Axial and appendicular bones, such as the femur, are formed by the central process of endochondral ossification, which occurs through progressive steps including chondrocyte proliferation, hypertrophy, and eventual replacement of the cartilaginous template with calcified bone matrix (Olsen et al., 2000). Bone growth clearly involves processes and signals originating and coordinated in the perichondrium and translated in the growth plate (Colnot et al., 2004). At the chondro-osseous junction, vascularization of the chondrocytic extracellular matrix leads to dramatic transformation of the tissue to induce osteogenesis. In this context, it is intriguing that the administration of voltage-sensitive calcium channel (VSCC) blockers, such as nifedipine or verapamil, during pregnancy is associated with increased occurrence of fetal limb defects in humans (http://www.reprotox.org). Further, reports indicate that such VSCC blockers also can inhibit osteogenesis and produce vertebral defects, lower mineral apposition rates, and interfere with bone formation in animal models (Duriez et al., 1993; Ridings et al., 1996; Li et al., 2003). Despite these observations, few investigations have addressed the expression of VSCCs during bone development, and none have examined both high (L-type) and low threshold (T-type) VSCCs.
VSCCs are key regulators of intracellular Ca2+ homeostasis and control plasma membrane Ca2+ permeability in osteoblasts (Duncan et al., 1998). L-type and T-type VSCCs each display unique electrophysiological and pharmacological characteristics (Catterall, 2000; Ertel et al., 2000). In osteoblasts, L-type VSCCs are composed of an α1 subunit and the auxiliary α2δ and β subunits, but devoid of a γ subunit (Bergh et al., 2003). The α1 subunit forms the membrane ion translocation pore, and it is the site for the binding of most channel modulators (Duncan et al., 1998; Catterall, 2000). The Cav1.1 (α1S), Cav1.2 (α1C), Cav1.3 (α1D), and Cav1.4 (α1F) genes comprise the L-type VSCCs (Lipscombe et al., 2004). Among them, Cav1.2 (α1C) is the most abundant L-type VSCC in growing osteoblasts, and also the primary site for 1,25-dihydroxyvitamin D3 (1,25-D3) stimulated Ca2+ influx across the cell membrane (Caffrey and Farach-Carson, 1989; Meszaros et al., 1996). In addition to the L-type VSCCs, osteoblasts express T-type VSCCs characterized by low voltage sensitivity, and the family includes Cav3.1 (α1G), Cav3.2 (α1H), and Cav3.3 (α1I) (Cribbs et al., 1998; Lee et al., 1999; Perez-Reyes, 2003). Functional data from our laboratory indicated that both the Cav1.2 (α1C) and Cav3.2 (α1H) mRNA and protein levels are modulated by 1,25-D3 (Bergh et al., unpublished data), and mice null for T-type Cav3.2 (α1H) subunits (Chen et al., 2003) have altered skeletal growth (Chen, Campbell, Farach-Carson, unpublished data). Because of the key roles that Cav1.2 (α1C) and Cav3.2 (α1H) play in bone growth, we investigated the expression of these two VSCCs during endochondral bone development and in appropriate in vitro models for osteoblastic and chondrocytic differentiation.
Expression of L-Type Cav1.2 (α1C) and T-Type Cav3.2 (α1H) Subunits in Mouse Femur at Various Developmental Stages
We chose the developing femur as the long bone upon which to focus our studies of growth plate because of its size, shape, and ready accessibility. Hematoxylin & Eosin (H&E) staining was used to visualize structural details of each developmental stage investigated (Fig. 1A–C). Expression and localization of the L-type VSCC Cav1.2 (α1C) first was determined from immunohistochemical studies that were performed using Cav1.2 (α1C) subunit specific polyclonal antibodies. Studies were initiated at day E12.5, when only bone or cartilage rudiments were observed (not shown). At E14.5, mouse femur is composed primarily of cartilage, although mineralization has begun at the mid-shaft of the femurs (Fig. 1D,G, the indentations indicated by arrows). Intense immunoreactivity for Cav1.2 (α1C) subunits was observed along the perichondrium, with additional staining in the proliferative chondrocytic regions of the non-mineralized cartilage at the growth plates (Fig. 1D). Significant osteogenesis was observed at E16.5 (Fig. 1E,H) and bone formation continued as the mineralized bone matrix expanded through E18.5 (Fig. 1F,I) and E21 bones (not shown). At E16.5 (Fig. 1E) and E18.5 (Fig. 1F), Cav1.2 (α1C) subunits were present in regions of rapid growth of long bone, including the perichondrium, periosteum, chondro-osseous junction, and along the surfaces of newly formed trabecular bone. Demineralization of the tissues did not change the immunostaining pattern (not shown). The subunit distribution pattern of E21 (newborn mice) sections was similar in appearance to E18.5 (not shown).
We next investigated the expression and localization of T-type VSCC Cav3.2 (α1H) subunits at different embryonic developmental stages using an antibody that we generated against the II–III cytoplasmic loop of mouse Cav3.2 (α1H) subunit sequence. Interestingly, expression of the Cav3.2 (α1H) subunits commenced simultaneously with the appearance of Cav1.2 (α1C) from E14.5 to E18.5 (Fig. 1G–I), and the distribution of both α1 subunits followed a similar pattern during skeletal growth. The immunostaining with Cav3.2 (α1H) was particularly intense at the site of neovascularization at the chondro-osseous junction (Fig. 1I, arrow). Specificity was clear, as addition of control blocking peptides abolished the specific staining (Fig. 1J,L). Collagen X is an expression marker for hypertrophic chondrocytes (O'Keefe et al., 1994). As shown in Figure 1K, as expected, collagen X was highly expressed in the hypertrophic chondrocyte region of the growth plate on E16.5 femur, an area where immunostaining with antibodies to both L-type Cav1.2 (α1C) and T-type Cav3.2 (α1H) channels was comparatively weak (compare Fig. 1E,H to 1K). No evidence of collagen X expression was found in the proliferating chondrocyte regions or in the bone matrix, providing further confidence that our immunostaining procedures were specific.
Under high magnification (40x oil lens), osteoblasts were observed along the bone surfaces on the vascular side of the chondro-osseous junction in E18.5 embryonic mouse femurs where they stained brightly with antibodies directed against Cav1.2 (α1C) (Fig. 2A,D). The actively growing periosteum that consists of several layers of organized osteogenic cells with differentiating osteoblasts in the inner layer and proliferating pre-osteoblasts in the outer layer also stained brightly (Fig. 2B,E). The Cav1.2 (α1C) subunits also appeared in chondrocytes located in the hypertrophic zone (Fig. 2A,D), as well as in the perichondrium (Fig. 2C,F). At high magnification, staining of osteoblasts, periosteum and perichondrium also was observed with the anti-Cav3.2 (α1H) subunit antibodies (not shown).
Bones from 6-week-old skeletally mature mice next were compared with developing embryonic bones. The calcified bone matrix of the femur was well developed in these sections, and abundant bone lining cells with flattened nuclei were evident as a monolayer on the surfaces of calcified bone (Fig. 3A,D). Cavities containing bone marrow stromal cells also were clearly evident (Fig. 3A,D, arrows). Both the osteoblasts lining the surfaces of spongy bone and the bone marrow stromal cells actively expressed the Cav1.2 (α1C) subunits (Fig. 3A,D). In contrast, the post-mitotic osteocytes embedded deep in the calcified bone matrix exhibited little or no immunoreactivity with the antibodies directed against the L-type Cav1.2 (α1C) subunits, but they were clearly evident with the Syto 13 nuclear counterstain (Fig. 3D). T-type Cav3.2 (α1H) VSCCs were intensely stained in osteoblasts along the periosteal surface of the cortical bone of the diaphysis (Fig. 3B,E). L-type Cav1.2 (α1C) subunits also were observed in the outer layer of those regions of compact bones (not shown). Interestingly, T-type Cav3.2 (α1H) subunit expression was identified readily in osteocytes embedded in the mineralized matrix of cortical bones (indicated by arrows in Fig. 3B,E). Pre-absorption with the peptide used for antibody preparation (peptide block) demonstrated the specificity of the primary antibodies and lack of nonspecific binding (Fig. 3C,F). Thus, while both L-type and T-type channels were present in osteoblasts and lining cells, only T-type channels were observed in osteocytes.
Expression of L-Type Cav1.2 (α1C) and T-Type Cav3.2 (α1H) in Osteoblastic (MC3T3-E1), Chondrogenic (ATDC5) and Osteocyte-Like (MLO-Y4) Cells
We next undertook to complement the in vivo localization study using well-established in vitro murine models of osteogenesis (MC3T3-E1) and chondrogenesis (ATDC5), as well as the murine osteocyte-like MLO-Y4 cell line (Kato et al., 1997). In subconfluent growth phase MC3T3-E1 cells, intense immunostaining of L-type Cav1.2 (α1C) subunits was seen at the membrane of the cell surface and intracellularly in the secretory route (endoplasmic reticulum and Golgi) (Fig. 4A). In a parallel study, we found that the Cav3.2 (α1H) subunit immunostaining was predominant at the plasma membrane of the subconfluent MC3T3-E1 cells (Fig. 4B), with much lesser amounts seen intracellularly. As before, specific staining was abolished when peptide block was used (not shown).
In the next study, confluent MC3T3-E1 cells were found to continue expressing Cav1.2 (α1C) subunits, but a greater proportion of the staining appeared to be intracellular (Fig. 4C). When confluent cells were treated with ascorbate and β-glycerolphosphate to stimulate differentiation, immunostaining of Cav1.2 (α1C) decreased dramatically (Fig. 4E). At confluence, a larger pool of T-type Cav3.2 (α1H) channel was seen as diffuse intracellular staining (Fig. 4D), and this staining decreased, but did not disappear, during differentiation after treatment with ascorbate and β-glycerolphosphate (Fig. 4F). The specificity of staining was demonstrated by peptide blocking (Fig. 4G,H). Total protein was then extracted from MC3T3-E1 cells at various stages of growth and differentiation in order to measure the levels of expression of Cav1.2 (α1C) and Cav3.2 (α1H) subunits. The molecular weight of α1 subunits predicted from the cDNA sequence of rabbit skeletal muscle is approximately 212 kDa (Tanabe et al., 1987). Anti-Cav1.2 (α1C) subunit antibodies recognized an intense band with an apparent molecular weight of ∼190 kDa in osteoblastic MC3T3-E1 cell extracts as detected by Western blot (Fig. 4I, lanes 1,2). An array of published data suggests that the sizes of α1 subunits purified from native skeletal muscle, heart, and brain vary from 160 to 235 kDa even with the presence of protease inhibitors (De Jongh et al., 1991; Brawley and Hosey, 1992; Hell et al., 1993; Wang et al., 2000). Some smaller bands also were present (Fig. 4I, lanes 1,2), presumably representing the partial proteolytic fragments. The levels of expression of Cav1.2 (α1C) subunits decreased during differentiation (compare lanes 1 and 2 in Fig. 4I), and when assessed by densitometry were reduced by approximately 75% in fully differentiated cells treated with β-glycerolphosphate. This was consistent with what had previously been observed by immunostaining (Fig. 4C,E). The Cav3.2 (α1H) subunit was recognized by its specific antibodies as a single immunoreactive band of ∼200 kDa under reducing conditions, suggesting that T-type channels may not experience the same post-translational proteolysis (Fig. 4J, lanes 1,2). Interestingly, the fully differentiated MC3T3-E1 cells (Fig. 4J, lane 2) contained approximately 30% less Cav3.2 (α1H) protein than the undifferentiated osteoblasts (compare lanes 1 and 2 in Fig. 4J). Non-immune rabbit IgG, substituted for either primary antibody, recognized no protein products around 200 kDa (not shown). The specific immunoreactivity was also blocked by adding the original peptide as a blocker (not shown).
Immunostaining of subconfluent growth phase chondrogenic ATDC5 cells also indicated there were significant amounts of Cav1.2 (α1C) present and that these channels were localized both intracellularly and at the cell membrane (Fig. 5A). Under these same conditions, the Cav3.2 (α1H) subunits were primarily located along periphery of the cells at the plasma membrane (Fig. 5B). No staining was seen when a peptide block was performed (Fig. 5C). When examined by Western blot, both Cav1.2 (α1C) and Cav3.2 (α1H) subunits were identified in the protein lysates of whole ATDC5 cells (Fig. 5D, lanes 1, 2). The apparent molecular weights of Cav1.2 (α1C) and Cav3.2 (α1H) subunits as estimated from Western blotting of chondrogenic ATDC5 whole cell lysates were ∼195 and ∼210 kDa, respectively.
The MLO-Y4 cell line is the best in vitro model of the osteocyte (Kato et al., 1997). We found that the morphology of the cells changed in the time period between 48 and 98 hr post plating. Immediately after plating, MLO-Y4 cells attached and spread, but they did not develop the long processes characteristic of osteocytes until they had been cultured in αMEM for several days. Consistent with our failure to see significant staining of osteocytes with antibodies against Cav1.2 (α1C) in bone sections (Fig. 3) and consistent with the down-regulation of Cav1.2 (α1C) in differentiated MC3T3-E1 cells (Fig. 4), we found that osteocytic MLO-Y4 cells exhibited very low levels of immunostaining for the L-type Cav1.2 (α1C) subunits (Fig. 5E,H). In contrast, T-type Cav3.2 (α1H) subunit expression clearly was evident and localized primarily along the cell plasma membrane (Fig. 5F,I). Significant staining could be seen along the osteocytic processes (Fig. 5I). Controls using non-immune rabbit IgG were blank (Fig. 5G,J). No band near the 200-kDa molecular weight marker was detectable when whole cell protein lysates were probed with anti-Cav1.2 (α1C) subunit antibodies (Fig. 5K, lane 1) in Western blot. In contrast, anti-Cav3.2 (α1H) subunit antibodies recognized a specific protein band of ∼200 kDa in the total protein extracts (Fig. 5K, lane 2).
Bone formation during development depends on the coordinated activity of several cell groups that include cells of the chondrogenic and osteoblastic lineages, each under exquisite regulation by morphogens and growth factors (Olsen et al., 2000). In the growth plates of long bones, chondrocytes proliferate, differentiate, undergo hypertrophy, and finally apoptose in an orderly progression (Wallis, 1996; Olsen et al., 2000). Osteoblasts derived from the periosteum and from bone marrow eventually replace the hypertrophic chondrocytes at the vascularized chondro-osseous junction. In the process of calcified matrix formation that ensues, certain osteoblasts become embedded as osteocytes deep in the mineralized bone matrix (Kato et al., 1997). Although they are the most abundant cells, almost ten times as many as the osteoblasts in bones, osteocytes are not well studied because they are relatively hard to access (Parfitt, 1977).
Both clinical and laboratory observations have demonstrated the critical roles of Ca2+ transport and intracellular Ca2+ signaling in early bone formation and growth. VSCCs are key regulators of Ca2+ permeability in osteoblasts (Duncan et al., 1998). Osteoblasts regulate the expression levels of different subtypes of VSCCs to maintain Ca2+ permeability and modulate intracellular Ca2+ concentration. VSCCs also have been identified in chondrocytes, such as N-type channels, but the functional roles of those channels in cartilage are poorly understood (Zuscik et al., 1997). Osteoblasts exhibit molecular diversity of VSCCs (Barry et al., 1995; Wang et al., 2000).
Previous data showed that L-type Cav1.2 (α1C) and T-type Cav3.2 (α1H) VSCCs are the only L- and T-type channels regulated by 1,25-D3 in osteoblasts, and they are likely to be the central regulators of voltage-sensitive Ca2+ flux in bone development (Bergh et al., unpublished data). The present study identified the presence and expression patterns of these two VSCCs during skeletal development, using mouse femur as the model.
Immunohistological analysis indicated that Cav1.2 (α1C) subunits are present at high levels in osteoblasts and chondrocytes in growth plates throughout skeletal development, particularly at sites where cells are actively involved in bone formation. This extends the earlier observations made in intact rat femur and skull (Wang et al., 2000), and is consistent with accumulating evidence pointing to a critical role for L-type VSCCs in bone. The previously undocumented presence of T-type Cav3.2 (α1H) subunits in osteoblasts, osteocytes, and chondrocytes is highly significant. In non-excitable cells such as osteoblasts, the unique gating properties of T-type channels may allow Ca2+ influx to occur at small variations of membrane depolarization across the plasma membrane, sequentially activating other classes of high-voltage activated channels (Perez-Reyes, 2003) or perhaps mechanoreceptors. T-type VSCCs inactivate quickly, preventing Ca2+ overload from occurring and preventing apoptosis (Ertel and Ertel, 1997).
Osteoblasts in mature 6-week postnatal mouse long bones continued to express both L-type Cav1.2 (α1C) and T-type Cav3.2 (α1H) channels. Although examination at high magnification showed that marrow cells, osteoblasts, bone lining cells, and chondrocytes in intact bone tissue actively expressed Cav1.2 (α1C) subunits, they were remarkably reduced in terminally differentiated osteocytes, consistent with the observations made in rat bones (Wang et al., 2000). Interestingly, Cav3.2 (α1H) subunits were present in deeply embedded osteocytes, suggesting that although osteocytes significantly decreased the major L-type VSCC expression, they can regulate Ca2+ influx through the T-type VSCCs. Western blot and immunohistochemical assessment using in vitro cell models of osteogenesis, chondrogenesis, and an osteocytic cell line also were performed and produced findings similar to the in vivo studies. Wang et al. (2000) did not observe Cav1.2 (α1C) subunit immunostaining in marrow cells of the rat bone marrow cavity, but we readily detected these L-type VSCCs in hematopoietic stromal cells. This may be the result of fixation methods, since they used paraffin-embedded sections, and our analysis was made using frozen bone sections. Consistent with our findings, whole cell patch-clamp electrophysiological studies demonstrated both L-type (high-voltage activated) and T-type (low-voltage activated) currents in rat bone marrow stromal cells (Preston et al., 1996).
MC3T3-E1 and ATDC5 cells retain many of the critical features of osteoblastic and chondrogenic phenotypes, thus they are excellent models for in vitro studies of the molecular expression of VSCCs. Immunohistochemical analysis of MC3T3-E1 cells confirmed the presence of both types of VSCCs in osteoblasts of growth phase and at all stages of differentiation. Cav1.2 (α1C) subunits were present along the cell surface and in the secretory route as well, similar to the distribution pattern previously seen in rat osteosarcoma ROS17/2.8 cells (Wang et al., 2000). The T-type Cav3.2 (α1H) subunit was found almost exclusively along the plasma membrane in MC3T3-E1 cells. The Cav1.2 (α1C) and Cav3.2 (α1H) subunit protein expression levels both were down-regulated during end stage osteoblastic differentiation, but the extent of this reduction was much higher for Cav1.2 (α1C). It is intriguing to speculate that the L-type channels support processes occurring in proliferative cells that are down regulated in post-mitotic osteocytes, which thus switch to T-type channels to support Ca2+ movements.
MLO-Y4 cells are a newly established cell line derived from murine long bones, and they phenotypically resemble primary osteocytes, including having long dendritic processes for communication, expressing high levels of osteocalcin, but low alkaline phosphatase activity, and lacking osteoblast-specific factor 2 (Kato et al., 1997). We found that MLO-Y4 cells required several days in culture to develop their characteristic long processes. Gu et al. reported a lack of both L- and T-type VSCC currents or gene transcripts encoding L-type Cav1.2 (α1C), Cav1.3 (α1D) subunits, and T-type Cav3.1 (α1G) subunits in untreated osteocytes (Gu et al., 2001). Our immunohistochemical studies of both bone tissue and of osteocytic MLO-Y4 cells indicated low to nearly absent levels of L-type Cav1.2 (α1C) subunits, consistent with the results achieved by Western blot with anti- Cav1.2 (α1C) subunit-specific antibodies. Interestingly, the presence of T-type Cav3.2 (α1H) subunits, but not Cav3.1 (α1G) subunits (not shown), was detected in osteocytes. The fact that this conflicts with the published data could be explained if Cav3.2 (α1H) subunit expression in osteocytes is relatively low compared to the level in osteoblasts. T-type VSCCs usually activate at near threshold membrane polarization with transient current and small single channel conductance (Perez-Reyes, 2003). Thus, it could be difficult to detect the T-type currents in osteocytes if they are present at low levels. Meanwhile, VSCC expression levels are also under the regulation of various hormones and growth factors in vivo. Physiological levels of hormones increased the expression of both L-type Cav1.2 (α1C), Cav1.3 (α1D) subunits and T-type Cav3.1 (α1G) subunits to detectable levels in MLO-Y4 cells; Cav3.2 (α1H) was not examined (Gu et al., 2001).
The expression and biological roles of VSCCs in chondrocytes have not been well investigated. Cav1.2 (α1C) subunits were found in chondrocytes of rat tracheal cartilage (Wang et al., 2000). Because of a lack of highly selective T-type VSCC blockers, Zuscik et al. were unable to prove the existence of a putative T-type channel in growth plate chondrocytes (Zuscik et al., 1997). Our Western blotting and immunohistochemical studies conclusively have identified the expression of L-type Cav1.2 (α1C) and T-type Cav3.2 (α1H) VSCCs in both bone tissue growth plate chondrocytes and the ATDC5 chondrogenic cell line.
To further understand the functional roles of VSCCs in human disease, several α1 subunit transgenic mouse models have been established. The L-type Cav1.2 (α1C) subunit null mice died in utero between 12.5 and 14.5 postcoitum, while the heterozygotes were phenotypically normal, suggesting that a dosage compensation or a compensatory expression of other VSCCs may occur (Seisenberger et al., 2000). Severe cardiac hypertrophy, early ventricular fibrosis and apoptosis were developed in transgenic mice with Cav1.2 (α1C) subunit over-expression (Muth et al., 2001). T-type Cav3.2 (α1H) subunit null mice had coronary dilation deficiency and focal myocardial fibrosis (Chen et al., 2003). The generation of those transgenic mice provides a new opportunity to study the VSCCs and their functions in skeletal tissues.
In conclusion, both L-type Cav1.2 (α1C) and T-type Cav3.2 (α1H) VSCCs are critical components actively involved in the early mouse long bone skeletal development. The simultaneous appearance of both channels hints that their expression is coordinated during development to maintain Ca2+ homeostasis and ensure normal bone cell proliferation and differentiation. The selective loss of L-type channels in osteocytes with the persistence of T-type channels suggests that the processes in post-proliferative cells of the osteoblast lineage can be supported by low-voltage activated channels.
Mouse osteoblastic MC3T3-E1 cell line (subclone 14), obtained from ATCC (Manassas, VA), was maintained in α-MEM containing ribonucleosides and deoxyribonucleosides, supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS), 100 μg/ml streptomycin, 100 IU/ml penicillin, and 10 mM HEPES buffer. Functional differentiation of MC3T3-E1 cells was performed as previously described (Bergh et al., 2003). Murine carcinoma–derived chondrogenic ATDC5 cells were obtained from Dr. Véronique Lefebvre (Cleveland Clinic, Cleveland, OH). ATDC5 cells were cultured in DMEM/F-12 (1:1) containing 5% (v/v) FBS, 100 IU/ml penicillin, and 100 μg/ml streptomycin. Murine ostocyte-like MLO-Y4 cells were a generous gift from Dr. Lynda Bonewald (University of Missouri at Kansas City School of Dentistry, Kansas City, MO). MLO-Y4 cells were maintained in α-MEM containing ribonucleosides and deoxyribonucleosides with 5% (v/v) FBS and 5% (v/v) calf serum (Hyclone Laboratories Inc., Logan, UT), 100 μg/ml streptomycin, and 100 IU/ml penicillin in 0.15 mg/ml rat tail type I collagen–coated plates. All cell culture reagents were purchased from Gibco Invitrogen Life Technologies (Carlsbad, CA) unless otherwise stated. All cells were incubated at 37°C with 5% (v/v) CO2 and supplied with fresh culture medium every two days.
Animals and Tissue Processing
All experiments involving laboratory animals have been approved by the University of Delaware IACUC. Mouse fetuses were obtained from overstock Swiss Webster females that had been crossed with FVB/N males and were harvested at embryonic days 12.5, 14.5, 16.5, 18.5, and 21. Intact mouse femurs also were dissected from 6-week-old skeletally mature mice. The 6-week bones were fixed in 4% (v/v) paraformaldehyde (Electron Microscopy Sciences, Fort Washington, PA) in phosphate buffered saline (PBS) solution for 24 hr at 4°C, and then demineralized with 10% (w/v) EDTA/PBS (PH = 7.6). Physical endpoint testing was administrated to ensure 100% decalcification. The tissues were embedded in Tissue-Tek O.C.T. compound (Miles, Elkhart, IN), frozen down on dry ice and stored at −80°C until ready for sectioning. Sections (10 μm) were prepared on a cryostat and mounted on charged slides (Tissue-Tack, Warrington, PA). All tissue sections were stored at −20°C until used for immunohistochemistry studies.
The affinity-purified rabbit anti-Cav3.2 (α1H) subunit polyclonal antibody was raised against a synthetic peptide sequence [HLEEDFDKLRDVRATE] corresponding to amino acids 1,034–1,049 of the T-type Cav3.2 (α1H) subunit sequence (Accession number NP_067390). The antibody was prepared for us commercially by ResGen Invitrogen Life Technologies. The affinity-purified rabbit anti-L-type VSCC Cav1.2 (α1C) subunit antibody was generated against a GST fusion protein with residues 1–46 amino acids of rabbit Cav1.2 (α1C) sequence with serine 44 replaced by alanine (accession number P16381) with confirmed cross-reactivity with mouse (Alomone Research, Jerusalem, Israel). Rabbit polyclonal antibody to β-actin loading control was purchase from Novus Biologicals Inc. (Littleton, CO).
H & E Staining
The sections were placed in Harris Modified Hematoxylin for 2 min and rinsed with running tap water until the water was cleared. The sections were then immersed into acid ethanol (hydrochloric acid:70% ethnol (v/v) = 1:400) for 10 dips. The specimen were counterstained with Eosin Y for 30 sec, and dehydrated in ascending ethanol solutions: 50, 70, 95, and 100% (3 × 3 min). Finally, the stained specimens were cleared in xylene (3 × 3 min), and mounted with Permount. All reagents used in H&E staining were purchased from Fisher Scientific (Fairlawn, NJ)
Cells were fixed in a 2% (w/v) paraformaldehyde/PBS solution for 15 min at room temperature (RT). Bone sections were fixed with 4% (w/v) paraformaldehyde in PBS for 30 min at RT. After rinsed in PBS, all samples were permeabilized and blocked for 30 min with PBS containing 0.2% (v/v) Triton X-100 and 1% (v/v) normal goat (for cells) or donkey serum (for tissues). The specimens were incubated subsequently with anti-Cav1.2 (α1C) subunit or anti-Cav3.2 (α1H) subunit polyclonal antibodies described as above diluted 1:50 in PBS containing 1% (v/v) normal serum for 1 hr at 37°C. After rinsing 3 × 5 min in PBS, a 1:40 dilution of fluorescein (FITC) conjugated goat anti-rabbit secondary antibody (Jackson ImmunoResarch Inc., West Grove, PA) was applied to the cells at 37°C for 40 min in the dark, followed by a 10-min nuclear counterstaining with ToPro3® (Molecular Probes, Eugene, OR). The sections were incubated with a 1:50 dilution of Texas Red conjugated donkey anti-rabbit secondary antibody (Amersham Biosciences, Piscataway, NJ) at 37°C for 45 min in the dark and counterstained by Syto13® (Molecular Probes). Tissue sections used for collagen X staining to identify the hypertrophic zones were fixed in 5% (v/v) glacial acetic acid and 95% (v/v) ethanol mixture for 15 min and penetrated with 0.2% (w/v) bovine Type IV-S testicular hyaluronidase/PBS solution (Sigma, St. Louis, MO) for 45 min. Rabbit anti-mouse collagen X primary antibody diluted 1:200 in PBS was added for 1 hr at 37°C, followed by a 1:10 dilution of Texas Red conjugated donkey anti-rabbit secondary antibody (Amersham) for 40 min in the dark at 37°C. Following three rinses with PBS, all samples were mounted in Biomeda Crystal/Mount (Foster City, CA) to prevent fading. Immunofluorescent images were collected with a Zeiss LSM 510 multi-photon confocal microscope (Carl Zeiss, Oberkochen, Germany). Negative controls were performed with pre-incubation of the primary antibodies with the specific antigenic peptides, with non-immune rabbit IgG replacing the primary antibodies, or without primary antibodies. For antigenic pre-absorption, 3 μg peptide was incubated with 1 μg of the Cav1.2 (α1C) antibody, and 1 μg antigen was incubated with 1 μg of the Cav3.2 (α1H) antibody for 1 hr at 37°C. Long bone sections obtained from T-type Cav3.2 (α1H) subunit knock-out transgenic mice (Chen et al., 2003) were used as additional negative controls. Compared to the wild type mouse femurs, null mouse femurs showed no immunoreactivity for anti-Cav3.2 (α1H) subunit antibody, and immunostaining of the sections was blank (not shown), which further demonstrated the specificity of the T-type Cav3.2 (α1H) subunit antibody.
Cells were scraped from the plates and solubilized in ice-cold lysis buffer containing 20 mM Na2HPO4 (PH 7.4), 150 mM NaCl, 2 mM EDTA, 2 mM EGTA, 0.3% (v/v) Triton X-100, 100 μM PMSF with protease inhibitor cocktail III (Calbiochem, San Diego, CA). The whole cell lysates were centrifuged at 10,000 rpm for 20 min at 4°C, and the protein concentration of the supernatant was determined using BCA assay (Pierce, Rockford, IL). Samples were diluted with Laemmli sample buffer (Bio-Rad, Hercules, CA), and heated at 95°C for 3 min. Protein bands were separated on a 10% (w/v) SDS-PAGE gel, and then transferred to a nibrocellulose membrane. The membrane was blocked overnight at 4°C in 5% (w/v) non-fat milk in PBS containing 0.1% (v/v) Tween-20 (PBS-T), followed by overnight incubation of anti-Cav1.2 (α1C) or anti-Cav3.2 (α1H) subunit primary antibodies at 2.5–3 μg/ml in blocking buffer at 4°C. After 3 × 10 min rinses with PBS-T, the blots were treated with a 1:2,000 dilution of HRP conjugated goat anti-rabbit IgG (Pierce) for 2 hr at 4°C. The immunoreactive bands were visualized by Supersignal West Dura Extended Duration Substrate (Pierce). Protein band density was measured by ImageQuant software version 5.2 (Molecular Dynamics, Sunnyvale, CA). β-actin was used to normalize the protein sample loading. Antigenic peptide pre-incubated primary antibodies and non-immune rabbit IgG of equivalent concentration were used to ensure specificity of primary antibodies.
This study was supported by a grant from the National Institute of Dental and Craniofacial Research (DE 12641 to M.C.F.-C.). We thank Dr. Greg Lunstrum (Shriners Hospital for Children, Portland, OR) for providing the anti-collagen X antibody. We especially want to thank Dr. Kevin P. Campbell and Dr. Chien-Chang Chen (University of Iowa, Iowa City, IA) for providing tissues of the T-type VSCC Cav3.2 (α1H) subunit knock out mice. We also thank Drs. Daniel Carson, Randall Duncan, Norman Karin, Joel Bergh, Catherine Kirn-Safran, Ronald Gomes, Kirk Czymmek, Weidong Yang, Ms. Anissa Brown, Ms. JoAnne Julian, and Mr. Ben Rohe for their excellent input to this work, and Mrs. Margie Barrett for graphics assistance.