Extracellular Calcium and Parathyroid Hormone-Related Peptide Signaling Modulate the Pace of Growth Plate Chondrocyte Differentiation
http://www.100md.com
《内分泌学杂志》
Endocrine Research Unit, Department of Veterans Affairs Medical Center, Department of Medicine, University of California, San Francisco, California 94121
Abstract
An adequate supply of Ca2+ is critical for normal growth plate development. Previous studies suggest that changes in extracellular [Ca2+] ([Ca2+]e) modulate the function of chondrocytes with high [Ca2+]e promoting cell differentiation. In contrast, signal transduction by the PTH/PTHrP type I receptor (PTH1R) slows down chondrocyte differentiation. This study addressed whether changes in [Ca2+]e modulate the differentiation of mouse growth plate chondrocytes by interacting with PTHrP/PTH1Rs. Raising [Ca2+]e from 0.5–3.0 mM dose-dependently promoted the development of mouse growth plate chondrocytes as indicated by decreases in proteoglycan accumulation and in the expression of early differentiation marker genes and by increases in mineral deposition and in the expression of markers of terminal differentiation. The effects of high [Ca2+]e on gene expression and matrix synthesis were blunted by incubating cells with PTHrP and vice versa. High [Ca2+]e also suppressed the expression of PTH1Rs. Chronic stimulation of PTHrP/PTH1R signaling by adenoviral expression of constitutively active human PTH1Rs (223hPTH1Rs) reduced the effects of high [Ca2+]e on proteoglycan synthesis and gene expression. Similar results were seen when we treated cells with forskolin or 8-bromo-cAMP. Taken together, these data support the idea that the pace of chondrocyte differentiation depends on a balance of interactions between PTHrP/PTH1R and extracellular Ca2+ signaling and that high [Ca2+]e promote cell differentiation potentially by reducing the availability of PTH1Rs and the level of cAMP-dependent signal transduction.
Introduction
GROWTH PLATE DEVELOPMENT is the critical step in endochondral bone formation that controls longitudinal growth in vertebrates. In the growth plate, chondrocytes progress through steps of resting, proliferation, maturation, hypertrophy, and terminal differentiation, eventually creating a mineralized matrix that supports skeletal development (1). At each step, chondrocytes produce distinct matrix proteins that signify the stages of differentiation (1, 2). Resting and proliferating chondrocytes produce high levels of the 1-subunit of type II collagen [1(II)] (1, 3, 4) and cartilage-specific proteoglycan core proteins such as aggrecan (Agg) (1, 5). Maturing and upper hypertrophic chondrocytes express the type X collagen [1(X)] and alkaline phosphatase (ALP) as well as the PTH1R. Terminally differentiated chondrocytes in the lower hypertrophic zone (or mineralization zone) demonstrate high levels of the osteopontin (OP) (6), osteonectin (ON) (7, 8), and osteocalcin (OC) (9) expression and deposit Ca2+- and phosphate-containing mineral (i.e. hydroxyapatite) into the matrix (1). At the chondro-osseus junction, vascular invasion takes place, and chondroclasts resorb cartilage matrix, providing a surface for the deposition of bone matrix by osteoblasts. This marks the first step of bone formation.
Chondrocyte differentiation is governed by factors that control the rate of gene expression and phenotypic change so that orderly rates of bone growth are maintained. Both delayed and accelerated cell differentiation in the growth plate can cause growth retardation (10, 11, 12, 13, 14). In vitro and in vivo studies showed that PTHrP and the PTH1R play critical roles in chondrocyte differentiation (15). Deletion of PTHrP (PTHrP–/–) (14) or PTH1R (PTH1R–/–) (16) genes produces a short-limbed dwarfism in mice due to accelerated chondrocyte differentiation, premature mineralization, and shortened growth plates. Growth plate defects in PTHrP–/– mice are partially rescued by overexpressing constitutively active PTH1Rs (223hPTH1Rs) in chondrocytes (17). This mutant receptor, containing a His Arg mutation at position 223, causes constitutive activation of adenylate cyclase with increased cellular cAMP in the absence of ligand (18). cAMP is a key second-messenger that delays chondrocyte differentiation (19, 20). This receptor, interestingly, does not stimulate phospholipase C, another pathway activated by wild-type PTH1Rs (18). These observations support a role for PTHrP and its receptor in delaying growth plate development, likely via the activation of cAMP-dependent signaling cascade. The accelerated chondrocyte differentiation in PTHrP–/– or PTH1R–/– mice also suggests the existence of pathways promoting differentiation which are unmasked in the absence of PTHrP/PTH1R signaling.
Sufficient extracellular Ca2+ is needed for growth plate development (21, 22, 23). Rickets of many etiologies, in which systemic levels of Ca2+ are low, produces abnormally expanded and demineralized growth plates and growth retardation. High Ca2+ diets can correct these defects in several instances (23, 24, 25). Similarly, the rickets and growth plate abnormalities of vitamin D receptor (VDR) knockout mice (21, 26) recede with a high Ca2+ diet. These findings highlight the potential importance of Ca2+ in controlling chondrocyte maturation and function. Because the attendant secondary hyperparathyroidism and renal phosphate wasting are also reversed by high Ca2+ diets in these models, it has been difficult to assess the role of Ca2+ on chondrocyte function in vivo (21). In vitro studies by us and others provide evidence that changes in [Ca2+]e directly modulate key aspects of chondrocyte differentiation—proteoglycan (PG) and matrix accumulation, gene expression, and mineralization (27, 28, 29, 30). These studies further explore effects of Ca2+ on chondrocytes independent of the associated metabolic changes seen in models of rickets.
These studies tested the hypothesis that mouse growth plate chondrocytes (mGPCs) sense changes in [Ca2+]e and that this promotes cell differentiation by counterbalancing PTHrP/PTH1R signal transduction. We found that the rate of chondrocyte differentiation depends on a balance of interactions between PTHrP/PTH1R and extracellular Ca2+ signaling. High [Ca2+]e enhanced mGPC development, whereas activation of PTH1Rs by PTHrP retarded it, by mechanisms that involve cAMP signal transduction. Reduced expression of PTH1Rs, due to high [Ca2+]e, appears to comprise at least part of the mechanism for advancing chondrocyte differentiation.
Materials and Methods
Materials
Fetal calf serum (FCS) was purchased from Tissue Culture Biologicals (Tulare, CA). PTHrP (1–34) was from Bachem (Torrance, CA). Trypsin, collagenase IA, hyaluronidase, deoxyribonuclease II, forskolin, and 8-bromo-cAMP (8-Br-cAMP) were purchased from Sigma-Aldrich (St. Louis, MO). Primers and probes for quantitative real-time PCR (qPCR) were synthesized by Integrated DNA Technologies (Skokie, IL) or purchased from Applied Biosystems Inc. (Foster City, CA). Antisera raised against PTH1Rs were from Upstate Cell Signaling Solution (Lake Placid, NY). Other supplies were from previously noted sources (28).
mGPC and RCJ3.1C5.18 (C5.18) cell cultures
Epiphyseal growth plates from newborn mice (2–4 d old, Black Swiss strain) were obtained and chondrocytes were released by enzymatic digestion. Briefly, growth plate cartilage was first incubated with trypsin (0.25%, wt/vol) in a chondrocyte isolation medium [CIM: DMEM + penicillin (100 μg/ml), streptomycin (100 U/ml), NaHCO3 (45 mM), HEPES (20 mM, pH 7.4), fungizone (0.25 μg/ml), and gentamicin (0.15 mg/ml)] for 15 min at 37 C, followed by digestion with a mixture of (wt/vol) collagenase IA (0.18%), hyaluronidase (0.1%), and deoxyribonuclease II (0.01%) in CIM for 30 min at 37 C with gentle agitation. After removing undigested cartilage with a nylon mesh and small particulate minerals by brief (15 sec) centrifugation at 100 x g, mGPCs in suspension were collected by centrifugation at 400 x g for 10 min. Remaining cartilage was continuously digested with fresh enzymes until no tissue remained. The first cell harvest was routinely discarded, but subsequent harvests were combined and plated (105 cells/cm2) in a chondrocyte maintenance medium [CMM: DMEM + penicillin (100 μg/ml), streptomycin (100 U/ml), and FCS (10%, vol/vol)] at 37 C. For time-course experiments, cells were grown in CMM until confluence and switched to chondrocyte differentiation medium [CDM, -MEM containing 1.8 mM Ca2+ plus ascorbic acid (50 μg/ml) and -glycerol phosphate (5 mM)] for different times. To test the effects of Ca2+, PTHrP, forskolin, and 8-Br-cAMP, confluent cultures were switched to CDM with different [CaCl2] (0.1–4.0 mM) in the presence or absence of the reagents to be tested for different times. Chondrogenic C5.18 cells were cultured as previously described (28), and the effects of Ca2+ and PTHrP were tested as described above. These studies were approved by the Animal Care Subcommittee of the San Francisco Department of Veterans Affairs Medical Center.
Alcian green, alizarin red, and von Kossa staining
The accumulation of PG and mineral in cultures was assessed by staining with alcian green 2GX and alizarin red S and quantified by absorbance at 340 and 540 nm, respectively (28, 29). Cells cultured on glass coverslips were examined after von Kossa staining to detect mineral deposits, alcian green staining for PG accumulation, and counterstaining with hematoxylin to reveal cell morphology (28, 29).
qPCR
Total RNA was isolated as described (28). First-strand cDNAs were reverse-transcribed with Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). Genes of interest were quantified using probe-based TaqMan qPCR kits and ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA). A Ct (minimal PCR cycles required to generate a fluorescent signal exceeding a preset threshold) was determined by SDS software (Applied Biosystems) for each gene and normalized to a Ct for a housekeeping gene (L19) determined in parallel. Expression of L19 in mGPCs was unchanged with [Ca2+]e, PTHrP, forskolin, and 8-Br-cAMP (data not shown). Fluorescent probes and PCR primers for each gene were designed based on published sequences (see Table 1) or purchased. Sequences of commercial primers are not available due to proprietary protection (Applied Biosystems).
Immunoblotting
Lysates from virally infected mGPCs were prepared, and immunoblotting was performed as previously described (31, 32, 33) using antisera directed against PTH1Rs.
Generation of adenoviral stocks and infection of chondrocytes
Adenoviruses carrying the cDNA encoding the 223hPTH1R were made using an Adeno-X Expression System II kit (BD Biosciences, Palo Alto, CA) according to the manufacturer’s instructions. The 223hPTH1R cDNA was provided by Dr. Harald Jueppner (Harvard Medical School, Boston, MA). This cDNA was first subcloned into a pDNR-CMV vector containing two loxP sites flanking the cloning site. The 223hPTH1R cDNA was then transferred by Cre/Lox recombination to an acceptor cosmid containing the human adenovirus type 5 genome lacking E1 and E3 genes, a CMV promoter, and simian virus 40 poly-adenylation sequence, creating the Ad-223hPTH1R construct. After removing nonviral components by PacI restriction digestion, the Ad-223hPTH1R construct was transfected into human embryonic kidney 293 cells, which stably express E1 proteins and allow for replication of viral DNAs and packaging of viruses. Because the E1 and E3 genes are not present in the Ad-223PTH1R construct, the resulting viruses are replication incompetent in other cells. After mass production in human embryonic kidney 293 cells, viral stocks were titered using an Adeno-X rapid titer kit (BD Biosciences) and used to infect mGPCs. Potential contamination of replication-competent viruses in the viral stocks was verified by their ability to propagate in Chinese hamster ovary and C5.18 cells.
For infection, confluent mGPCs were incubated with viruses (titers as specified) for 72 h in CMM and switched to CDM containing different [Ca2+]e with or without PTHrP (10–7 M) for time points specified before samples were prepared for analysis. Titers used in the experiments were calculated based on the numbers of cells seeded at the beginning of culture.
cAMP assay
To determine cAMP content in response to short-term PTHrP treatment, mGPCs were plated in 24-well plates, cultured for 7 or 14 d in CMM, and then treated with different [Ca2+]e in the presence or absence of PTHrP for 15 min. cAMP was quantified by RIA (32, 34) and presented as total cAMP in picomoles per well.
To determine cAMP content in cells overexpressing 223hPTH1Rs, mGPCs were cultured in six-well or 12-well plates for 3–4 d until they were confluent and infected with Ad-223hPTH1R [20–100 plaque-forming units (pfu)/cell] or control Ad-pAX (100 pfu/cell) for 72 h. In parallel, uninfected or Ad-pAX-infected cells were incubated continuously with PTHrP (10–7 M) for 3 d or exposed briefly (15 min) to PTHrP before sample collections. After cAMP was eluted from the cells with ethanol (100%) as described (34), cell number was determined by quantifying crystal violet (CV) staining (35). cAMP content was normalized to levels of CV staining in each well and expressed as fold change over control levels.
CV staining
After eluting cAMP, cells were incubated with CV (0.2%, wt/vol) in 2% ethanol (vol/vol) for 1 h. After four washes with distilled H2O, stain was eluted by Triton X-100 (0.2%, vol/vol) and quantified by absorbance at 590 nm (35).
TUNEL [terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate (dUTP) nick end labeling] staining
Apoptosis was assessed by TUNEL using a commercial Apoptag kit (Oncor, Gaithersburg, MD) according to manufacturer’s instructions. Briefly, mGPCs were cultured on coverslips in 12-well plates until confluence, followed by infections with Ad-223hPTH1R or Ad-pAX or treatments with PTHrP, forskolin, or 8-Br-cAMP for 72 h. Cells on coverslips were fixed with paraformaldehyde (1%) for 10 min and permeabilized in ethanol/acetic acid (2:1, vol/vol) mixture for 5 min at –20 C. Endogenous peroxidase was quenched by H2O2 (3%) in PBS. After preincubation with a reaction buffer (pH 8.0) for 10 min, cells were incubated with a reaction mixture containing terminal deoxynucleotidyl transferase and digoxigenin-conjugated dUTP (Dig-dUTP) for 1 h at 37 C. Nicked DNA labeled with Dig-dUTP was detected by peroxidase-conjugated anti-digoxigenin antibody and 3,3'-diaminobenzidine substrate. After counterstaining with methyl green, cells were examined microscopically.
Results
Differentiation of mGPCs in culture
The characteristics of primary cultures of growth plate chondrocytes from the limbs of newborn mice recapitulate many key steps in growth plate development. Under subconfluent conditions, the cells produced a PG-containing matrix that stained with alcian green (Fig. 1A). The amount of PG increased with time (Fig. 1, B and C) and decreased when the cells began to deposit minerals, indicated by von Kossa staining, at 7–9 d after confluence (Fig. 1D). Thereafter, mineral deposition increased as PG accumulation declined (Fig. 1E). Levels of RNA encoding matrix proteins, assessed by qPCR, correlated with these biochemical changes. RNA levels for early differentiation markers, Agg and 1(II), were highest in subconfluent and confluent cultures and decreased (by >90%) in postconfluent cultures (Fig. 1, F and G). Conversely, expression of late differentiation markers, ALP, 1(X), OP, and OC, were low in subconfluent cells and increased in postconfluent cultures (4- to 300-fold) (Fig. 1, H–K). These observations indicated that mGPCs in culture demonstrate a temporal sequence of gene expression and matrix protein synthesis characteristic of key steps in growth plate differentiation.
Effects of changing [Ca2+]e on cultured mGPCs
We next tested the effects of changing [Ca2+]e on the profiles of differentiation markers in mGPCs. Raising [Ca2+]e dose-dependently decreased PG accumulation (Fig. 2A) with an ID50 (dosage for 50% inhibition) of approximately 2 mM Ca2+ at 11 d after confluence (Fig. 2B). Changes in [Ca2+]e, however, had minimal, if any, effects on PG synthesis in cultures at 4 d after confluence (Fig. 2B), suggesting that the responsiveness of PG production to changing [Ca2+]e increased as cell differentiation proceeded. On the other hand, high [Ca2+]e promoted mineral accumulation in the matrix surrounding the cells (Fig. 2A) in a concentration- and time-dependent manner (Fig. 2C), beginning at d 7 and as shown for d 14, 21, and 28. The EC50 for mineralization was approximately 2–3 mM Ca2+ at 21 and 28 d. These morphological changes were accompanied by changes in expression of several marker genes. In cultures grown at 0.5 mM Ca2+ for 14 d, early differentiation markers, Agg and 1(II) (Fig. 2, D and E), were coexpressed with markers of maturing and early hypertrophic chondrocytes-1(X), ALP, and the PTH1R (Fig. 2, F–H). At this stage of culture, OP, a marker of terminally differentiated chondrocytes, was low compared with its expression at high [Ca2+]e (Fig. 2I). We interpreted these data as either that the cells were slowly transitioning from resting and proliferation to the maturing and hypertrophic stages of differentiation or that this was a mixed population of cells that were at distinct stages of differentiation.
In cultures maintained at normal and high [Ca2+]e (1.0–3.0 mM) for 14 d, RNA levels for Agg and 1(II) (Fig. 2, D and E) were reduced (by up to 72 and 76%, respectively) in a dose-dependent manner (P < 0.01). The effects of high [Ca2+]e on gene expression were observed as early as 3 d after changes in [Ca2+]e and remained evident at d 21 (data not shown). The effects of changing [Ca2+]e on expression of 1(X), ALP, and PTH1Rs were biphasic. Maintaining cells at 1.0 vs. 0.5 mM Ca2+ produced slightly but significantly (P < 0.05) greater levels of RNA encoding these three genes (Fig. 2, F–H), suggesting that cells grown at 1.0 mM Ca2+ were more mature than cells at 0.5 Ca2+ mM after 14 d in culture. Culturing at 3 mM Ca2+ for the same duration of time reduced expression of 1(X), ALP, and PTH1R RNA (P < 0.001) and significantly increased OP expression (Fig. 2I), indicating that the majority of cells at this [Ca2+]e had matured and reached terminal differentiation at which point the expression of 1(X), ALP, and PTH1R genes are typically decreased. Thus, our data indicate that cultured mGPCs were influenced by changes in [Ca2+]e, and high [Ca2+]e tended to promote features of chondrocyte differentiation.
Effects of PTHrP on mGPCs
PTHrP and its receptor play critical roles in pacing chondrocyte differentiation. To examine whether the impact of high [Ca2+]e on this process is influenced by the level of PTHrP/PTH1R signaling, we compared the effects of [Ca2+]e in the presence and absence of PTHrP. Culturing mGPCs with PTHrP (1–34) (10–7 M) for 14 d significantly blunted the ability of high [Ca2+]e to suppress PG accumulation and promote mineral deposition (Fig. 3A; –PTHrP vs. +PTHrP). This was confirmed by right-shifted Ca2+ dose-response curves (Fig. 3, B and C). The EC50 for PG accumulation and mineralization increased from approximately 1.2 to >3 mM Ca2+ with PTHrP treatment (Fig. 3, B and C). PTHrP also blocked the ability of high [Ca2+]e to influence gene expression. In mGPCs maintained at 0.5 mM Ca2+, treatment with PTHrP for 14 d significantly (P < 0.001) increased RNA levels for Agg by 2.5-fold (Fig. 3D) and markedly reduced the expression of 1(X), ALP, and PTH1R—markers of maturing and hypertrophic chondrocytes—by 65–75% (P < 0.001) (Figs. 3, G–I). These changes in gene expression indicate that PTHrP delayed the transition of mGPCs into more mature stages. Incubation with PTHrP also reduced the ability of high [Ca2+]e (3 mM) to inhibit Agg and 1(II) expression (Fig. 3, D and E) and to increase OP RNA levels (Fig. 3F), suggesting that increased PTHrP/PTH1R signaling directly or indirectly counteracted the effects of high [Ca2+]e on cell differentiation. Similar studies were performed with chondrogenic C5.18 cells, and comparable results were obtained (data not shown).
Effects of changing [Ca2+]e and PTHrP on cAMP signal transduction in mGPCs
We noted that in both mGPCs and C5.18 cells even maximal doses of PTHrP (10–7 M) did not completely block the high [Ca2+]e-induced mineralization, inhibition of PG accumulation, and changes in gene expression (Fig. 3 and data not shown). This could be explained by down-regulation of PTH1R expression at high [Ca2+]e (Fig. 2H), leading to a reduction in PTHrP/PTH1R signaling capacity. We cannot, however, rule out the possibilities: 1) that a portion of high [Ca2+]e-induced changes in chondrocyte function are independent of PTHrP/PTH1R signaling; and or 2) that high [Ca2+]e affect the ability of PTH1Rs to activate downstream signal transduction.
To test the latter possibility, we first examined the impact of changing [Ca2+]e on PTHrP-induced signal transduction in these cells. The best-characterized signaling pathway activated by the PTH1R is the G protein (Gs)-activated increase in adenylate cyclase activity causing cAMP accumulation. We assessed the effects of different [Ca2+]e on PTHrP-induced cAMP accumulation. Treating mGPC cells with PTHrP for 15 min dose-dependently increased cAMP production by up to approximately 40-fold with an ED50 of approximately 3 x 10–9 M at 0.5 mM Ca2+ (Fig. 4A). Raising [Ca2+]e from 0.5 to 3, 5, or 10 mM, however, had no significant effect on cAMP production induced by PTHrP (10–9 M) in either 7- or 14-d-old cultures (Fig. 4B), indicating that different [Ca2+]e did not affect the ability of PTH1Rs to couple to adenylate cyclase activation in short-term incubations.
We, however, found that continuous incubation with PTHrP (10–7 M) desensitized responses of mGPCs to that ligand. As shown in Fig. 5, confluent mGPCs cultured in the absence of PTHrP for 3 d responded to a short-term (15 min) treatment of PTHrP (10–7 M) and significantly (P < 0.001) increased cAMP production by approximately 51-fold (, P < 0.001; +PTHrP vs. –PTHrP). In cells continuously cultured with PTHrP (10–7 M) for 3 d, cAMP levels at the end of the treatment (Fig. 5, ; + PTHrP pretreatment/–PTHrP) were indistinguishable from the levels in cultures without PTHrP (Fig. 5, ; –PTHrP pretreatment/–PTHrP). The cells continuously exposed to PTHrP also failed to increase cAMP production in response to an additional dose of PTHrP (Fig. 5, ; –PTHrP vs. +PTHrP), confirming the insensitivity of the cells to the ligand. These data demonstrated that prolonged activation of PTH1Rs down-regulate their signaling capacity. This was likely due at least in part to the down-regulation of PTH1R gene as we demonstrated in Fig. 3I.
Overexpression of 223hPTH1R in mGPCs
In our experiments of long-term growth of mGPCs at high [Ca2+]e, we observed a decrease in PTH1R mRNA (Fig. 2H), suggesting that high [Ca2+]e caused a decrease in receptor number in the differentiated chondrocytes and reduced the ability of PTHrP to mediate signal transduction. To test this possibility, we overexpressed a constitutively active PTH1R by adenoviral vectors and determined whether sustained activation of PTH1R signaling affected the ability of high [Ca2+]e to promote cell differentiation. We used a previously characterized activating human PTH1R (223hPTH1R) that elevates cellular cAMP in the absence of ligand (18). Infecting mGPCs with Ad-223hPTH1R (5–200 pfu/cell) dose-dependently increased RNA levels for the receptor, assessed by qPCR using primers specific to the human PTH1R gene (Fig. 6A). Increased RNA levels were evident at 3 d after infection (Fig. 6A) and lasted for at least 10 d (see Fig. 7D). In contrast, RNA levels for endogenous PTH1Rs (mPTH1R) were significantly decreased in the 223hPTH1R-expressing mGPCs (Fig. 6B), assessed with primers specific to the mouse PTH1R gene, indicating that signaling activated by the 223hPTH1Rs could down-regulate the expression of the endogenous PTH1R gene. Despite the decreased expression of endogenous receptors, Western blotting, using antisera immunoreactive with both mouse and human PTH1Rs, showed an overall increase in protein expression in the cells infected with Ad-223hPTH1R at titers of at least 20 pfu/cell (Fig. 6C).
Constitutive activity of the 223hPTH1R was confirmed by increases in cAMP levels in the infected mGPCs (Fig. 6D, ). In mGPCs 3 d after infection with Ad-223hPTH1R (50 and 100 pfu/cell), we observed modest but significant (P < 0.01) increases in cAMP accumulation by 1.8- and 2.9-fold, respectively, compared with cells that were infected with control viruses (100 pfu/cell) (Fig. 6D, ). In control cells, incubation with PTHrP for 15 min (Fig. 6D, +PTHrP) increased cAMP production by approximately 47-fold, indicating that the procedure that we used to infect mGPCs did not impact on the responsiveness of mGPCs to PTHrP.
Effects of 223hPTH1R on the differentiation of mGPCs
We next examined the effects of overexpressing 223hPTH1Rs on high [Ca2+]e-induced chondrocyte differentiation. In control cells infected with Ad-pAX (50 pfu/cell), raising [Ca2+]e for 7 d suppressed RNA levels for early differentiation markers, Agg by 82% and 1(II) by 60% at 3.0 mM vs. 0.5 mM Ca2+ (Fig. 7, A and B; ), and increased expression of the terminal differentiation marker OP by 10-fold at 3.0 mM vs. 0.5 mM Ca2+ (Fig. 7C; ) with a potency comparable to that in noninfected cells (see Fig. 2). The effects of high [Ca2+]e (i.e. 3 mM) on the expression of the above matrix genes, however, were significantly altered in cells infected with Ad-223hPTH1R (Fig. 7, A–C; ). RNA levels for Agg and 1(II) in the 223hPTH1R-expressing cells cultured at 3 mM Ca2+ were 2.6- and 1.5-fold higher than the levels in control cells grown at 0.5 mM Ca2+ (P < 0.001) and 12- and 4-fold higher than the levels in control cells cultured at 3 mM Ca2+ (P < 0.001), respectively. This indicates that constitutive activation of PTH1R signaling counteracted in a dramatic way the effects of high [Ca2+]e on the expression of the matrix genes, particularly Agg and 1(II). Overexpression of 223hPTH1Rs had modest (30% inhibition) but significant (P < 0.05) impact on high [Ca2+]e-induced OP expression (Fig. 7C). Expression of 223hPTH1Rs remained evident in cells at this stage of culture (10 d after infection) (Fig. 7D). Its levels, however, were substantially lower (<5% of the level at 3 d after infection; Fig. 6A).
Adenoviral expression of 223hPTH1Rs also blocked the ability of high [Ca2+]e to inhibit PG accumulation as indicated by alcian green staining shown in Fig. 8A. In control cultures, raising [Ca2+]e for 10 d dose-dependently suppressed PG content by up to 35% (Fig. 8A, ), but these changes in [Ca2+]e had no effect on cultures infected with Ad-223hPTH1R (Fig. 8A, ). Incubating the control cells with PTHrP (10–7 M) mimicked the effects of the infection with Ad-223hPTH1Rs on PG accumulation as expected (Fig. 8A, +PTHrP), suggesting that overexpressing 223hPTH1Rs produced sufficient signaling and downstream responses to counteract the effects of high [Ca2+]e on PG synthesis.
Interestingly, in cultures infected with Ad-223hPTH1Rs (Fig. 8B, ), raising [Ca2+]e for 14 d enhanced mineral deposition with a potency similar to that in cells infected with control viruses (Fig. 8B, ). The high [Ca2+]e-induced mineralization in both populations of cells was blocked by coincubation with PTHrP (10–7 M) during the Ca2+ treatment (Fig. 8B, +PTHrP). This suggests that signaling by 223hPTH1Rs was insufficient to block high [Ca2+]e-induced mineralization unless PTHrP was added to cultures.
Effects of 223hPTH1R on the growth of mGPCs
In vitro (20) and in vivo (36) studies suggest that activation of PTHrP/PTH1R signaling enhances the proliferation of chondrocytes. We, therefore, tested whether overexpressing 223hPTH1R increases the growth of mGPCs. On the contrary, we found that in mGPC cultures infected with Ad-223hPTH1R (50 and 100 pfu/cell), cell numbers decreased by approximately 45 and 70% (Fig. 9A, ) (P < 0.01), respectively, when compared with those of cultures infected with control viruses (Ad-pAX, 100 pfu/cell) (Fig. 9A, ).
The 223hPTH1R has been proposed to activate adenylate cyclase activity predominantly and not phospholipase C (18). We tested whether sustained activation of this signaling network was the cause for the decreased cell numbers in Ad-223hPTH1R-infected cultures. Continuous incubation with either 8-Br-cAMP (10–3 M) or forskolin (10–4 M), two activators of cAMP-mediated signaling, did not decrease the number of cells (Fig. 9B). Nor did continuous incubation with PTHrP (10–7 M) impact on the growth of mGPCs (Fig. 9B). In contrast, we observed a modest (20%) but significant (P < 0.05) increase in cell numbers in the cultures treated with forskolin (Fig. 9B).
We further tested whether the decreased cell numbers in mGPCs overexpressing 223hPTH1Rs were due to apoptosis by performing TUNEL staining (Fig. 9, C–H). In mGPCs 3 d after infection with Ad-223hPTH1R (50 and 100 pfu/cell) (Fig. 9, D and E), we observed increased numbers of apoptotic cells as indicated by brown 3,3'-diaminobenzidine-TUNEL staining (arrows). In cultures infected with control vectors (Fig. 9C) or grown with PTHrP (10–7 M), forskolin (10–4 M), or 8-Br-cAMP (10–3 M) for 3 d (Fig. 9, F–H), there were rare apoptotic cells. These observations suggest that constitutively activating 223hPTH1Rs induced apoptosis through pathways independent of cAMP signaling.
Effects of cAMP signaling on the differentiation of mGPCs
To test whether direct activation of cAMP signaling mimics the effects of PTHrP and overexpressing 223hPTH1R on the differentiation of mGPCs, we assessed matrix gene expression and mineralization in cultures treated with forskolin (10–4 M) or 8-Br-cAMP (10–3 M). In mGPCs cultured without these reagents, raising [Ca2+]e from 0.5–1.5 and 3.0 mM for 10 d decreased RNA levels for Agg by 74% and 1(II) by 60% and increased OP expression by 2.5-fold (Fig. 10A, ) as we demonstrated in the earlier experiments. Treating cells with forskolin significantly (P < 0.001) blocked the ability of high [Ca2+]e (i.e. 3 mM) to inhibit the expression of Agg and 1(II) and to increase OP expression (Fig. 10A, ). Incubating cells with 8-Br-cAMP profoundly impacted on matrix mineralization. Raising [Ca2+]e from 0.5 to 2 and 3 mM for 10 d in the absence of 8-Br-cAMP increased mineral deposition by 2.5-fold in mGPC cultures as expected (Fig. 10B, ). The effects of [Ca2+]e on the mineralization were blocked in cultures treated with 8-Br-cAMP (Fig. 10B, ). Blockade of high [Ca2+]e-induced mineralization was also observed in cultures treated with forskolin (data not shown). Taken together, these data support the idea that activation of cAMP signal transduction plays a role in counteracting the differentiation-promoting actions of high [Ca2+]e.
Discussion
Chondrocytes in the growth plate differentiate at an orderly rate that maintains normal bone growth. Both delay in and acceleration of chondrocyte differentiation cause growth retardation (10, 15). The PTHrP/PTH1R/Indian hedgehog feedback mechanism is critical in slowing down and thereby pacing chondrocyte differentiation (10, 15). The factors that counterbalance the PTHrP/PTH1R/Indian hedgehog system and promote chondrocyte transit into the steps of terminal differentiation remain to be proven, but there are several candidate molecules such as IGF-I, IGF-I receptor, VDR, thyroid hormone, and TGF- (37). Our previous studies suggested that extracellular Ca2+, possibly acting via extracellular Ca2+-sensing receptors, contributes to controlling the rate of differentiation in C5.18 cells (28). The experiments reported herein addressed whether extracellular Ca2+ interacts with or influences the PTHrP/PTH1R signaling pathways. This is a first step in determining a mechanism by which extracellular Ca2+ might influence growth plate differentiation. We confirmed that PTHrP and its receptor regulate chondrocyte maturation and differentiation, which is well established from in vivo models (10) and demonstrated that changes in [Ca2+]e significantly modify this regulation.
We found that cultured mGPCs recapitulate key steps in the progressive development of the growth plate. These cells express early differentiation markers shortly after plating and markers of terminally differentiated chondrocytes in later cultures. These characteristics indicate this is a suitable model for studying factors that modulate the time-dependent progression of chondrocyte differentiation. Raising [Ca2+]e advances cell differentiation by suppressing the expression of early chondrogenic genes [i.e. Agg and 1(II)] and PG synthesis and enhancing the expression of terminal differentiation markers [i.e. OP, OC, and ON] and mineral deposition. These observations are compatible with our previous findings with the chondrogenic C5.18 cell model system (28) in which raising [Ca2+]e promoted cell differentiation. Consistent with this differentiation scheme, Wu and colleagues (30) showed that high [Ca2+]e enhanced terminal differentiation of hypertrophic chondrocytes in bone rudiment cultures. Bonen and Schmid (27) also reported that a shorter incubation for 3 d of chick embryonic growth plate chondrocytes at 5 mM Ca2+ increased the synthesis and accumulation of 1(X) protein in the matrix, indicating that these cells displayed signs of a more mature phenotype shortly after treatment with high [Ca2+]e. According to the expression pattern of 1(X) in growth plates—beginning in maturing chondrocytes, peaking in upper hypertrophic chondrocytes, and decreasing in lower hypertrophic chondrocytes in the mineralization zone (38)—this increase in 1(X) expression indicated a transition of the cells into the stages of maturation and early hypertrophy, but not yet into the terminal differentiation. Although long-term treatment with high [Ca2+]e was not performed in the study of Bonen and Schmid, we predict that high [Ca2+]e would eventually reduce 1(X) content in the cultures as cells become terminally differentiated after long-term culture. This prediction is based on our observations that high [Ca2+]e ( 3 mM) suppressed the RNA levels for 1(X) in mGPCs cultured for 14 d. Together these observations lend support to a model in which low systemic Ca2+ levels participate in the delay of hypertrophic chondrocyte differentiation in conditions like rickets. Our study does not exclude important roles for the other systemic derangements (hyperparathyroidism and hypophosphatemia) that accompany rickets and contribute to the skeletal phenotype. Studies by Donohue and Demay (39) in fact suggest that reduced cell apoptosis, as a result of hypophosphatemia in VDR-knockout mice, may cause the expansion of hypertrophic zone in growth plates. It is therefore likely that optimal levels of Ca2+ and phosphorus, as well as other growth factors, are required for normal growth plate development.
In conjunction with advancing cell differentiation, increases in [Ca2+]e profoundly suppressed RNA levels for the PTH1Rs in mGPC cultures, suggesting that reduced receptor numbers, in the face of stable levels of ligand (PTHrP), may explain why cell differentiation is promoted at high [Ca2+]e. This idea is supported by the finding that overexpressing constitutively active 223hPTH1Rs was able to counteract the effects of high [Ca2+]e on PG accumulation and matrix gene expression. The ability of forskolin and 8-Br-cAMP to alter matrix gene expression and matrix mineralization further suggests that increased cAMP production, through the actions of PTHrP or 223hPTH1R, is a key step in counterbalancing the effects of high [Ca2+]e. These data are consistent with other observations that cAMP/protein kinase A signaling is the major pathway that delays chondrocyte maturation (19, 20).
Despite its impact on PG accumulation and gene expression, overexpressing 223hPTH1Rs, interestingly, did not affect high [Ca2+]e-induced mineralization of mGPC cultures. We hypothesize that this was due to the specific characteristics of the signaling capacity of these constitutively active receptors. Previous studies showed that 223hPTH1Rs transfected in COS-7 cells modestly increased cAMP levels, equivalent to approximately 10% of the levels induced by fully activated wild-type PTH1Rs (18). In mGPCs, adenoviral overexpression of the same mutant (223hPTH1Rs) also modestly increased cAMP production (2- to 3-fold over controls), which is approximately 5–10% of the peak levels induced by acutely incubating mGPCs directly with PTHrP (10–7 M) for short time points. We hypothesize that more cAMP is required to achieve changes in mineralization. The dose-response curves (i.e. pharmacology) for PG vs. mineral accumulation, with respect to intracellular cAMP, may be different. Alternatively, the decline of 223hPTH1Rs expression in the cultures after infection may be affecting our results. When compared qPCR data shown in Figs. 6A and 7D, RNA levels for 223hPTH1Rs in Ad-223hPTH1R-infected cultures decreased by >95% from 3–10 d after infection. We think this was due to a dilution of adenoviral vectors in cells that continued to proliferate and or apoptosis in cells expressing higher levels of 223hPTH1Rs. As matrix mineralization takes place late in cultures, this decrease in expression, plus low constitutive activity of the receptor, may significantly lessen its impact on mineralization. This scenario is supported by the observation that continuously incubating mGPCs with 8-Br-cAMP or forskolin were able to block the high [Ca2+]e-induced mineralization.
It is unclear what causes apoptosis in cells that overexpressed 223hPTH1Rs. It was unlikely due to the propagation of replication-competent viruses that might have contaminated our viral stocks for the following reasons (1). Apoptosis was not observed in cells infected with a slightly lower titer (20 pfu/cell). With this titer, replication-competent viruses, if present, would be expected to propagate and eventually kill the cells as cultures continued to grow, and this did not happen in the cells for up to 21 d (2). Although we observed apoptotic cells when infections were done at at least 50 pfu/cells, nonapoptotic cells in the same cultures continued to grow and differentiate for as long as 21 d. This argues further against the presence of replication-competent viruses.
Does prolonged activation of cAMP signaling by 223hPTH1Rs cause cell death We think this is unlikely because continuously incubating cells with high concentrations of 8-Br-cAMP or forskolin did not induce apoptosis in the cultures. We, therefore, speculate that another signaling pathway(s) activated by 223hPTH1Rs may be involved.
It has been shown by Turner et al. (40) that incubating Chinese hamster ovary cells, which overexpressed opossum PTH1Rs, with PTH (1–34), induced apoptosis in a dose-dependent manner. This observation supports our finding in mGPCs overexpressing 223hPTH1Rs that demonstrated increased apoptosis. We, however, did not observe apoptotic responses in uninfected cultures continuously treated with PTHrP with concentrations as high as 10–7 M. We think this is due to the ability of ligand to desensitize endogenous PTH1Rs (41, 42) and down-regulate the expression of the receptor. This type of postreceptor regulation may protect cells from overstimulation by ligand and apoptosis. These feedback mechanisms may be overwhelmed in mGPCs overexpressing 223hPTH1Rs, as the viral constructs are driven by constitutively active CMV promoters and appear to be unregulated, and the receptors are active without added ligand.
We were surprised to see minimal impact of PTHrP, forskolin, and 8-Br-cAMP on the growth of mGPCs, as cAMP-dependent PTHrP/PTH1R signaling critically promotes chondrocyte proliferation in vivo and in vitro (20). We think this is most likely due to the timing of PTHrP treatment in our experiments. We tested the reagents in postconfluent cultures, whereas other groups studied subconfluent cells (20). According to our data, confluent cultures contain substantially less resting and proliferating cells. Because these cells are prime targets of PTHrP in promoting growth (10, 43), reduction in their numbers could readily lessen the impact of PTHrP on cell proliferation. Deviations in anatomical sites and ages of the animals (prenatal vs. postnatal) from which chondrocytes were isolated may also contribute to the different observations. Nonetheless, our data demonstrate that PTH1R and cAMP signaling can mediate matrix synthesis and mineralization, and gene expression independently of their actions on proliferation.
It remains unclear how changes in [Ca2+]e regulate the expression of several matrix genes and critical components in PTHrP/PTH1R signaling. Our preliminary data indicate changes in [Ca2+]e alter the expression of several transcription factors in Sox, Runx, and AP1 families (Shoback, D., and W. Chang, unpublished data), supporting a transcriptional regulation by Ca2+. We also found that Ca2+ impacts on the expression of molecules in IGF-I, fibroblast growth factor, and TGF- signaling pathways families (Shoback, D., and W. Chang, unpublished data), suggesting regulation through local signaling networks in the growth plate. Thus, extracellular Ca2+ has a broad influence on growth plate development, and this may be necessary to couple information about systemic Ca2+ availability to changes in longitudinal skeletal growth.
Acknowledgments
We acknowledge technical support and critical review of the manuscript by Dr. Dolores Shoback and helpful discussions of Dr. Robert Nissenson in the Endocrine Research Unit of the San Francisco Veterans Affairs Medical Center.
Footnotes
This work was supported by National Institutes of Health Grants AG-21353 and AR-050662 and a Veterans Affairs Merit Review.
Abbreviations: Agg, Aggrecan; ALP, alkaline phosphatase; 8-Br-cAMP, 8-bromo-cAMP; [Ca2+]e, extracellular [Ca2+]; CV, crystal violet; Dig-dUTP, digoxigenin-conjugated dUTP; dUTP, deoxyuridine triphosphate; FCS, fetal calf serum; mGPC, mouse growth plate chondrocyte; OC, osteocalcin; ON, osteonectin; OP, osteopontin; pfu, plaque-forming units; PG, proteoglycan; PTH1R, PTH/PTHrP type I receptor; qPCR, quantitative real-time PCR; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling; VDR, vitamin D receptor.
References
Poole AR 1991 The growth plate: cellular physiology, cartilage assembly, and mineralization. In: Hall BK, Newman SA, eds. Cartilage: molecular aspects. Boca Raton, FL: CRC Press, Inc.; 179–212
Stevens DA, Williams GR 1999 Hormone regulation of chondrocyte differentiation and endochondral bone formation. Mol Cell Endocrinol 151:195–204
Sandell LJ, Sugai JV, Trippel SB 1994 Expression of collagens I, II, X, and XI and aggrecan mRNAs by bovine growth plate chondrocytes in situ. J Orthop Res 12:1–14
Leboy PS, Shapiro IM, Uschmann BD, Oshima O, Lin D 1988 Gene expression in mineralizing chick epiphyseal cartilage. J Biol Chem 263:8515–8520
Mundlos S, Meyer R, Yamada Y, Zabel B 1991 Distribution of cartilage proteoglycan (aggrecan) core protein and link protein gene expression during human skeletal development. Matrix 11:339–346
Nakase T, Takaoka K, Hirakawa K, Hirota S, Takemura T, Onoue H, Takebayashi K, Kitamura Y, Nomura S 1994 Alterations in the expression of osteonectin, osteopontin and osteocalcin mRNAs during the development of skeletal tissues in vivo. Bone Miner 26:109–122
Mundlos S, Schwahn B, Reichert T, Zabel B 1992 Distribution of osteonectin mRNA and protein during human embryonic and fetal development. J Histochem Cytochem 40:283–291
Pacifici M, Oshima O, Fisher LW, Young MF, Shapiro IM, Leboy PS 1990 Changes in osteonectin distribution and levels are associated with mineralization of the chicken tibial growth cartilage. Calcif Tissue Int 47:51–61
McKee MD, Glimcher MJ, Nanci A 1992 High-resolution immunolocalization of osteopontin and osteocalcin in bone and cartilage during endochondral ossification in the chicken tibia. Anat Rec 234:479–492
Kronenberg HM 2003 Developmental regulation of the growth plate. Nature 423:332–336
Kronenberg HM, Lanske B, Kovacs CS, Chung UI, Lee K, Segre GV, Schipani E, Juppner H 1998 Functional analysis of the PTH/PTHrP network of ligands and receptors. Recent Prog Horm Res 53:283–301
Nissenson RA 1998 Parathyroid hormone (PTH)/PTHrP receptor mutations in human chondrodysplasia. Endocrinology 139:4753–4755
Woodard JC, Donovan GA, Fisher LW 1997 Pathogenesis of vitamin (A and D)-induced premature growth-plate closure in calves. Bone 21:171–182
Lee K, Lanske B, Karaplis AC, Deeds JD, Kohno H, Nissenson RA, Kronenberg HM, Segre GV 1996 Parathyroid hormone-related peptide delays terminal differentiation of chondrocytes during endochondral bone development. Endocrinology 137:5109–5118
Strewler GJ 2000 The physiology of parathyroid hormone-related protein. N Engl J Med 342:177–185
Lanske B, Karaplis AC, Lee K, Luz A, Vortkamp A, Pirro A, Karperien M, Defize LH, Ho C, Mulligan RC, Abou-Samra AB, Juppner H, Segre GV, Kronenberg HM 1996 PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273:663–666
Schipani E, Lanske B, Hunzelman J, Luz A, Kovacs CS, Lee K, Pirro A, Kronenberg HM, Juppner H 1997 Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide. Proc Natl Acad Sci USA 94:13689–13694
Schipani E, Langman CB, Parfitt AM, Jensen GS, Kikuchi S, Kooh SW, Cole WG, Juppner H 1996 Constitutively activated receptors for parathyroid hormone and parathyroid hormone-related peptide in Jansen’s metaphyseal chondrodysplasia. N Engl J Med 335:708–714
Guo J, Chung UI, Kondo H, Bringhurst FR, Kronenberg HM 2002 The PTH/PTHrP receptor can delay chondrocyte hypertrophy in vivo without activating phospholipase C. Dev Cell 3:183–194
Ionescu AM, Schwarz EM, Vinson C, Puzas JE, Rosier R, Reynolds PR, O’Keefe RJ 2001 PTHrP modulates chondrocyte differentiation through AP-1 and CREB signaling. J Biol Chem 276:11639–11647
Amling M, Priemel M, Holzmann T, Chapin K, Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987
Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139:4391–4396
Bergstrom WH 1998 When you see rickets, consider calcium deficiency. J Pediatr 133:722–724
Thacher TD 2003 Calcium-deficiency rickets. Endocr Dev 6:105–125
Clark SA, Boass A, Toverud SU 1987 Effects of high dietary contents of calcium and phosphorus on mineral metabolism and growth of vitamin D-deficient suckling and weaned rats. Bone Miner 2:257–270
Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835
Bonen DK, Schmid TM 1991 Elevated extracellular calcium concentrations induce type X collagen synthesis in chondrocyte cultures. J Cell Biol 115:1171–1178
Chang W, Tu C, Pratt S, Chen TH, Shoback D 2002 Extracellular Ca(2+)-sensing receptors modulate matrix production and mineralization in chondrogenic RCJ3.1C5.18 cells. Endocrinology 143:1467–1474
Chang W, Tu C, Bajra R, Komuves L, Miller S, Strewler G, Shoback D 1999 Calcium sensing in cultured chondrogenic RCJ3.1C5.18 cells. Endocrinology 140:1911–1919
Wu S, Palese T, Mishra OP, Delivoria-Papadopoulos M, De Luca F 2004 Effects of Ca2+ sensing receptor activation in the growth plate. FASEB J 18:143–145
Chang W, Pratt S, Chen TH, Shoback D 1999 Protein kinase C activation blocks calcium receptor signaling in Xenopus laevis oocytes. Mol Cell Endocrinol 158:13–23
Chang W, Tu C, Chen TH, Komuves L, Oda Y, Pratt SA, Miller S, Shoback D 1999 Expression and signal transduction of calcium-sensing receptors in cartilage and bone. Endocrinology 140:5883–5893
Chang W, Chen T-H, Pratt SA, Shoback D 2000 Amino acids in the second and third intracellular loops of the parathyroid Ca2+-sensing receptor mediate efficient coupling to phospholipase C. J Biol Chem 275:19955–19963
Chang W, Pratt S, Chen TH, Nemeth E, Huang Z, Shoback D 1998 Coupling of calcium receptors to inositol phosphate and cyclic AMP generation in mammalian cells and Xenopus laevis oocytes and immunodetection of receptor protein by region-specific antipeptide antisera. J Bone Miner Res 13:570–580
Gillies RJ, Didier N, Denton M 1986 Determination of cell number in monolayer cultures. Anal Biochem 159:109–113
MacLean HE, Guo J, Knight MC, Zhang P, Cobrinik D, Kronenberg HM 2004 The cyclin-dependent kinase inhibitor p57(Kip2) mediates proliferative actions of PTHrP in chondrocytes. J Clin Invest 113:1334–1343
van der Eerden BC, Karperien M, Wit JM 2003 Systemic and local regulation of the growth plate. Endocr Rev 24:782–801
van der Eerden BC, Karperien M, Gevers EF, Lowik CW, Wit JM 2000 Expression of Indian hedgehog, parathyroid hormone-related protein, and their receptors in the postnatal growth plate of the rat: evidence for a locally acting growth restraining feedback loop after birth. J Bone Miner Res 15:1045–1055
Donohue MM, Demay MB 2002 Rickets in VDR null mice is secondary to decreased apoptosis of hypertrophic chondrocytes. Endocrinology 143:3691–3694
Turner PR, Mefford S, Christakos S, Nissenson RA 2000 Apoptosis mediated by activation of the G protein-coupled receptor for parathyroid hormone (PTH)/PTH-related protein (PTHrP). Mol Endocrinol 14:241–254
Mitchell J, Goltzman D 1990 Mechanisms of homologous and heterologous regulation of parathyroid hormone receptors in the rat osteosarcoma cell line UMR-106. Endocrinology 126:2650–2660
Pun KK, Ho PW, Nissenson RA, Arnaud CD 1990 Desensitization of parathyroid hormone receptors on cultured bone cells. J Bone Miner Res 5:1193–1200
Kronenberg HM, Chung U 2001 The parathyroid hormone-related protein and Indian hedgehog feedback loop in the growth plate. Novartis Found Symp. 232:144–152; discussion 152–157(Luis Rodriguez, Zhiqiang )
Abstract
An adequate supply of Ca2+ is critical for normal growth plate development. Previous studies suggest that changes in extracellular [Ca2+] ([Ca2+]e) modulate the function of chondrocytes with high [Ca2+]e promoting cell differentiation. In contrast, signal transduction by the PTH/PTHrP type I receptor (PTH1R) slows down chondrocyte differentiation. This study addressed whether changes in [Ca2+]e modulate the differentiation of mouse growth plate chondrocytes by interacting with PTHrP/PTH1Rs. Raising [Ca2+]e from 0.5–3.0 mM dose-dependently promoted the development of mouse growth plate chondrocytes as indicated by decreases in proteoglycan accumulation and in the expression of early differentiation marker genes and by increases in mineral deposition and in the expression of markers of terminal differentiation. The effects of high [Ca2+]e on gene expression and matrix synthesis were blunted by incubating cells with PTHrP and vice versa. High [Ca2+]e also suppressed the expression of PTH1Rs. Chronic stimulation of PTHrP/PTH1R signaling by adenoviral expression of constitutively active human PTH1Rs (223hPTH1Rs) reduced the effects of high [Ca2+]e on proteoglycan synthesis and gene expression. Similar results were seen when we treated cells with forskolin or 8-bromo-cAMP. Taken together, these data support the idea that the pace of chondrocyte differentiation depends on a balance of interactions between PTHrP/PTH1R and extracellular Ca2+ signaling and that high [Ca2+]e promote cell differentiation potentially by reducing the availability of PTH1Rs and the level of cAMP-dependent signal transduction.
Introduction
GROWTH PLATE DEVELOPMENT is the critical step in endochondral bone formation that controls longitudinal growth in vertebrates. In the growth plate, chondrocytes progress through steps of resting, proliferation, maturation, hypertrophy, and terminal differentiation, eventually creating a mineralized matrix that supports skeletal development (1). At each step, chondrocytes produce distinct matrix proteins that signify the stages of differentiation (1, 2). Resting and proliferating chondrocytes produce high levels of the 1-subunit of type II collagen [1(II)] (1, 3, 4) and cartilage-specific proteoglycan core proteins such as aggrecan (Agg) (1, 5). Maturing and upper hypertrophic chondrocytes express the type X collagen [1(X)] and alkaline phosphatase (ALP) as well as the PTH1R. Terminally differentiated chondrocytes in the lower hypertrophic zone (or mineralization zone) demonstrate high levels of the osteopontin (OP) (6), osteonectin (ON) (7, 8), and osteocalcin (OC) (9) expression and deposit Ca2+- and phosphate-containing mineral (i.e. hydroxyapatite) into the matrix (1). At the chondro-osseus junction, vascular invasion takes place, and chondroclasts resorb cartilage matrix, providing a surface for the deposition of bone matrix by osteoblasts. This marks the first step of bone formation.
Chondrocyte differentiation is governed by factors that control the rate of gene expression and phenotypic change so that orderly rates of bone growth are maintained. Both delayed and accelerated cell differentiation in the growth plate can cause growth retardation (10, 11, 12, 13, 14). In vitro and in vivo studies showed that PTHrP and the PTH1R play critical roles in chondrocyte differentiation (15). Deletion of PTHrP (PTHrP–/–) (14) or PTH1R (PTH1R–/–) (16) genes produces a short-limbed dwarfism in mice due to accelerated chondrocyte differentiation, premature mineralization, and shortened growth plates. Growth plate defects in PTHrP–/– mice are partially rescued by overexpressing constitutively active PTH1Rs (223hPTH1Rs) in chondrocytes (17). This mutant receptor, containing a His Arg mutation at position 223, causes constitutive activation of adenylate cyclase with increased cellular cAMP in the absence of ligand (18). cAMP is a key second-messenger that delays chondrocyte differentiation (19, 20). This receptor, interestingly, does not stimulate phospholipase C, another pathway activated by wild-type PTH1Rs (18). These observations support a role for PTHrP and its receptor in delaying growth plate development, likely via the activation of cAMP-dependent signaling cascade. The accelerated chondrocyte differentiation in PTHrP–/– or PTH1R–/– mice also suggests the existence of pathways promoting differentiation which are unmasked in the absence of PTHrP/PTH1R signaling.
Sufficient extracellular Ca2+ is needed for growth plate development (21, 22, 23). Rickets of many etiologies, in which systemic levels of Ca2+ are low, produces abnormally expanded and demineralized growth plates and growth retardation. High Ca2+ diets can correct these defects in several instances (23, 24, 25). Similarly, the rickets and growth plate abnormalities of vitamin D receptor (VDR) knockout mice (21, 26) recede with a high Ca2+ diet. These findings highlight the potential importance of Ca2+ in controlling chondrocyte maturation and function. Because the attendant secondary hyperparathyroidism and renal phosphate wasting are also reversed by high Ca2+ diets in these models, it has been difficult to assess the role of Ca2+ on chondrocyte function in vivo (21). In vitro studies by us and others provide evidence that changes in [Ca2+]e directly modulate key aspects of chondrocyte differentiation—proteoglycan (PG) and matrix accumulation, gene expression, and mineralization (27, 28, 29, 30). These studies further explore effects of Ca2+ on chondrocytes independent of the associated metabolic changes seen in models of rickets.
These studies tested the hypothesis that mouse growth plate chondrocytes (mGPCs) sense changes in [Ca2+]e and that this promotes cell differentiation by counterbalancing PTHrP/PTH1R signal transduction. We found that the rate of chondrocyte differentiation depends on a balance of interactions between PTHrP/PTH1R and extracellular Ca2+ signaling. High [Ca2+]e enhanced mGPC development, whereas activation of PTH1Rs by PTHrP retarded it, by mechanisms that involve cAMP signal transduction. Reduced expression of PTH1Rs, due to high [Ca2+]e, appears to comprise at least part of the mechanism for advancing chondrocyte differentiation.
Materials and Methods
Materials
Fetal calf serum (FCS) was purchased from Tissue Culture Biologicals (Tulare, CA). PTHrP (1–34) was from Bachem (Torrance, CA). Trypsin, collagenase IA, hyaluronidase, deoxyribonuclease II, forskolin, and 8-bromo-cAMP (8-Br-cAMP) were purchased from Sigma-Aldrich (St. Louis, MO). Primers and probes for quantitative real-time PCR (qPCR) were synthesized by Integrated DNA Technologies (Skokie, IL) or purchased from Applied Biosystems Inc. (Foster City, CA). Antisera raised against PTH1Rs were from Upstate Cell Signaling Solution (Lake Placid, NY). Other supplies were from previously noted sources (28).
mGPC and RCJ3.1C5.18 (C5.18) cell cultures
Epiphyseal growth plates from newborn mice (2–4 d old, Black Swiss strain) were obtained and chondrocytes were released by enzymatic digestion. Briefly, growth plate cartilage was first incubated with trypsin (0.25%, wt/vol) in a chondrocyte isolation medium [CIM: DMEM + penicillin (100 μg/ml), streptomycin (100 U/ml), NaHCO3 (45 mM), HEPES (20 mM, pH 7.4), fungizone (0.25 μg/ml), and gentamicin (0.15 mg/ml)] for 15 min at 37 C, followed by digestion with a mixture of (wt/vol) collagenase IA (0.18%), hyaluronidase (0.1%), and deoxyribonuclease II (0.01%) in CIM for 30 min at 37 C with gentle agitation. After removing undigested cartilage with a nylon mesh and small particulate minerals by brief (15 sec) centrifugation at 100 x g, mGPCs in suspension were collected by centrifugation at 400 x g for 10 min. Remaining cartilage was continuously digested with fresh enzymes until no tissue remained. The first cell harvest was routinely discarded, but subsequent harvests were combined and plated (105 cells/cm2) in a chondrocyte maintenance medium [CMM: DMEM + penicillin (100 μg/ml), streptomycin (100 U/ml), and FCS (10%, vol/vol)] at 37 C. For time-course experiments, cells were grown in CMM until confluence and switched to chondrocyte differentiation medium [CDM, -MEM containing 1.8 mM Ca2+ plus ascorbic acid (50 μg/ml) and -glycerol phosphate (5 mM)] for different times. To test the effects of Ca2+, PTHrP, forskolin, and 8-Br-cAMP, confluent cultures were switched to CDM with different [CaCl2] (0.1–4.0 mM) in the presence or absence of the reagents to be tested for different times. Chondrogenic C5.18 cells were cultured as previously described (28), and the effects of Ca2+ and PTHrP were tested as described above. These studies were approved by the Animal Care Subcommittee of the San Francisco Department of Veterans Affairs Medical Center.
Alcian green, alizarin red, and von Kossa staining
The accumulation of PG and mineral in cultures was assessed by staining with alcian green 2GX and alizarin red S and quantified by absorbance at 340 and 540 nm, respectively (28, 29). Cells cultured on glass coverslips were examined after von Kossa staining to detect mineral deposits, alcian green staining for PG accumulation, and counterstaining with hematoxylin to reveal cell morphology (28, 29).
qPCR
Total RNA was isolated as described (28). First-strand cDNAs were reverse-transcribed with Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA). Genes of interest were quantified using probe-based TaqMan qPCR kits and ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA). A Ct (minimal PCR cycles required to generate a fluorescent signal exceeding a preset threshold) was determined by SDS software (Applied Biosystems) for each gene and normalized to a Ct for a housekeeping gene (L19) determined in parallel. Expression of L19 in mGPCs was unchanged with [Ca2+]e, PTHrP, forskolin, and 8-Br-cAMP (data not shown). Fluorescent probes and PCR primers for each gene were designed based on published sequences (see Table 1) or purchased. Sequences of commercial primers are not available due to proprietary protection (Applied Biosystems).
Immunoblotting
Lysates from virally infected mGPCs were prepared, and immunoblotting was performed as previously described (31, 32, 33) using antisera directed against PTH1Rs.
Generation of adenoviral stocks and infection of chondrocytes
Adenoviruses carrying the cDNA encoding the 223hPTH1R were made using an Adeno-X Expression System II kit (BD Biosciences, Palo Alto, CA) according to the manufacturer’s instructions. The 223hPTH1R cDNA was provided by Dr. Harald Jueppner (Harvard Medical School, Boston, MA). This cDNA was first subcloned into a pDNR-CMV vector containing two loxP sites flanking the cloning site. The 223hPTH1R cDNA was then transferred by Cre/Lox recombination to an acceptor cosmid containing the human adenovirus type 5 genome lacking E1 and E3 genes, a CMV promoter, and simian virus 40 poly-adenylation sequence, creating the Ad-223hPTH1R construct. After removing nonviral components by PacI restriction digestion, the Ad-223hPTH1R construct was transfected into human embryonic kidney 293 cells, which stably express E1 proteins and allow for replication of viral DNAs and packaging of viruses. Because the E1 and E3 genes are not present in the Ad-223PTH1R construct, the resulting viruses are replication incompetent in other cells. After mass production in human embryonic kidney 293 cells, viral stocks were titered using an Adeno-X rapid titer kit (BD Biosciences) and used to infect mGPCs. Potential contamination of replication-competent viruses in the viral stocks was verified by their ability to propagate in Chinese hamster ovary and C5.18 cells.
For infection, confluent mGPCs were incubated with viruses (titers as specified) for 72 h in CMM and switched to CDM containing different [Ca2+]e with or without PTHrP (10–7 M) for time points specified before samples were prepared for analysis. Titers used in the experiments were calculated based on the numbers of cells seeded at the beginning of culture.
cAMP assay
To determine cAMP content in response to short-term PTHrP treatment, mGPCs were plated in 24-well plates, cultured for 7 or 14 d in CMM, and then treated with different [Ca2+]e in the presence or absence of PTHrP for 15 min. cAMP was quantified by RIA (32, 34) and presented as total cAMP in picomoles per well.
To determine cAMP content in cells overexpressing 223hPTH1Rs, mGPCs were cultured in six-well or 12-well plates for 3–4 d until they were confluent and infected with Ad-223hPTH1R [20–100 plaque-forming units (pfu)/cell] or control Ad-pAX (100 pfu/cell) for 72 h. In parallel, uninfected or Ad-pAX-infected cells were incubated continuously with PTHrP (10–7 M) for 3 d or exposed briefly (15 min) to PTHrP before sample collections. After cAMP was eluted from the cells with ethanol (100%) as described (34), cell number was determined by quantifying crystal violet (CV) staining (35). cAMP content was normalized to levels of CV staining in each well and expressed as fold change over control levels.
CV staining
After eluting cAMP, cells were incubated with CV (0.2%, wt/vol) in 2% ethanol (vol/vol) for 1 h. After four washes with distilled H2O, stain was eluted by Triton X-100 (0.2%, vol/vol) and quantified by absorbance at 590 nm (35).
TUNEL [terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate (dUTP) nick end labeling] staining
Apoptosis was assessed by TUNEL using a commercial Apoptag kit (Oncor, Gaithersburg, MD) according to manufacturer’s instructions. Briefly, mGPCs were cultured on coverslips in 12-well plates until confluence, followed by infections with Ad-223hPTH1R or Ad-pAX or treatments with PTHrP, forskolin, or 8-Br-cAMP for 72 h. Cells on coverslips were fixed with paraformaldehyde (1%) for 10 min and permeabilized in ethanol/acetic acid (2:1, vol/vol) mixture for 5 min at –20 C. Endogenous peroxidase was quenched by H2O2 (3%) in PBS. After preincubation with a reaction buffer (pH 8.0) for 10 min, cells were incubated with a reaction mixture containing terminal deoxynucleotidyl transferase and digoxigenin-conjugated dUTP (Dig-dUTP) for 1 h at 37 C. Nicked DNA labeled with Dig-dUTP was detected by peroxidase-conjugated anti-digoxigenin antibody and 3,3'-diaminobenzidine substrate. After counterstaining with methyl green, cells were examined microscopically.
Results
Differentiation of mGPCs in culture
The characteristics of primary cultures of growth plate chondrocytes from the limbs of newborn mice recapitulate many key steps in growth plate development. Under subconfluent conditions, the cells produced a PG-containing matrix that stained with alcian green (Fig. 1A). The amount of PG increased with time (Fig. 1, B and C) and decreased when the cells began to deposit minerals, indicated by von Kossa staining, at 7–9 d after confluence (Fig. 1D). Thereafter, mineral deposition increased as PG accumulation declined (Fig. 1E). Levels of RNA encoding matrix proteins, assessed by qPCR, correlated with these biochemical changes. RNA levels for early differentiation markers, Agg and 1(II), were highest in subconfluent and confluent cultures and decreased (by >90%) in postconfluent cultures (Fig. 1, F and G). Conversely, expression of late differentiation markers, ALP, 1(X), OP, and OC, were low in subconfluent cells and increased in postconfluent cultures (4- to 300-fold) (Fig. 1, H–K). These observations indicated that mGPCs in culture demonstrate a temporal sequence of gene expression and matrix protein synthesis characteristic of key steps in growth plate differentiation.
Effects of changing [Ca2+]e on cultured mGPCs
We next tested the effects of changing [Ca2+]e on the profiles of differentiation markers in mGPCs. Raising [Ca2+]e dose-dependently decreased PG accumulation (Fig. 2A) with an ID50 (dosage for 50% inhibition) of approximately 2 mM Ca2+ at 11 d after confluence (Fig. 2B). Changes in [Ca2+]e, however, had minimal, if any, effects on PG synthesis in cultures at 4 d after confluence (Fig. 2B), suggesting that the responsiveness of PG production to changing [Ca2+]e increased as cell differentiation proceeded. On the other hand, high [Ca2+]e promoted mineral accumulation in the matrix surrounding the cells (Fig. 2A) in a concentration- and time-dependent manner (Fig. 2C), beginning at d 7 and as shown for d 14, 21, and 28. The EC50 for mineralization was approximately 2–3 mM Ca2+ at 21 and 28 d. These morphological changes were accompanied by changes in expression of several marker genes. In cultures grown at 0.5 mM Ca2+ for 14 d, early differentiation markers, Agg and 1(II) (Fig. 2, D and E), were coexpressed with markers of maturing and early hypertrophic chondrocytes-1(X), ALP, and the PTH1R (Fig. 2, F–H). At this stage of culture, OP, a marker of terminally differentiated chondrocytes, was low compared with its expression at high [Ca2+]e (Fig. 2I). We interpreted these data as either that the cells were slowly transitioning from resting and proliferation to the maturing and hypertrophic stages of differentiation or that this was a mixed population of cells that were at distinct stages of differentiation.
In cultures maintained at normal and high [Ca2+]e (1.0–3.0 mM) for 14 d, RNA levels for Agg and 1(II) (Fig. 2, D and E) were reduced (by up to 72 and 76%, respectively) in a dose-dependent manner (P < 0.01). The effects of high [Ca2+]e on gene expression were observed as early as 3 d after changes in [Ca2+]e and remained evident at d 21 (data not shown). The effects of changing [Ca2+]e on expression of 1(X), ALP, and PTH1Rs were biphasic. Maintaining cells at 1.0 vs. 0.5 mM Ca2+ produced slightly but significantly (P < 0.05) greater levels of RNA encoding these three genes (Fig. 2, F–H), suggesting that cells grown at 1.0 mM Ca2+ were more mature than cells at 0.5 Ca2+ mM after 14 d in culture. Culturing at 3 mM Ca2+ for the same duration of time reduced expression of 1(X), ALP, and PTH1R RNA (P < 0.001) and significantly increased OP expression (Fig. 2I), indicating that the majority of cells at this [Ca2+]e had matured and reached terminal differentiation at which point the expression of 1(X), ALP, and PTH1R genes are typically decreased. Thus, our data indicate that cultured mGPCs were influenced by changes in [Ca2+]e, and high [Ca2+]e tended to promote features of chondrocyte differentiation.
Effects of PTHrP on mGPCs
PTHrP and its receptor play critical roles in pacing chondrocyte differentiation. To examine whether the impact of high [Ca2+]e on this process is influenced by the level of PTHrP/PTH1R signaling, we compared the effects of [Ca2+]e in the presence and absence of PTHrP. Culturing mGPCs with PTHrP (1–34) (10–7 M) for 14 d significantly blunted the ability of high [Ca2+]e to suppress PG accumulation and promote mineral deposition (Fig. 3A; –PTHrP vs. +PTHrP). This was confirmed by right-shifted Ca2+ dose-response curves (Fig. 3, B and C). The EC50 for PG accumulation and mineralization increased from approximately 1.2 to >3 mM Ca2+ with PTHrP treatment (Fig. 3, B and C). PTHrP also blocked the ability of high [Ca2+]e to influence gene expression. In mGPCs maintained at 0.5 mM Ca2+, treatment with PTHrP for 14 d significantly (P < 0.001) increased RNA levels for Agg by 2.5-fold (Fig. 3D) and markedly reduced the expression of 1(X), ALP, and PTH1R—markers of maturing and hypertrophic chondrocytes—by 65–75% (P < 0.001) (Figs. 3, G–I). These changes in gene expression indicate that PTHrP delayed the transition of mGPCs into more mature stages. Incubation with PTHrP also reduced the ability of high [Ca2+]e (3 mM) to inhibit Agg and 1(II) expression (Fig. 3, D and E) and to increase OP RNA levels (Fig. 3F), suggesting that increased PTHrP/PTH1R signaling directly or indirectly counteracted the effects of high [Ca2+]e on cell differentiation. Similar studies were performed with chondrogenic C5.18 cells, and comparable results were obtained (data not shown).
Effects of changing [Ca2+]e and PTHrP on cAMP signal transduction in mGPCs
We noted that in both mGPCs and C5.18 cells even maximal doses of PTHrP (10–7 M) did not completely block the high [Ca2+]e-induced mineralization, inhibition of PG accumulation, and changes in gene expression (Fig. 3 and data not shown). This could be explained by down-regulation of PTH1R expression at high [Ca2+]e (Fig. 2H), leading to a reduction in PTHrP/PTH1R signaling capacity. We cannot, however, rule out the possibilities: 1) that a portion of high [Ca2+]e-induced changes in chondrocyte function are independent of PTHrP/PTH1R signaling; and or 2) that high [Ca2+]e affect the ability of PTH1Rs to activate downstream signal transduction.
To test the latter possibility, we first examined the impact of changing [Ca2+]e on PTHrP-induced signal transduction in these cells. The best-characterized signaling pathway activated by the PTH1R is the G protein (Gs)-activated increase in adenylate cyclase activity causing cAMP accumulation. We assessed the effects of different [Ca2+]e on PTHrP-induced cAMP accumulation. Treating mGPC cells with PTHrP for 15 min dose-dependently increased cAMP production by up to approximately 40-fold with an ED50 of approximately 3 x 10–9 M at 0.5 mM Ca2+ (Fig. 4A). Raising [Ca2+]e from 0.5 to 3, 5, or 10 mM, however, had no significant effect on cAMP production induced by PTHrP (10–9 M) in either 7- or 14-d-old cultures (Fig. 4B), indicating that different [Ca2+]e did not affect the ability of PTH1Rs to couple to adenylate cyclase activation in short-term incubations.
We, however, found that continuous incubation with PTHrP (10–7 M) desensitized responses of mGPCs to that ligand. As shown in Fig. 5, confluent mGPCs cultured in the absence of PTHrP for 3 d responded to a short-term (15 min) treatment of PTHrP (10–7 M) and significantly (P < 0.001) increased cAMP production by approximately 51-fold (, P < 0.001; +PTHrP vs. –PTHrP). In cells continuously cultured with PTHrP (10–7 M) for 3 d, cAMP levels at the end of the treatment (Fig. 5, ; + PTHrP pretreatment/–PTHrP) were indistinguishable from the levels in cultures without PTHrP (Fig. 5, ; –PTHrP pretreatment/–PTHrP). The cells continuously exposed to PTHrP also failed to increase cAMP production in response to an additional dose of PTHrP (Fig. 5, ; –PTHrP vs. +PTHrP), confirming the insensitivity of the cells to the ligand. These data demonstrated that prolonged activation of PTH1Rs down-regulate their signaling capacity. This was likely due at least in part to the down-regulation of PTH1R gene as we demonstrated in Fig. 3I.
Overexpression of 223hPTH1R in mGPCs
In our experiments of long-term growth of mGPCs at high [Ca2+]e, we observed a decrease in PTH1R mRNA (Fig. 2H), suggesting that high [Ca2+]e caused a decrease in receptor number in the differentiated chondrocytes and reduced the ability of PTHrP to mediate signal transduction. To test this possibility, we overexpressed a constitutively active PTH1R by adenoviral vectors and determined whether sustained activation of PTH1R signaling affected the ability of high [Ca2+]e to promote cell differentiation. We used a previously characterized activating human PTH1R (223hPTH1R) that elevates cellular cAMP in the absence of ligand (18). Infecting mGPCs with Ad-223hPTH1R (5–200 pfu/cell) dose-dependently increased RNA levels for the receptor, assessed by qPCR using primers specific to the human PTH1R gene (Fig. 6A). Increased RNA levels were evident at 3 d after infection (Fig. 6A) and lasted for at least 10 d (see Fig. 7D). In contrast, RNA levels for endogenous PTH1Rs (mPTH1R) were significantly decreased in the 223hPTH1R-expressing mGPCs (Fig. 6B), assessed with primers specific to the mouse PTH1R gene, indicating that signaling activated by the 223hPTH1Rs could down-regulate the expression of the endogenous PTH1R gene. Despite the decreased expression of endogenous receptors, Western blotting, using antisera immunoreactive with both mouse and human PTH1Rs, showed an overall increase in protein expression in the cells infected with Ad-223hPTH1R at titers of at least 20 pfu/cell (Fig. 6C).
Constitutive activity of the 223hPTH1R was confirmed by increases in cAMP levels in the infected mGPCs (Fig. 6D, ). In mGPCs 3 d after infection with Ad-223hPTH1R (50 and 100 pfu/cell), we observed modest but significant (P < 0.01) increases in cAMP accumulation by 1.8- and 2.9-fold, respectively, compared with cells that were infected with control viruses (100 pfu/cell) (Fig. 6D, ). In control cells, incubation with PTHrP for 15 min (Fig. 6D, +PTHrP) increased cAMP production by approximately 47-fold, indicating that the procedure that we used to infect mGPCs did not impact on the responsiveness of mGPCs to PTHrP.
Effects of 223hPTH1R on the differentiation of mGPCs
We next examined the effects of overexpressing 223hPTH1Rs on high [Ca2+]e-induced chondrocyte differentiation. In control cells infected with Ad-pAX (50 pfu/cell), raising [Ca2+]e for 7 d suppressed RNA levels for early differentiation markers, Agg by 82% and 1(II) by 60% at 3.0 mM vs. 0.5 mM Ca2+ (Fig. 7, A and B; ), and increased expression of the terminal differentiation marker OP by 10-fold at 3.0 mM vs. 0.5 mM Ca2+ (Fig. 7C; ) with a potency comparable to that in noninfected cells (see Fig. 2). The effects of high [Ca2+]e (i.e. 3 mM) on the expression of the above matrix genes, however, were significantly altered in cells infected with Ad-223hPTH1R (Fig. 7, A–C; ). RNA levels for Agg and 1(II) in the 223hPTH1R-expressing cells cultured at 3 mM Ca2+ were 2.6- and 1.5-fold higher than the levels in control cells grown at 0.5 mM Ca2+ (P < 0.001) and 12- and 4-fold higher than the levels in control cells cultured at 3 mM Ca2+ (P < 0.001), respectively. This indicates that constitutive activation of PTH1R signaling counteracted in a dramatic way the effects of high [Ca2+]e on the expression of the matrix genes, particularly Agg and 1(II). Overexpression of 223hPTH1Rs had modest (30% inhibition) but significant (P < 0.05) impact on high [Ca2+]e-induced OP expression (Fig. 7C). Expression of 223hPTH1Rs remained evident in cells at this stage of culture (10 d after infection) (Fig. 7D). Its levels, however, were substantially lower (<5% of the level at 3 d after infection; Fig. 6A).
Adenoviral expression of 223hPTH1Rs also blocked the ability of high [Ca2+]e to inhibit PG accumulation as indicated by alcian green staining shown in Fig. 8A. In control cultures, raising [Ca2+]e for 10 d dose-dependently suppressed PG content by up to 35% (Fig. 8A, ), but these changes in [Ca2+]e had no effect on cultures infected with Ad-223hPTH1R (Fig. 8A, ). Incubating the control cells with PTHrP (10–7 M) mimicked the effects of the infection with Ad-223hPTH1Rs on PG accumulation as expected (Fig. 8A, +PTHrP), suggesting that overexpressing 223hPTH1Rs produced sufficient signaling and downstream responses to counteract the effects of high [Ca2+]e on PG synthesis.
Interestingly, in cultures infected with Ad-223hPTH1Rs (Fig. 8B, ), raising [Ca2+]e for 14 d enhanced mineral deposition with a potency similar to that in cells infected with control viruses (Fig. 8B, ). The high [Ca2+]e-induced mineralization in both populations of cells was blocked by coincubation with PTHrP (10–7 M) during the Ca2+ treatment (Fig. 8B, +PTHrP). This suggests that signaling by 223hPTH1Rs was insufficient to block high [Ca2+]e-induced mineralization unless PTHrP was added to cultures.
Effects of 223hPTH1R on the growth of mGPCs
In vitro (20) and in vivo (36) studies suggest that activation of PTHrP/PTH1R signaling enhances the proliferation of chondrocytes. We, therefore, tested whether overexpressing 223hPTH1R increases the growth of mGPCs. On the contrary, we found that in mGPC cultures infected with Ad-223hPTH1R (50 and 100 pfu/cell), cell numbers decreased by approximately 45 and 70% (Fig. 9A, ) (P < 0.01), respectively, when compared with those of cultures infected with control viruses (Ad-pAX, 100 pfu/cell) (Fig. 9A, ).
The 223hPTH1R has been proposed to activate adenylate cyclase activity predominantly and not phospholipase C (18). We tested whether sustained activation of this signaling network was the cause for the decreased cell numbers in Ad-223hPTH1R-infected cultures. Continuous incubation with either 8-Br-cAMP (10–3 M) or forskolin (10–4 M), two activators of cAMP-mediated signaling, did not decrease the number of cells (Fig. 9B). Nor did continuous incubation with PTHrP (10–7 M) impact on the growth of mGPCs (Fig. 9B). In contrast, we observed a modest (20%) but significant (P < 0.05) increase in cell numbers in the cultures treated with forskolin (Fig. 9B).
We further tested whether the decreased cell numbers in mGPCs overexpressing 223hPTH1Rs were due to apoptosis by performing TUNEL staining (Fig. 9, C–H). In mGPCs 3 d after infection with Ad-223hPTH1R (50 and 100 pfu/cell) (Fig. 9, D and E), we observed increased numbers of apoptotic cells as indicated by brown 3,3'-diaminobenzidine-TUNEL staining (arrows). In cultures infected with control vectors (Fig. 9C) or grown with PTHrP (10–7 M), forskolin (10–4 M), or 8-Br-cAMP (10–3 M) for 3 d (Fig. 9, F–H), there were rare apoptotic cells. These observations suggest that constitutively activating 223hPTH1Rs induced apoptosis through pathways independent of cAMP signaling.
Effects of cAMP signaling on the differentiation of mGPCs
To test whether direct activation of cAMP signaling mimics the effects of PTHrP and overexpressing 223hPTH1R on the differentiation of mGPCs, we assessed matrix gene expression and mineralization in cultures treated with forskolin (10–4 M) or 8-Br-cAMP (10–3 M). In mGPCs cultured without these reagents, raising [Ca2+]e from 0.5–1.5 and 3.0 mM for 10 d decreased RNA levels for Agg by 74% and 1(II) by 60% and increased OP expression by 2.5-fold (Fig. 10A, ) as we demonstrated in the earlier experiments. Treating cells with forskolin significantly (P < 0.001) blocked the ability of high [Ca2+]e (i.e. 3 mM) to inhibit the expression of Agg and 1(II) and to increase OP expression (Fig. 10A, ). Incubating cells with 8-Br-cAMP profoundly impacted on matrix mineralization. Raising [Ca2+]e from 0.5 to 2 and 3 mM for 10 d in the absence of 8-Br-cAMP increased mineral deposition by 2.5-fold in mGPC cultures as expected (Fig. 10B, ). The effects of [Ca2+]e on the mineralization were blocked in cultures treated with 8-Br-cAMP (Fig. 10B, ). Blockade of high [Ca2+]e-induced mineralization was also observed in cultures treated with forskolin (data not shown). Taken together, these data support the idea that activation of cAMP signal transduction plays a role in counteracting the differentiation-promoting actions of high [Ca2+]e.
Discussion
Chondrocytes in the growth plate differentiate at an orderly rate that maintains normal bone growth. Both delay in and acceleration of chondrocyte differentiation cause growth retardation (10, 15). The PTHrP/PTH1R/Indian hedgehog feedback mechanism is critical in slowing down and thereby pacing chondrocyte differentiation (10, 15). The factors that counterbalance the PTHrP/PTH1R/Indian hedgehog system and promote chondrocyte transit into the steps of terminal differentiation remain to be proven, but there are several candidate molecules such as IGF-I, IGF-I receptor, VDR, thyroid hormone, and TGF- (37). Our previous studies suggested that extracellular Ca2+, possibly acting via extracellular Ca2+-sensing receptors, contributes to controlling the rate of differentiation in C5.18 cells (28). The experiments reported herein addressed whether extracellular Ca2+ interacts with or influences the PTHrP/PTH1R signaling pathways. This is a first step in determining a mechanism by which extracellular Ca2+ might influence growth plate differentiation. We confirmed that PTHrP and its receptor regulate chondrocyte maturation and differentiation, which is well established from in vivo models (10) and demonstrated that changes in [Ca2+]e significantly modify this regulation.
We found that cultured mGPCs recapitulate key steps in the progressive development of the growth plate. These cells express early differentiation markers shortly after plating and markers of terminally differentiated chondrocytes in later cultures. These characteristics indicate this is a suitable model for studying factors that modulate the time-dependent progression of chondrocyte differentiation. Raising [Ca2+]e advances cell differentiation by suppressing the expression of early chondrogenic genes [i.e. Agg and 1(II)] and PG synthesis and enhancing the expression of terminal differentiation markers [i.e. OP, OC, and ON] and mineral deposition. These observations are compatible with our previous findings with the chondrogenic C5.18 cell model system (28) in which raising [Ca2+]e promoted cell differentiation. Consistent with this differentiation scheme, Wu and colleagues (30) showed that high [Ca2+]e enhanced terminal differentiation of hypertrophic chondrocytes in bone rudiment cultures. Bonen and Schmid (27) also reported that a shorter incubation for 3 d of chick embryonic growth plate chondrocytes at 5 mM Ca2+ increased the synthesis and accumulation of 1(X) protein in the matrix, indicating that these cells displayed signs of a more mature phenotype shortly after treatment with high [Ca2+]e. According to the expression pattern of 1(X) in growth plates—beginning in maturing chondrocytes, peaking in upper hypertrophic chondrocytes, and decreasing in lower hypertrophic chondrocytes in the mineralization zone (38)—this increase in 1(X) expression indicated a transition of the cells into the stages of maturation and early hypertrophy, but not yet into the terminal differentiation. Although long-term treatment with high [Ca2+]e was not performed in the study of Bonen and Schmid, we predict that high [Ca2+]e would eventually reduce 1(X) content in the cultures as cells become terminally differentiated after long-term culture. This prediction is based on our observations that high [Ca2+]e ( 3 mM) suppressed the RNA levels for 1(X) in mGPCs cultured for 14 d. Together these observations lend support to a model in which low systemic Ca2+ levels participate in the delay of hypertrophic chondrocyte differentiation in conditions like rickets. Our study does not exclude important roles for the other systemic derangements (hyperparathyroidism and hypophosphatemia) that accompany rickets and contribute to the skeletal phenotype. Studies by Donohue and Demay (39) in fact suggest that reduced cell apoptosis, as a result of hypophosphatemia in VDR-knockout mice, may cause the expansion of hypertrophic zone in growth plates. It is therefore likely that optimal levels of Ca2+ and phosphorus, as well as other growth factors, are required for normal growth plate development.
In conjunction with advancing cell differentiation, increases in [Ca2+]e profoundly suppressed RNA levels for the PTH1Rs in mGPC cultures, suggesting that reduced receptor numbers, in the face of stable levels of ligand (PTHrP), may explain why cell differentiation is promoted at high [Ca2+]e. This idea is supported by the finding that overexpressing constitutively active 223hPTH1Rs was able to counteract the effects of high [Ca2+]e on PG accumulation and matrix gene expression. The ability of forskolin and 8-Br-cAMP to alter matrix gene expression and matrix mineralization further suggests that increased cAMP production, through the actions of PTHrP or 223hPTH1R, is a key step in counterbalancing the effects of high [Ca2+]e. These data are consistent with other observations that cAMP/protein kinase A signaling is the major pathway that delays chondrocyte maturation (19, 20).
Despite its impact on PG accumulation and gene expression, overexpressing 223hPTH1Rs, interestingly, did not affect high [Ca2+]e-induced mineralization of mGPC cultures. We hypothesize that this was due to the specific characteristics of the signaling capacity of these constitutively active receptors. Previous studies showed that 223hPTH1Rs transfected in COS-7 cells modestly increased cAMP levels, equivalent to approximately 10% of the levels induced by fully activated wild-type PTH1Rs (18). In mGPCs, adenoviral overexpression of the same mutant (223hPTH1Rs) also modestly increased cAMP production (2- to 3-fold over controls), which is approximately 5–10% of the peak levels induced by acutely incubating mGPCs directly with PTHrP (10–7 M) for short time points. We hypothesize that more cAMP is required to achieve changes in mineralization. The dose-response curves (i.e. pharmacology) for PG vs. mineral accumulation, with respect to intracellular cAMP, may be different. Alternatively, the decline of 223hPTH1Rs expression in the cultures after infection may be affecting our results. When compared qPCR data shown in Figs. 6A and 7D, RNA levels for 223hPTH1Rs in Ad-223hPTH1R-infected cultures decreased by >95% from 3–10 d after infection. We think this was due to a dilution of adenoviral vectors in cells that continued to proliferate and or apoptosis in cells expressing higher levels of 223hPTH1Rs. As matrix mineralization takes place late in cultures, this decrease in expression, plus low constitutive activity of the receptor, may significantly lessen its impact on mineralization. This scenario is supported by the observation that continuously incubating mGPCs with 8-Br-cAMP or forskolin were able to block the high [Ca2+]e-induced mineralization.
It is unclear what causes apoptosis in cells that overexpressed 223hPTH1Rs. It was unlikely due to the propagation of replication-competent viruses that might have contaminated our viral stocks for the following reasons (1). Apoptosis was not observed in cells infected with a slightly lower titer (20 pfu/cell). With this titer, replication-competent viruses, if present, would be expected to propagate and eventually kill the cells as cultures continued to grow, and this did not happen in the cells for up to 21 d (2). Although we observed apoptotic cells when infections were done at at least 50 pfu/cells, nonapoptotic cells in the same cultures continued to grow and differentiate for as long as 21 d. This argues further against the presence of replication-competent viruses.
Does prolonged activation of cAMP signaling by 223hPTH1Rs cause cell death We think this is unlikely because continuously incubating cells with high concentrations of 8-Br-cAMP or forskolin did not induce apoptosis in the cultures. We, therefore, speculate that another signaling pathway(s) activated by 223hPTH1Rs may be involved.
It has been shown by Turner et al. (40) that incubating Chinese hamster ovary cells, which overexpressed opossum PTH1Rs, with PTH (1–34), induced apoptosis in a dose-dependent manner. This observation supports our finding in mGPCs overexpressing 223hPTH1Rs that demonstrated increased apoptosis. We, however, did not observe apoptotic responses in uninfected cultures continuously treated with PTHrP with concentrations as high as 10–7 M. We think this is due to the ability of ligand to desensitize endogenous PTH1Rs (41, 42) and down-regulate the expression of the receptor. This type of postreceptor regulation may protect cells from overstimulation by ligand and apoptosis. These feedback mechanisms may be overwhelmed in mGPCs overexpressing 223hPTH1Rs, as the viral constructs are driven by constitutively active CMV promoters and appear to be unregulated, and the receptors are active without added ligand.
We were surprised to see minimal impact of PTHrP, forskolin, and 8-Br-cAMP on the growth of mGPCs, as cAMP-dependent PTHrP/PTH1R signaling critically promotes chondrocyte proliferation in vivo and in vitro (20). We think this is most likely due to the timing of PTHrP treatment in our experiments. We tested the reagents in postconfluent cultures, whereas other groups studied subconfluent cells (20). According to our data, confluent cultures contain substantially less resting and proliferating cells. Because these cells are prime targets of PTHrP in promoting growth (10, 43), reduction in their numbers could readily lessen the impact of PTHrP on cell proliferation. Deviations in anatomical sites and ages of the animals (prenatal vs. postnatal) from which chondrocytes were isolated may also contribute to the different observations. Nonetheless, our data demonstrate that PTH1R and cAMP signaling can mediate matrix synthesis and mineralization, and gene expression independently of their actions on proliferation.
It remains unclear how changes in [Ca2+]e regulate the expression of several matrix genes and critical components in PTHrP/PTH1R signaling. Our preliminary data indicate changes in [Ca2+]e alter the expression of several transcription factors in Sox, Runx, and AP1 families (Shoback, D., and W. Chang, unpublished data), supporting a transcriptional regulation by Ca2+. We also found that Ca2+ impacts on the expression of molecules in IGF-I, fibroblast growth factor, and TGF- signaling pathways families (Shoback, D., and W. Chang, unpublished data), suggesting regulation through local signaling networks in the growth plate. Thus, extracellular Ca2+ has a broad influence on growth plate development, and this may be necessary to couple information about systemic Ca2+ availability to changes in longitudinal skeletal growth.
Acknowledgments
We acknowledge technical support and critical review of the manuscript by Dr. Dolores Shoback and helpful discussions of Dr. Robert Nissenson in the Endocrine Research Unit of the San Francisco Veterans Affairs Medical Center.
Footnotes
This work was supported by National Institutes of Health Grants AG-21353 and AR-050662 and a Veterans Affairs Merit Review.
Abbreviations: Agg, Aggrecan; ALP, alkaline phosphatase; 8-Br-cAMP, 8-bromo-cAMP; [Ca2+]e, extracellular [Ca2+]; CV, crystal violet; Dig-dUTP, digoxigenin-conjugated dUTP; dUTP, deoxyuridine triphosphate; FCS, fetal calf serum; mGPC, mouse growth plate chondrocyte; OC, osteocalcin; ON, osteonectin; OP, osteopontin; pfu, plaque-forming units; PG, proteoglycan; PTH1R, PTH/PTHrP type I receptor; qPCR, quantitative real-time PCR; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling; VDR, vitamin D receptor.
References
Poole AR 1991 The growth plate: cellular physiology, cartilage assembly, and mineralization. In: Hall BK, Newman SA, eds. Cartilage: molecular aspects. Boca Raton, FL: CRC Press, Inc.; 179–212
Stevens DA, Williams GR 1999 Hormone regulation of chondrocyte differentiation and endochondral bone formation. Mol Cell Endocrinol 151:195–204
Sandell LJ, Sugai JV, Trippel SB 1994 Expression of collagens I, II, X, and XI and aggrecan mRNAs by bovine growth plate chondrocytes in situ. J Orthop Res 12:1–14
Leboy PS, Shapiro IM, Uschmann BD, Oshima O, Lin D 1988 Gene expression in mineralizing chick epiphyseal cartilage. J Biol Chem 263:8515–8520
Mundlos S, Meyer R, Yamada Y, Zabel B 1991 Distribution of cartilage proteoglycan (aggrecan) core protein and link protein gene expression during human skeletal development. Matrix 11:339–346
Nakase T, Takaoka K, Hirakawa K, Hirota S, Takemura T, Onoue H, Takebayashi K, Kitamura Y, Nomura S 1994 Alterations in the expression of osteonectin, osteopontin and osteocalcin mRNAs during the development of skeletal tissues in vivo. Bone Miner 26:109–122
Mundlos S, Schwahn B, Reichert T, Zabel B 1992 Distribution of osteonectin mRNA and protein during human embryonic and fetal development. J Histochem Cytochem 40:283–291
Pacifici M, Oshima O, Fisher LW, Young MF, Shapiro IM, Leboy PS 1990 Changes in osteonectin distribution and levels are associated with mineralization of the chicken tibial growth cartilage. Calcif Tissue Int 47:51–61
McKee MD, Glimcher MJ, Nanci A 1992 High-resolution immunolocalization of osteopontin and osteocalcin in bone and cartilage during endochondral ossification in the chicken tibia. Anat Rec 234:479–492
Kronenberg HM 2003 Developmental regulation of the growth plate. Nature 423:332–336
Kronenberg HM, Lanske B, Kovacs CS, Chung UI, Lee K, Segre GV, Schipani E, Juppner H 1998 Functional analysis of the PTH/PTHrP network of ligands and receptors. Recent Prog Horm Res 53:283–301
Nissenson RA 1998 Parathyroid hormone (PTH)/PTHrP receptor mutations in human chondrodysplasia. Endocrinology 139:4753–4755
Woodard JC, Donovan GA, Fisher LW 1997 Pathogenesis of vitamin (A and D)-induced premature growth-plate closure in calves. Bone 21:171–182
Lee K, Lanske B, Karaplis AC, Deeds JD, Kohno H, Nissenson RA, Kronenberg HM, Segre GV 1996 Parathyroid hormone-related peptide delays terminal differentiation of chondrocytes during endochondral bone development. Endocrinology 137:5109–5118
Strewler GJ 2000 The physiology of parathyroid hormone-related protein. N Engl J Med 342:177–185
Lanske B, Karaplis AC, Lee K, Luz A, Vortkamp A, Pirro A, Karperien M, Defize LH, Ho C, Mulligan RC, Abou-Samra AB, Juppner H, Segre GV, Kronenberg HM 1996 PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273:663–666
Schipani E, Lanske B, Hunzelman J, Luz A, Kovacs CS, Lee K, Pirro A, Kronenberg HM, Juppner H 1997 Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide. Proc Natl Acad Sci USA 94:13689–13694
Schipani E, Langman CB, Parfitt AM, Jensen GS, Kikuchi S, Kooh SW, Cole WG, Juppner H 1996 Constitutively activated receptors for parathyroid hormone and parathyroid hormone-related peptide in Jansen’s metaphyseal chondrodysplasia. N Engl J Med 335:708–714
Guo J, Chung UI, Kondo H, Bringhurst FR, Kronenberg HM 2002 The PTH/PTHrP receptor can delay chondrocyte hypertrophy in vivo without activating phospholipase C. Dev Cell 3:183–194
Ionescu AM, Schwarz EM, Vinson C, Puzas JE, Rosier R, Reynolds PR, O’Keefe RJ 2001 PTHrP modulates chondrocyte differentiation through AP-1 and CREB signaling. J Biol Chem 276:11639–11647
Amling M, Priemel M, Holzmann T, Chapin K, Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987
Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139:4391–4396
Bergstrom WH 1998 When you see rickets, consider calcium deficiency. J Pediatr 133:722–724
Thacher TD 2003 Calcium-deficiency rickets. Endocr Dev 6:105–125
Clark SA, Boass A, Toverud SU 1987 Effects of high dietary contents of calcium and phosphorus on mineral metabolism and growth of vitamin D-deficient suckling and weaned rats. Bone Miner 2:257–270
Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835
Bonen DK, Schmid TM 1991 Elevated extracellular calcium concentrations induce type X collagen synthesis in chondrocyte cultures. J Cell Biol 115:1171–1178
Chang W, Tu C, Pratt S, Chen TH, Shoback D 2002 Extracellular Ca(2+)-sensing receptors modulate matrix production and mineralization in chondrogenic RCJ3.1C5.18 cells. Endocrinology 143:1467–1474
Chang W, Tu C, Bajra R, Komuves L, Miller S, Strewler G, Shoback D 1999 Calcium sensing in cultured chondrogenic RCJ3.1C5.18 cells. Endocrinology 140:1911–1919
Wu S, Palese T, Mishra OP, Delivoria-Papadopoulos M, De Luca F 2004 Effects of Ca2+ sensing receptor activation in the growth plate. FASEB J 18:143–145
Chang W, Pratt S, Chen TH, Shoback D 1999 Protein kinase C activation blocks calcium receptor signaling in Xenopus laevis oocytes. Mol Cell Endocrinol 158:13–23
Chang W, Tu C, Chen TH, Komuves L, Oda Y, Pratt SA, Miller S, Shoback D 1999 Expression and signal transduction of calcium-sensing receptors in cartilage and bone. Endocrinology 140:5883–5893
Chang W, Chen T-H, Pratt SA, Shoback D 2000 Amino acids in the second and third intracellular loops of the parathyroid Ca2+-sensing receptor mediate efficient coupling to phospholipase C. J Biol Chem 275:19955–19963
Chang W, Pratt S, Chen TH, Nemeth E, Huang Z, Shoback D 1998 Coupling of calcium receptors to inositol phosphate and cyclic AMP generation in mammalian cells and Xenopus laevis oocytes and immunodetection of receptor protein by region-specific antipeptide antisera. J Bone Miner Res 13:570–580
Gillies RJ, Didier N, Denton M 1986 Determination of cell number in monolayer cultures. Anal Biochem 159:109–113
MacLean HE, Guo J, Knight MC, Zhang P, Cobrinik D, Kronenberg HM 2004 The cyclin-dependent kinase inhibitor p57(Kip2) mediates proliferative actions of PTHrP in chondrocytes. J Clin Invest 113:1334–1343
van der Eerden BC, Karperien M, Wit JM 2003 Systemic and local regulation of the growth plate. Endocr Rev 24:782–801
van der Eerden BC, Karperien M, Gevers EF, Lowik CW, Wit JM 2000 Expression of Indian hedgehog, parathyroid hormone-related protein, and their receptors in the postnatal growth plate of the rat: evidence for a locally acting growth restraining feedback loop after birth. J Bone Miner Res 15:1045–1055
Donohue MM, Demay MB 2002 Rickets in VDR null mice is secondary to decreased apoptosis of hypertrophic chondrocytes. Endocrinology 143:3691–3694
Turner PR, Mefford S, Christakos S, Nissenson RA 2000 Apoptosis mediated by activation of the G protein-coupled receptor for parathyroid hormone (PTH)/PTH-related protein (PTHrP). Mol Endocrinol 14:241–254
Mitchell J, Goltzman D 1990 Mechanisms of homologous and heterologous regulation of parathyroid hormone receptors in the rat osteosarcoma cell line UMR-106. Endocrinology 126:2650–2660
Pun KK, Ho PW, Nissenson RA, Arnaud CD 1990 Desensitization of parathyroid hormone receptors on cultured bone cells. J Bone Miner Res 5:1193–1200
Kronenberg HM, Chung U 2001 The parathyroid hormone-related protein and Indian hedgehog feedback loop in the growth plate. Novartis Found Symp. 232:144–152; discussion 152–157(Luis Rodriguez, Zhiqiang )