Main

Longitudinal growth of the long bones occurs by a process termed endochondral bone formation. In the growth plate, cartilage is formed by chondrocyte proliferation, hypertrophy, and extracellular matrix secretion. The newly formed cartilage is then remodeled into bone at the adjacent metaphysis (1).

Recent evidence suggests that longitudinal bone growth is regulated by CNP (2–4). Unlike atrial natriuretic peptide and BNP, which act as hormones inducing natriuresis and vasodilation, CNP is thought to act primarily as a paracrine factor regulating growth and differentiation in a variety of tissues (3–7). Natriuretic peptides interact with guanylyl cyclase-coupled natriuretic peptide receptors, NPR-A and NPR-B, leading to increased intracellular production of cGMP (8). Some of the effects of cGMP, in turn, are mediated by activation of two known protein kinases, cGMP-dependent protein kinase I and II (9).

Several lines of evidence suggest that CNP regulates longitudinal bone growth. Both CNP and its high-affinity receptor, NPR-B, are expressed in the fetal mouse tibia (2–3). In cultured fetal mouse tibiae, CNP reportedly stimulated cGMP production and accelerated longitudinal growth (4). These effects were blocked by a NPR antagonist. The antagonist also decreased cGMP concentration and growth in the absence of exogenous CNP, suggesting that endogenous CNP serves to stimulate longitudinal bone growth (4).

To investigate the underlying cellular mechanisms, we used a fetal rat metatarsal organ culture system. Fetal bones were cultured with varying concentrations of either CNP or 8-Br-cGMP. Because the rate of longitudinal bone growth depends on the rate of chondrogenesis, we assessed the effects of these compounds on each component of chondrogenesis: chondrocyte proliferation, chondrocyte hypertrophy, and cartilage matrix synthesis.

METHODS

Organ culture.

The second, third, and fourth metatarsal bone rudiments were dissected from Sprague-Dawley rat fetuses at E20 and were cultured individually in 24-well plates (10). Each well contained 0.5 mL of minimal essential medium (GIBCO BRL, Gaithersburg, MD) supplemented with 0.05 mg/mL ascorbic acid (GIBCO BRL), 0.3 mg/mL L-glutamine (GIBCO BRL), 1 mM sodium glycerophosphate (Sigma Chemical Co., St. Louis, MO), 0.2% BSA (Sigma Chemical Co.), 100 U/mL penicillin plus 100 μg/mL streptomycin (GIBCO BRL), and 1% PBS. The culture medium was changed daily. 8-Br-cGMP (Sigma Chemical Co.) or C-type natriuretic peptide (1–22) (Sigma Chemical Co.) was added to the culture medium at the indicated concentrations on all 3 d of culture. Plates were incubated in humidified air containing 5% CO2 at 37°C. The National Institute of Child Health and Human Development, Animal Care and Use Committee, approved animal procedures. Animals were cared for in accordance with the Guide for the Care and Use of Laboratory Animals [DHEW Publication (National Institutes of Health) 85–23, revised 1988].

Measurement of longitudinal growth.

The length of each bone rudiment was measured daily by using an eyepiece micrometer in a dissecting microscope (10). Culture medium was briefly removed immediately before each measurement.

Assessment of cell proliferation.

After fetal rat metatarsal bones had been cultured in serum-free medium for 3 d, cell proliferation was assessed by measuring 3H-labeled thymidine incorporation into newly synthesized DNA as previously described (11). The 3H-labeled thymidine (Amersham, specific activity 25 Ci/mmol) was added to the culture medium at a concentration of 5 μCi/mL, and the rudiments were incubated for an additional 3 h. The metatarsals were then washed three times for 10 min each with minimal essential medium and solubilized with 1.0 mL NCS-II Tissue Solubilizer (Amersham) 0.5 N solution for 22 h. Ten milliliters of scintillation liquid (Biosafe NA, Research Product International Corp., Mount Prospect, IL) and 40 μL of 10% acetic acid were added to each sample. Liquid scintillation counting was delayed at least 7 d to minimize chemiluminescence.

Incorporation of [3H]thymidine was localized to specific populations of chondrocytes by autoradiography. Bone rudiments were labeled with [3H]thymidine as described above and then fixed in 10% phosphate-buffered formalin. After routine processing, samples were embedded in paraffin, and longitudinal 5-μm sections were prepared. Autoradiography was performed by standard techniques, with a 1-wk exposure time. Sections were counterstained with hematoxylin and eosin. A single observer blinded to treatment category determined labeling index (labeled cells/total cells). Using a microscope with a 10× objective, video camera, and video display, the observer selected the largest possible rectangle that was completely encompassed within the resting zone or the proliferative zone of interest. The labeled cells present within that rectangle were counted. The number of unlabeled cells was determined with a 40× objective, using a smaller rectangular counting region (one-forth the length and width of the larger rectangle).

Assessment of cellular hypertrophy.

Metatarsals were fixed in 10% phosphate-buffered formalin for 24 h. After routine processing, the metatarsals were embedded in plastic. Longitudinal 5-μm sections were stained with toluidine blue. The number of hypertrophic cells (defined by a lacunar height along the longitudinal axis ≥ 9 μm) per section was determined (10). Relative hypertrophic cell size was assessed by measuring the height (along the long axis of the bone) of the two largest-appearing hypertrophic cells in each of the two hypertrophic zones of each metatarsal section. For each metatarsal bone, three sections were examined and the cell heights averaged. The largest chondrocytes were selected because smaller-appearing chondrocytes may represent tangential sections through the cell. All quantitative histology was performed by a single observer blinded to the treatment category.

Assessment of glycosaminoglycan synthesis.

Glycosaminoglycan synthesis was assessed by measuring 35SO4 incorporation as previously described (12). Bones were labeled with 5 μCi/mL Na235SO4 (Amersham, specific activity up to 100 mCi/mmol) for 3 h. The bone rudiments were then rinsed three times for 10 min each with minimal essential medium and then digested in 1.5 mL of fresh medium with 0.3% papain (Sigma Chemical Co.) for 24 h at 60°C. Next, 0.5 mL of 10% cetylpyridinium chloride (Sigma Chemical Co.) in 0.2 M NaCl was added, and the samples were incubated at room temperature for 18 h to precipitate glycosaminoglycans. The precipitate was collected by vacuum filtration through glass filter paper (Whatman Cat. No. 1001090) and washed three times with 1 mL of a solution containing 0.1% cetylpyridinium chloride in 0.2 M NaCl. Samples were then dissolved in 1 mL of 23 N formic acid, and 35SO4 content was determined by liquid scintillation counting.

Statistics.

All data were expressed as mean ± SEM. Statistical significance was determined by ANOVA and posthoc Fisher's protected least significant differences test. [3H]thymidine and 35SO4 incorporation were normalized to control values.

RESULTS

Effects of CNP on longitudinal bone growth.

CNP increased the longitudinal growth rate of cultured fetal rat metatarsal bone (p < 0.01, n = 26/group) (Fig. 1A). At the highest concentration, CNP increased the growth rate by 49% compared with the control group (122 ± 3 versus 82 ± 3 μm/day, mean ± SEM, p < 0.01) (Fig. 1A). At this stage of development, longitudinal bone growth arises primarily from three processes: chondrocyte proliferation, chondrocyte hypertrophy, and cartilage matrix synthesis. We therefore assessed the effects of CNP on each of these three components.

Figure 1
figure 1

Effects of CNP and 8-Br-cGMP on longitudinal bone growth (mean ± SEM). Fetal rat metatarsals (E20) were cultured for 3 d in serum-free medium containing 0–1000 nM CNP (panel A, n = 26/group) or 0–100 μM 8-Br-cGMP (panel B, n = 50/group). Bone length was measured daily with an eyepiece micrometer in a dissecting microscope. Growth rate (μm/day) after 3 days of culture was calculated (*p < 0.05; **p < 0.01 vs control).

Full size image

To assess chondrocyte proliferation, we measured incorporation of 3H-thymidine in metatarsal rudiments after 72 h of culture in the presence of CNP (0, 10, or 1000 nM;n = 26/group). CNP (1000 nM) significantly increased [3H]thymidine incorporation (112 ± 6% of control, p < 0.05). This increase in [3H]thymidine incorporation was not observed at 24 or 48 h of culture (data not shown). To determine the histologic distribution of the proliferative effects, we cultured metatarsals for 72 h in the presence of CNP (0, 10, or 1000 nM), labeled with [3H]thymidine and then performed autoradiography (n = 28–32/group) (Fig. 2, A and B). CNP did not affect the labeling index in the resting zone (Fig. 3A). However, in the proliferative zone, CNP (10 and 1000 nM) induced a significant increase in labeling index (p < 0.005) (Fig. 3C).

Figure 2
figure 2

Autoradiography depicting the effect of CNP and 8-Br-cGMP on [3H]thymidine incorporation in resting zone and proliferative zone chondrocytes. Fetal rat metatarsal bones (E20) were cultured in serum-free medium with 1000 nM CNP (B), 100 μM 8-Br-cGMP (C), or no treatment (A) for 3 d, labeled with [3H]thymidine for 3 h, and fixed in 10% phosphate-buffered formalin. After routine processing, samples were embedded in paraffin, and 5-μm longitudinal sections were prepared. Autoradiography was performed by standard techniques using a 1-week exposure time. RZ, resting zone;PZ, proliferative zone;HZ, hypertrophic zone;OC, primary center of ossification. The primary center of ossification could be distinguished from the hypertrophic zone by the presence of cells incorporating [3H]thymidine, differential staining with Masson-Trichrome stain and toluidine blue stain, and avid 45Ca uptake (data not shown).

Full size image
Figure 3
figure 3

Effect of CNP (A, C) and 8-Br-cGMP (B, D) on [3H]thymidine-labeling index (mean ± SEM), of resting zone (A, B) and proliferative zone (C, D) chondrocytes. Fetal rat metatarsal bones (E20) were cultured in serum-free medium with indicated concentrations of CNP (n = 28–32/group) or 8-Br-cGMP (n = 10–12/group) for 3 d, labeled with [3H]thymidine for 3 h, and prepared for autoradiography as in Figure 2. Labeling index (number of labeled nuclei/total nuclei) was determined by a single observer blinded to the treatment category (*p < 0.02; **p < 0.005 vs control).

Full size image

After 72 h of culture, CNP (10 and 1000 nM) also caused an increase in the number of hypertrophic cells per section (p < 0.01, n = 8–28/group) (Fig. 4A). CNP also increased the relative height of the terminal hypertrophic cell (20 ± 0.3 versus 17 ± 0.5 μm, 1000 nM CNP versus control, p < 0.05, n = 12–28/group) (Fig. 4C).

Figure 4
figure 4

Effect of CNP and 8-Br-cGMP on (mean ± SEM) hypertrophic cell number (A, B) and height (C, D). Fetal rat metatarsals (E20) were cultured for 3 days in serum-free medium with the indicated concentrations of CNP (panels A, C;n = 8–28/group) or 8-Br-cGMP (panels B, D;n = 12–16/group). After routine histologic processing, bones were embedded in plastic. Longitudinal sections (5-μm) were stained with toluidine blue. In A and B, the number of hypertrophic cells (defined by a lacunar height along the longitudinal axis ≥ 9 μm) per section was determined. In C and D, relative hypertrophic cell size was assessed by measuring the height (along the long axis of the bone) of the two largest appearing hypertrophic cells on each side of a metatarsal section. Quantification was performed by a single observer blinded to the treatment category (*p < 0.05; **p < 0.01 vs control).

Full size image

To assess cartilage matrix formation, we measured 35SO4 incorporation into newly synthesized glycosaminoglycans. After 72 h of culture, CNP (1000 nM) increased glycosaminoglycan synthesis (156 ± 14% of control, n = 26/group, p < 0.05) (Fig. 5A). 35SO4 incorporation was also greater in CNP-treated fetal bones at 12 h (166 ± 20% of control, p < 0.005) and at 24 h (106 ± 15% of control, P = NS).

Figure 5
figure 5

Effect of CNP (A) and 8-Br-cGMP (B) on glycosaminoglycan synthesis. Fetal rat metatarsal bones (E20) were cultured in serum-free medium with indicated concentrations of CNP (n = 26/group) or 8-Br-cGMP (n = 11–20/group) for 3 d before radioisotope labeling. Bones were labeled with Na235SO4 for 3 h and then were digested with papain. Glycosaminoglycans were precipitated with 10% cetylpyridinium chloride and dissolved with 23 N formic acid. 35SO4 content (mean ± SEM) was determined by liquid scintillation counting (*p < 0.05; **p < 0.001 vs control).

Full size image

Effects of 8-Br-cGMP on longitudinal bone growth.

Fetal rat metatarsal bones (E20) were cultured in serum-free medium with varying concentrations of 8-Br-cGMP (0, 10, or 100 μM) for 3 d. 8-Br-cGMP (10 and 100 μM) increased longitudinal bone growth by 12% and 21%, respectively, versus controls (p < 0.05, n = 50/group) (Fig. 1B). We next assessed the effect of 8-Br-cGMP on each of the major components of longitudinal bone growth.

Addition of 8-Br-cGMP (10 or 100 μM) for 72 h did not significantly affect [3H]thymidine incorporation (n = 18/group). To assess proliferation within specific regions of the growth plate, we cultured metatarsals for 72 h in the presence of 8-Br-cGMP (0, 10, or 100 μM), labeled with [3H]thymidine, and then performed autoradiography (n = 10–12/group) (Fig. 2C). 8-Br-cGMP at a concentration of 100 μM decreased the labeling index in both resting (p < 0.001) (Fig. 3B) and proliferative (p < 0.02) (Fig. 3D) chondrocytes.

We next assessed the number of hypertrophic chondrocytes per slide and the height of the terminal hypertrophic cell. Neither hypertrophic cell number nor height was influenced by 10 or 100 μM 8-Br-cGMP (n = 12–16/group) (Fig. 4, B and D).

To assess cartilage matrix formation, we measured 35SO4 incorporation into newly synthesized glycosaminoglycans. 8-Br-cGMP (100 μM) increased glycosaminoglycans synthesis (169 ± 25% of control, n = 11–20/group, p < 0.001) (Fig. 5B).

DISCUSSION

CNP accelerated longitudinal growth of cultured fetal rat metatarsal bones. Our findings suggest that this growth acceleration is mediated by three mechanisms. First, CNP stimulated chondrocyte proliferation in the proliferative zone, the region thought to be most important for longitudinal growth (1). Second, CNP stimulated cell hypertrophy as assessed by the number and size of hypertrophic chondrocytes present in the cultured bone. Third, CNP stimulated cartilage matrix production as assessed by incorporation of 35SO4 into glycosaminoglycans.

Our results provide a mechanistic explanation for anatomical and histologic effects of natriuretic peptides previously observed in vivo and in vitro. In vivo, mice overexpressing BNP show an increased length of the long bones and increased height of the proliferative and hypertrophic zones of the growth plates (2). Although BNP is thought to serve normally as an endocrine signal that regulates sodium balance and vascular tone (13), BNP overexpression may have caused skeletal overgrowth, because it acted on the same receptors as endogenous CNP. In vitro, CNP has been shown to cause increased longitudinal growth and increased height of the proliferative and hypertrophic zones of cultured fetal mouse tibiae (4). In that study, hypertrophic cell size and immunohistochemical staining of type X collagen and incorporated 5-bromo-2′-deoxyuridine were reportedly increased by CNP, but quantification and reproducibility were not described.

Our observation that CNP acts on growth plate cartilage is consistent with the known distribution of natriuretic peptide receptors. mRNA for both NPR-A and NPR-B is expressed in the growth plate. In particular, mRNA for NPR-B, the receptor which has a high affinity and specificity for CNP, appears to be more abundant than mRNA for NPR-A in growth plate cartilage (2). Furthermore, no skeletal abnormalities have been reported in mice lacking NPR-A (4). Thus, NPR-B may be the receptor mediating the effects of CNP on the growth plate.

Both NPR-A and NPR-B contain an intracellular guanylyl cyclase catalytic domain, and at least some of the biologic actions of natriuretic peptides appear to be mediated by production of the second messenger cGMP (8, 13, 14). We therefore hypothesized that cGMP would reproduce the effects of CNP on the fetal bones. Consistent with this hypothesis, 8-Br-cGMP stimulated longitudinal growth, a finding consistent with a previous study using cultured mouse tibiae (4). Like CNP, 8-Br-cGMP stimulated glycosaminoglycan synthesis by fetal chondrocytes. This increased synthetic activity could explain the observed increase in growth rate. However, unlike CNP, cGMP decreased proliferation of growth plate chondrocytes and had no effect on hypertrophy. These findings suggest that some of the effects of CNP on the growth plate might be mediated by cGMP whereas other effects may involve different signal transduction mechanisms. These mechanisms could reflect the activation of other signaling pathways by NPR-A or NPR-B or the action of CNP on other receptors (15). CNP might affect chondrocyte proliferation and differentiation indirectly by altering expression of other growth factors present in growth cartilage.

The lowest concentration of 8-Br-cGMP (10 μM) increased longitudinal growth rate slightly. The cellular mechanism mediating this increased growth rate is not clear. We speculate that, like the higher concentration (100 μM), 10 μM 8-Br-cGMP may have increased matrix synthesis but that this modest increase, when assessed by 35SO4 incorporation, did not reach statistical significance, or that increased synthesis of other unmeasured cartilage matrix components might be involved.

The effects of cGMP on growth plate chondrocytes appear to differ from those of cAMP (16). cAMP, another cyclic nucleotide, acts in chondrocytes as a second messenger for intercellular signals such as PTH-related peptide and prostaglandins. Unlike cGMP, cAMP inhibits hypertrophic differentiation. cAMP appears to have a biphasic effect on DNA synthesis, stimulating [3H]thymidine incorporation at low concentrations and inhibiting it at high concentrations. At high concentrations, cAMP, like cGMP, increases cartilage matrix production.

Our findings provide a partial explanation for the phenotype of mice with homozygous inactivating mutations in the gene for cGMP-dependent protein kinase-II (17). These mice show impaired longitudinal bone growth. Our results suggest that this growth impairment occurs because the mutation blocks the stimulatory effect of cGMP on chondrogenesis. However, our findings do not explain the growth plate disorganization observed. Deficient cGMP action throughout development may well have effects on the growth plate which would not be precisely mirrored by a brief exposure to increased cGMP.

We conclude that CNP stimulates longitudinal bone growth by increasing chondrocyte proliferation, chondrocyte hypertrophy, and cartilage matrix production, and thus may serve as an endogenous positive regulator of endochondral bone formation in the growth plate. cGMP, a second messenger for CNP, reproduces some but not all of the effects of CNP, suggesting that other signal transduction mechanisms may also be involved.