sections were incubated in a multilink solution containing biotinylated anti-mouse/rabbit antibodies for 30 min, at RT, washed in PBS, and incubated with streptavidin–peroxidase complex. Peroxidase activity was revealed by 3,3′-diaminobenzidine (DAB) (DAKO Corporation, USA) and the sections were counterstained with Carazzi’s hematoxylin. Then, the sections were dehydrated in ethanol, cleared in xylene, and mounted with Entellan mounting medium. As negative control, sections were subjected to the von Kossa reaction and subsequently, the sections underwent the same steps of immunohistochemical reactions, except that the sections were incubated with non-immune serum replacing the primary antibodies (anti-Osx and anti-OCN).

Immunostaining Reactions

Deparaffinized sections were hydrated and incubated for 20 min in 3% H₂O₂ to block endogenous peroxidase. Subsequently, sections were incubated for 40 min in 0.001 M sodium citrate buffer solution (pH 6.0) at 90–94°C in a vapor cooker, for antigen retrieval. Sections were then washed in PBS at pH 7.4 and incubated for 20 min with 2% BSA. Afterward, sections were incubated in a humidified chamber at 4°C overnight in each of the following primary antibodies: mouse monoclonal anti-OCN antibody (MAB1419, R&D Systems, diluted 1:600) and rabbit polyclonal anti-ALP antibody (sc-30203; Santa Cruz Biotechnology, Dallas, TX, diluted 1:1000). After washes in PBS, immunoreactivity was detected by the Labeled Streptavidin–Biotin system (LSAB + System-HRP Kit; Dako). Sections were incubated in a multilink solution containing biotinylated anti-mouse/rabbit secondary antibodies for 30 min at RT, washed in PBS, and incubated with streptavidin–peroxidase complex. Peroxidase activity was revealed by DAB (DAKO Corporation) and sections were counterstained with Carazzi’s hematoxylin. As negative controls of all immunoreactions, the step of incubation in primary antibodies was replaced by incubating sections in non-immune serum (Sigma-Aldrich, Germany).

Numerical Density of OCN- and ALP-Immunostained Cells

The number of OCN- and ALP-immunostained cells per mm² was computed in the six samples of each period (E16, E18, E20, P05, P10, and P15). In each sample, two non-serial sections with a minimum distance of 100 μm between sections were selected. In each section, three standardized fields (0.020 mm² each) in regions containing the developing alveolar process surrounding the first molar germ were captured. Images were captured using a light microscope (Axiolab 2.0; Carl Zeiss) at ×1745 magnification attached to a camera (AxioCam ICc5; Carl Zeiss). Considering that the AxioVision image analysis system used in this study does not discriminate shapes and cellular geometry, the number of OCN- and ALP-immunolabelled cells was performed by a calibrated and blinded examiner. Using the image analysis system (AxioVision 4.2 REL; Carl Zeiss), each image previously captured was uploaded; with the help of the image analysis system, the examiner selected each cell that exhibited brown-yellow cytoplasm (OCN-immunolabelled cells or ALP-immunolabelled cells). At the end of the count, the image analysis system provided the total number of cells computed by the examiner in each image (standardized field). The number of immunopositive cells/mm² was then estimated.

Statistical Analysis

Quantitative data were evaluated using the GraphPad Prism 5 software (GraphPad Software, Inc.; San Diego, CA). Data were expressed as mean ± SD. Differences among groups were analyzed by one-way analysis of variance, followed by Tukey’s multiple comparison tests. A value of p≤0.05 was considered statistically significant.

Ultrastructural Analysis

Three heads from each period (E16, E18, E20, P05, P10, and P20) were used for ultrastructural analyses. Fragments of maxilla containing the ROI were fixed for 16–20 hr in a solution containing 4% glutaraldehyde and 4% formaldehyde (freshly prepared from paraformaldehyde) buffered at pH 7.2 with 0.1 M sodium cacodylate. Then, samples from E18, E20, P05, P10, and P15 periods were decalcified for 3, 5, 7, 14, and 21 days, respectively, in 7% EDTA buffered with 0.1 M sodium cacodylate at pH 7.2. Subsequently, the samples were washed in 0.1 M sodium cacodylate buffer (pH 7.2) and treated with 1% osmium tetroxide for 1 hr at RT. After washing in dH₂O, the samples were stained in aqueous solution of 2% uranyl acetate for 2 hr. Afterward, the samples were dehydrated in graded ethanol, treated with propylene oxide, and then embedded in Araldite resin. Using an ultramicrotome (Ultracut UCT; Leica), semithin sections (600–800 nm) stained in aqueous solution of 1% toluidine blue were examined under a light microscope for selection of areas of interest. Ultrathin sections (80 nm) were collected on copper grids and stained in alcoholic 2% uranyl acetate and in lead citrate. Ultrathin sections were analyzed using FEI transmission electron microscope (TEM; TECNAI model) of the Biosciences Institute, Botucatu-UNESP.

OCN Immunolocalization in Ultrathin Sections

For ultrastructural immunolocalization of OCN, 18 samples from all periods (three heads per period) were processed for embedding in acrylic resin (LR White). Small fragments of maxilla containing first molars were fixed with 0.5% glutaraldehyde and 2% formaldehyde solution (freshly prepared from paraformaldehyde) buffered at pH 7.2 with 0.1 M sodium cacodylate for 12 hr. Thereafter, the samples were decalcified as previously described for samples embedded in Araldite. After washings in dH₂O, the samples were dehydrated in graded concentrations of ethanol at 4°C (30° GL, 50° GL, and 70° GL). After 1 hr immersed in 70° GL ethanol, the in bloc contrast was performed using 70° alcoholic solution of 2% uranyl acetate at 4°C for 16–18 hr. Subsequently, the samples were dehydrated and embedded in acrylic resin (LR White Embedding Resin Kit; Electron Microscopy Sciences, London, England), which was polymerized at 4°C under ultraviolet light. Semithin sections were obtained and stained with 1% toluidine blue for selection of the ROI, that is, early stages of alveolar process formation. The ultrathin sections were obtained and collected onto parlodion-coated 300 mesh nickel grids.

The grids with ultrathin sections were incubated for 10 min in the following solutions: (1) PBS with 0.2% Triton, (2) PBS containing with 5% BSA, and 0.1% fish jelly, and (3) PBS. Then, the sections were incubated for 12–16 hr in a humid chamber at 4°C with a rabbit monoclonal anti-OCN antibody (Sigma), diluted 1:50 in PBS containing 1% BSA. After washing and incubation in 2% BSA, the sections were incubated with anti-rabbit antibody conjugated to 10-nm colloidal gold particles (Goat-anti-rabbit/Electron Microscopy Sciences), diluted 1:100, for 2 hr at RT. Afterward, the sections were stained in 2% osmium tetroxide, 2% uranyl acetate, and lead citrate. For specificity control, ultrathin sections adhered to the nickel grids underwent the same steps, except that the sections were incubated with non-immune serum replacing the primary antibody. The ultrathin sections were analyzed in a TEM (FEI Company, TECNAI model).

Results

Histological and Histochemical Evaluations of the ROI

The morphological analysis revealed that the formation of the alveolar process surrounding the upper first molar germs was concomitant with the development of the tooth germ (Fig. 1A to F). In the E16 specimens, the ectomesenchyme in the corresponding region of the future alveolar process exhibited some large and polarized cells; newly synthesized thin aggregates of bone matrix were observed among osteoblasts in differentiation (Figs. 1A and 2A). In the E18 and E20 specimens, bone spicules of the developing alveolar process surrounded almost completely the molar tooth germs (Fig. 1B and C). In these specimens, osteoblasts and osteoclasts on the bone surface as well as newly embedded osteocytes were observed (Fig. 2B and C). In the P05, P10, and P15 specimens, the alveolar process showed a network of interconnected trabeculae and large spaces occupied by bone marrow (Fig. 1D to F).

Figure 2.

(A–L) Light micrographs of frontal sections of embryo heads at 16 (E16), 18 (E18), and 20 (E20) days of age showing portions of maxillae stained with H&E (A–C), subjected to the Gomori (D–F) and von Kossa methods (G–I). (A–C) Osteoblasts (Ob), osteocytes (Ot), osteoclasts (Oc), and newly synthesized bone matrix (BM) of the developing alveolar process are seen. (D–F) ALP activity is observed in the periphery of some Ob [inset of (D) and (F)] and BM. (G–I) von Kossa method showing the BM of the developing alveolar process in the early stage of mineralization in E16 (G), and at a more advanced stage in E18 (H) and E20 (I) specimens. Scale bars: 31 µm (A and B), 93 µm (C), 56 µm (D–F), 35 µm (G), 40 µm (H), 46 µm (I), 15 µm (insets). Abbreviations: ALP, alkaline phosphatase; BV, blood vessel.

Undecalcified sections of embryo heads subjected to the histochemical method for detection of ALP showed strong positivity (black/dark brown color) in the bone matrix during the early formation of the alveolar process (Fig. 2D to F). Evident ALP activity was also observed in the peripheral portion of the osteoblast cytoplasm (Fig. 2D and F). The developing alveolar process of E16, E18, and E20 specimens showed a varied pattern in the von Kossa reaction (Fig. 2G to I). The embryo heads at 16 days revealed von Kossa-positive granular deposits (in yellow color) on the thin bone spicules (Fig. 2G), whereas in E18 and E20 specimens, the developing alveolar process showed bone spicules with brown/black-stained areas (Fig. 2H and I). Note that von Kossa-positive staining was not observed in the osteoblasts near the developing alveolar process (Fig. 2G to I).

Combined von Kossa Method and Immunohistochemical Detection of OCN or Osx

Undecalcified sections of embryo heads revealed thin bone spicules weakly stained in yellow color by von Kossa reaction in the E16 and E18 specimens (Fig. 3A, B, D, and E). In the E20 specimens, irregular bone trabeculae exhibited strong positivity to the von Kossa, in brown color (Fig. 3C and F). von Kossa-positive granulation was seen in continuity with the newly formed bone spicules in the E16, E18, and E20 specimens (Fig. 3A, C, E, and F). In E16 and E18 specimens, several cells with Osx-immunostained nucleus were observed next to the developing bone spicules (Fig. 3A and B). Osteoblasts exhibiting Osx-immunolabelled nucleus were observed adjacent to the developing bone trabeculae in E20 specimens (Fig. 3C). Large cells showing strong OCN immunolabelling in the cytoplasm were located near thin bone spicules with varied positivity to the von Kossa reaction in E16, E18, and E20 specimens (Fig. 3D to F). As osteoblasts were not stained by von Kossa (Fig. 2G to I), brown-yellow color observed in the nucleus (Fig. 3A to C) and in the cytoplasm (Fig. 3D to F) of osteoblasts indicates the immunoreactivity to the Osx and OCN, respectively. Moreover, Osx- and OCN-immunolabelled osteoblasts were not observed in the sections used as negative controls (data not shown).

Figure 3.

(A–F) Light micrographs showing portions of non-decalcified heads of 16-, 18- and 20-day-old rat fetuses. The sections were subjected to the von Kossa histochemical method followed by immunohistochemistry for the detection of Osx (A–C) and OCN (D–F). (A–C) Several cells exhibiting Osx-immunopositive nucleus (arrowheads) are observed in the bone blastema (B) of E16 and E18 specimens. In these specimens (A and B), the thin bone spicules (asterisks) are weakly stained by von Kossa (in yellow color). In the E20 specimen (C), osteoblast nucleus exhibiting strong Osx immunolabelling (arrowheads) are juxtaposed to bone spicules (asterisks) exhibiting strong positivity to the von Kossa reaction (brown color). von Kossa-positive granular deposits (arrows) are seen in some regions of bone spicules in formation. (D–F) Polarized osteoblasts (arrowheads) showing strong cytoplasmic OCN immunolabelling are seen juxtaposed to bone spicules (asterisks). The bone spicules (asterisks) exhibit varied positivity to the von Kossa reaction: dark brown/black staining and regions exhibiting von Kossa-positive granular material (arrows). von Kossa-negative portions of bone spicules (black asterisks) are also observed. Scale bars: 30 µm. Abbreviations: BV, blood vessel; OCN, osteocalcin; Osx, osterix.

Immunohistochemical Detection of ALP and OCN

In the sections subjected to immunohistochemistry for detection of ALP, numerous immunolabelled cells were detected in the dental follicle of rat fetuses at 16 and 18 days (Fig. 4A and B). Moreover, a strong immunolabelling for ALP was seen in the bone blastema, that is, the region corresponding to the future alveolar process (Fig. 4A). In the E20, P05, P10, and P15 specimens, the ALP immunoexpression was detected mainly in the osteoblasts adjacent to the bone surface, whereas few immunolabelled cells were seen in the dental follicle (Fig. 4C to E) and in the developing periodontal ligament (Fig. 4F). An accentuated OCN immunoexpression was noted in regions of the bone blastema of alveolar process in rat fetuses at 16 days (Fig. 5A) and near to the bone spicules of developing alveolar process of E18 (Fig. 5B), E20 (Fig. 5C), P05 (Fig. 5D), P10 (Fig. 5E), and P15 (Fig. 5F) specimens.

Figure 4.

(A–F) Light micrographs of portions of heads from rat fetuses (A–C) and 5-, 10-, and 15-day-old rats (D–F). The sections were subjected to immunohistochemistry for detection of ALP and counterstained with hematoxylin. In (A), ALP-immunolabelled cells (arrows) adjacent to an amorphous material (asterisks) are seen in the bone blastema in the 16-day-old embryo. (B) ALP immunolabelling is observed in cells of dental follicle (DF) and in differentiating osteoblasts (Ob) of the developing alveolar process (AP). (C–F) Strong ALP immunolabelling is observed in the Ob adjacent to the bone spicules in the E20, P05, and P10 specimens in comparison with the weak reaction observed in P15 specimen. Strong ALP immunolabelling is observed in the Ob cytoplasm [inset of (E)]. In the 15-day-old rat (F), a few ALP-immunolabelled Ob are observed on the alveolar bone surface. Scale bars: 35 and 11 µm [inset in (E)]. Abbreviations: ALP, alkaline phosphatase; BM, bone marrow; De, dentine; G, tooth germ; PL, developing periodontal ligament.

Figure 5.

(A–F) Light micrographs of portions of heads from rat fetuses (A–C) and 5-, 10-, and 15-day-old rats (D–F). The sections were subjected to immunohistochemistry for detection of OCN and counterstained with hematoxylin. (A) OCN immunolabelling is observed in ectomesenchymal cells (EC) around the tooth germ (G). In the region corresponding to the bone blastema (asterisk), osteoblasts (Ob) in differentiation exhibit evident OCN immunoreactivity (arrows). (B–F) OCN immunolabelling is observed in Ob on the surface of the developing alveolar process (AP). Scale bars: 32 µm. Abbreviations: BV, blood vessel; DF, dental follicle; OCN, osteocalcin; PL, developing periodontal ligament.

As shown in Fig. 6A, the greatest number of ALP-immunolabelled cells was observed in the E16 specimens. A gradual reduction in the number of ALP-immunolabelled cells was seen over time, except in the P15. In the developing periodontal ligament of 15-day-old rats, the number of ALP-immunostained cells was greater than that in the 10-day-old rats (p<0.0001). According to Fig. 6B, the number of OCN-immunolabelled cells in the dental follicle was significantly elevated in E16 specimens in comparison with E18 (p<0.01), E20, P05, and P10 (p<0.0001), but no significant difference was observed between E16 and P15. On the other hand, a low OCN immunoexpression was observed in the dental follicle of E20 and P05 specimens, and a significant difference was not verified between these groups.

Figure 6.

(A and B) Number of ALP-immunolabelled (A) and OCN-immunolabelled (B) cells (mean ± SD) in cells of dental follicle/developing periodontal ligament of rat fetuses at 16, 18, and 20 days (E16, E18, and E20) and 5-, 10-, and 15-day-old rats (P05, P10, and P15). Superscript asterisks indicate significant differences between the groups. ANOVA/Tukey’s test. Abbreviations: ALP, alkaline phosphatase; OCN, osteocalcin. *p<0.05, **p<0.01, ****p<0.0001.

Ultrastructural Analysis and OCN Immunolocalization

The examination of the region of the developing alveolar process at 16 days by TEM revealed osteoblasts in differentiation juxtaposed to a thin bone matrix containing sparse collagen fibrils. Round or ovoid structures delimited by plasma membrane and showing heterogeneous electron opacity, identified as matrix vesicles, were observed among collagen fibrils in the developing bone matrix (Fig. 7A to C). The ultrathin sections subjected to the immunolocalization of OCN revealed gold particles in the bone matrix as well as in osteoblasts in differentiation during the early stages of bone formation (Fig. 8A to E). Gold particles indicating OCN immunoreactivity were observed inside the rough endoplasmic reticulum cisternae of osteoblasts (Fig. 8B). In the early stage of deposition of bone matrix, gold particles were often associated with collagen fibrils; OCN immunoreactive particles apparently free in the bone matrix and within the matrix vesicles were also observed (Fig. 8C to E). In these vesicles, gold particles were often associated with electron-opaque material (Fig. 8E). Gold particles were not found in the ultrathin sections used as negative controls (Fig. 8F).

Figure 7.

(A–C) Electron micrographs showing a portion of bone matrix (BM) in an early stage of formation from rat fetus at 16 days (Araldite-embedded specimen). Osteoblasts (Ob) in differentiation are juxtaposed to developing BM. (B) The outlined area of (A) shows small vesicles (Ve) in close proximity to the plasma membrane of Ob. Matrix vesicles (MV) are present among collagen fibrils (Cf) in the BM. (C) MVs with varied electron opacity; some Cf (arrows) are observed in close association with MVs. Scale bars: 2 µm (A) and 0.25 µm (B and C). Abbreviations: P, osteoblast process; rER, rough endoplasmic reticulum.

Figure 8.

(A–F) Electron micrographs showing portions of bone matrix (BM) in early stages of formation from rat fetuses at 16 (A–D and F) and 20 (E) days. Sections of LR White-embedded specimens subjected to the immunolocalization of osteocalcin (OCN). (A) Portion of an osteoblast (Ob) adjacent to newly formed BM. Several matrix vesicles (MV) with round shape and varied electron opacity are situated among collagen fibrils (Cf) in the BM. (B) Higher magnification view of Ob showing gold particles (arrows) inside rough endoplasmic reticulum (rER) cisternae. (C and D) Higher magnification of (A) showing Cf and MV in the BM. Gold particles indicating OCN immunoreactivity (arrows) are observed in the BM; some gold particles are in close association with Cf. OCN immunoreactivity (white arrows) is also observed within an MV. Large white arrows, plasma membrane surrounding partially the MV. (E) Several gold particles within MV, exhibiting varied electron opacity, are seen in the developing bone of a rat fetus at 20 days. (F) Negative control showing a portion of neoformed BM. Note absence of gold particles within MV or associated with Cf. Scale bars: 2 µm (A), 0.1 µm (B, D, E), 0.2 µm (C and F).

In 5-, 10-, and 15-day-old rats, the alveolar process in advanced stages of development revealed rich collagen bone spicules at different mineralization phases (Fig. 9A to D). Large osteoblasts exhibiting several profiles of rough endoplasmic reticulum were juxtaposed to a dense bone matrix (Fig. 9A and B). An osteoid layer with randomly arranged collagen fibrils was present between osteoblasts and the mineralized bone matrix (Fig. 9B and D). Several gold particles indicating OCN immunoreactivity were usually seen in association with collagen fibrils in the osteoid, whereas a few gold particles were found in the mineralized matrix (Fig. 9C).

Figure 9.

(A–D) Electron micrographs of portions of developing bone matrix (BM) of 5- (A), 10- (B and C), and 15-day-old (D) rats. Sections of Araldite-embedded specimens (A and D) and sections of LR White-embedded specimens (B and C). (A) Polarized osteoblasts (Ob) containing several profiles of rough endoplasmic reticulum (rER) are adjacent to mineralizing BM. The inset, outlined area in (A), shows a matrix vesicle (MV) in the BM. (B and C) Sections subjected to the immunolocalization of OCN. An Ob adjacent to the non-mineralized BM in formation (asterisks). Higher magnification of non-mineralized matrix (C), gold particles (arrows) indicating OCN immunoreactivity are observed mainly in close association with collagen fibrils (Cf). Few gold particles (white arrows) are seen in the BM in the advanced stage of mineralization. In (D), Ob are juxtaposed to the BM in the advanced stage of development of the alveolar process. Scale bars: 2.5 µm (A and D), 1 µm (B), 0.1 µm (C), 0.4 µm [inset of (A)].

Discussion

The findings of this study clearly showed that osteoblasts in differentiation produce and release OCN, which may have a pivotal role in the early stages of bone matrix formation of the alveolar process in rat molars. Furthermore, OCN may be involved in the initial mineralization process because this glycoprotein was immunolocalized within matrix vesicles during the initial stage of alveolar process formation in rats.

Stem cells derived from the neural crest migrate to the maxillary primordia, where they interact with local tissue and differentiate into ectomesenchymal cells. ¹ A subpopulation of ectomesenchymal cells proliferates and form a blastema with round-shaped cells and, then, differentiate into osteoblasts that produce the bone matrix of the alveolar process. ² Here, the initial formation of alveolar process was recognized by thin bone spicules found in the base of the dental alveolus. ALP-positive osteoblasts in differentiation were adjacent to the bone spicules, which also exhibited strong ALP positivity. ALP is a specific glycoprotein found on the osteoblast surface and plays a key role in bone matrix mineralization.^31–33 An enhanced ALP immunolabelling was also detected in the dental follicle cells in rat fetuses at 16 and 18 days in comparison with other time points evaluated. These findings confirm the concept that ALP constitutes an important marker of preosteoblasts and this enzyme participates in the initial osteogenesis of the alveolar process. ³²

Thus, the next step of the study was to evaluate the mineralization in the ROI. Therefore, undecalcified sections of the embryo heads submitted to the von Kossa method revealed a varied degree of mineralization because the bone matrix of the E16 specimens was weakly stained in yellow. This suggests the presence of calcium deposits on proteins previously produced by ALP-positive osteoblasts. On the other hand, the bone spicules presented calcium possibly associated with phosphate inorganic deposits in brown/dark color at 18- and 20-day-old rat fetuses, which point to a more advanced stage of bone matrix mineralization. ³⁴ Therefore, these findings indicate a greater secretory phenotype of osteoblasts in the E18 and E20 specimens than in the E16 specimens, as previously described. ³

The differentiation of osteoblast is a complex process in which several factors interact in a coordinated spatiotemporal manner.^6,7 Among several factors involved in osteoblastic differentiation, the Osx is a transcription factor that stimulates preosteoblasts to go through maturation stages until they become mature osteoblasts.^6–8,10 Thus, the next step of our study was to identify the osteoblast precursor cells in the ROI. For this purpose, undecalcified sections of E16, E18, and E20 specimens were subjected to the von Kossa method followed by immunohistochemistry for Osx or OCN detection. Although the von Kossa histochemical reaction was combined with immunohistochemistry detection using DAB chromogen, which provides a reaction product in brown-yellow color similar to the von Kossa-positive deposits, it is important to empathize that combined methods in this study were useful because no positivity for the von Kossa reaction is seen in osteoblasts. In these specimens, osteoblasts with strong nuclear Osx immunolabelling were juxtaposed to von Kossa-positive bone spicules around tooth germs. Furthermore, other sections revealed that osteoblasts in differentiation, situated in corresponding regions to those immunolabelled for Osx, showed evident cytoplasmic OCN immunolabelling. Therefore, these data indicate that osteoblasts in differentiation are able to secrete OCN in the early stages of the alveolar process development.

On the other hand, OCN immunolabelling was also observed in some dental follicle cells, mainly in the ectomesenchyme near the developing alveolar process. As shown in this study, in rat fetuses at 16 days, a population of ectomesenchymal cells begins to differentiate into osteoblasts, which start the deposition of bone matrix in the base of the dental alveolus. In subsequent periods, there is a gradual decrease in the OCN immunoexpression until the period of 5 days after birth. The increase in OCN immunoexpression observed in 10- and 15-day-old rats coincides with the beginning of root formation of first upper molar and, consequently, with the onset of alveolar bone formation.^35,36 Therefore, the increase in the OCN production may be involved in the initial stage of bone matrix formation in E16 and PN15 specimens.

There is evidence that Osx stimulates the expression of OCN, osteopontin, and osteonectin, which culminates in the advance of the mineralization process.^6,8 In the final stages of mineralization, the organic fraction of bone matrix contains high concentration of OCN. ¹⁰ In fact, OCN is the most abundant non-collagenous protein in the bone matrix, which together with ALP, regulates the process of bone matrix mineralization.^20,22,23 Our findings indicate that ALP may have an important participation in the differentiation of ectomesenchymal cells into osteoblasts because a great ALP immunolabelling was observed in the maxillae of embryos at 16 and 18 days, that is, the early phase of synthesis of bone matrix of the alveolar process. The highest values of OCN were also associated with the initial stages of bone matrix formation (E16 and E18 specimens) as well as with the growth of the alveolar process (P15 days) in parallel to the development of the germ molars. Thus, the increase in the OCN immunoexpression in P15 specimens may be associated with an intense differentiation of osteoblasts required for the development of the alveolar process of the first upper molars. Therefore, these findings indicate that OCN is produced and released by osteoblasts in differentiation during the initial deposition of bone matrix.

Ultrastructural analysis of the regions corresponding to the initial formation of the alveolar process of rat fetuses (E16, E18, and E20) revealed large and polarized osteoblasts in differentiation juxtaposed to the bone matrix containing a few irregularly arranged collagen fibrils and several matrix vesicles. Moreover, the immunolocalization of OCN inside well-developed rough endoplasmic reticulum cisternae and in the bone matrix confirms that this glycoprotein is produced and released by osteoblasts in differentiation. Before secretion by osteoblasts, OCN is carboxylated on its glutamate residues in the rough endoplasmic reticulum by c-glutamyl carboxylase, providing an increased affinity of OCN to calcium and hydroxyapatite crystals.^37,38 In bone matrix, the gold particles indicating OCN immunoreactivity were associated with collagen fibrils, whereas some of them were apparently free among the extracellular bone matrix components. Gold particles were also observed within matrix vesicles in the initial phases of bone formation.

Matrix vesicles with 50–200 nm in diameter sprout from the surface of osteoblasts and remain dispersed among other organic components of bone matrix. It is well known that matrix vesicles are rich in ALP.^16,20,39 This enzyme cleaves pyrophosphate, thus leading to the increase in the inorganic phosphate and subsequent supersaturation inside matrix vesicles.^32,40,41 Phosphatidylserine-bound annexin in membrane increases the permeability of matrix vesicles to calcium. As calcium ions flow into the matrix vesicles, they interact with phosphate ions, forming calcium phosphate, which crystallizes and form hydroxyapatite crystals. After the crystal growth from the vesicle lumen and rupture of the vesicle membrane, the crystals spread through the newly synthesized bone matrix, favoring its mineralization.^33,42 In addition to ALP, the presence of non-collagenous proteins including bone sialoprotein, osteopontin, and OCN was detected by Western blot analyses in matrix vesicles from chondrocytes of rat growth plate. ⁴³ Here, our findings showed, for the first time, the ultrastructural immunolocalization of OCN during the early stage of bone formation of the alveolar process in rat molars. It is conceivable to suggest that OCN is accumulated in the surface of the osteoblast and, subsequently, the matrix vesicles containing OCN are released by osteoblasts. The presence of OCN within matrix vesicles reinforces the concept that this structure plays an important role in the early process of mineralization of the bone matrix. In the advanced stages of the alveolar process formation in postnatal rats, few gold particles were dispersed in the unmineralized bone matrix. However, we cannot exclude the possibility that a great amount of OCN along with other non-collagenous proteins could have been extracted during the decalcification of these specimens. In addition to extraction of non-collagenous proteins, ultrastructural analysis of decalcified specimens made it impossible to examine the interactions between hydroxyapatite and OCN.

Although studies have shown that OCN promotes bone matrix mineralization,^10,44,45 in vitro and in vivo studies also indicate that OCN could act as an inhibitor of bone mineralization.^21,46,47 Although it is not yet known whether the inhibitory action of OCN on the bone matrix mineralization is attributed to its carboxylated or undercarboxylated form, a recent clinical study has shown a correlation between increased serum levels of undercarboxylated OCN and a reduction in bone mineral density, both in men and in women. ⁴⁸ Therefore, the exact role of OCN during bone matrix mineralization remains controversial. Nevertheless, the results of our study support the idea that carboxylated OCN stimulates bone matrix mineralization, at least during the earliest stages of alveolar process formation in rat maxillae.

Meanwhile, it has been reported that OCN may serve as an early marker of in vitro osteogenic differentiation of mesenchymal stem cells.^24,25 In addition to the role of OCN in the maturation of hydroxyapatite crystals, evidence suggests that OCN downregulation reduces the expression of osteoblast differentiation markers, including Runx2 and ALP during mesenchymal stem cell differentiation. ²⁵ In this context, our results corroborate those studies and provide in vivo evidence that OCN participates in the control of osteoblast differentiation.

In conclusion, our results show that OCN is expressed not only by differentiated osteoblasts but also by their precursors, suggesting a possible role of this glycoprotein in the osteoblast differentiation. Moreover, OCN may regulate the nucleation of hydroxyapatite crystals inside matrix vesicles indicating that OCN may play a pivotal role in the beginning of the bone matrix mineralization of alveolar process. Thus, this study contributes to a better understanding of the mechanisms involved in osteoblast differentiation and bone formation, providing new insights into the pathogenesis of craniofacial malformations involving disturbances during intramembranous ossification.