Search

Thiram-induced tibial dyschondroplasia (TD) and vitamin-D deficiency rickets are avian bone disorders of different etiologies characterized by abnormal chondrocyte differentiation, enlarged and unvascularized growth plates, and lameness. Heat-shock protein 90 (Hsp90) is a proangiogenic factor in mammalian tissues and in tumors; therefore, Hsp90 inhibitors were developed as antiangiogenic factors. In this study, we evaluated the association between Hsp90, hypoxia, and angiogenesis in the chick growth plate. Administration of the Hsp90 inhibitor to TD- and rickets-afflicted chicks at the time of induction resulted in reduction in growth-plate size and, contrary to its antiangiogenic effect in tumors, a major invasion of blood vessels occurred in the growth plates. This was the result of upregulation of the VEGF receptor Flk-1, the major rate-limiting factor of vascularization in TD and rickets. In addition, the abnormal chondrocyte differentiation, as characterized by collagen type II expression and alkaline phosphatase activity, and the changes in hypoxia-inducible factor-1α (HIF-1α) in both disorders were restored. All these changes resulted in prevention of lameness. Inhibition of Hsp90 activity reduced growth-plate size, increased vascularization, and mitigated lameness also in TD chicks with established lesions. In summary, this is the first reported demonstration of involvement of Hsp90 in chondrocyte differentiation and growth-plate vascularization. In contrast to the antiangiogenic effect of Hsp90 inhibitors observed in mammals, inhibition of Hsp90 activity in the unvascularized TD- and rickets-afflicted chicks resulted in activation of the angiogenic switch and reinstated normal growth-plate morphology.

Enzymes of the matrix metalloproteinase (MMP) family regulate angiogenesis and are involved in the endochondral ossification process. Tibial dyschondroplasia (TD) and rickets are 2 disorders associated with impairments in this process, mainly in the vascularization of the avian growth plate. In this paper, we induced TD and rickets and studied the expression patterns of 4 members of the MMP family known to be important for endochondral ossification, MMP-2, 3, 9, and 13, in normal and impaired avian growth plates. The expression of MMP-3, 9, and 13 was reduced in the lesions and lined up parallel to the expulsion of blood vessels, which was extended up to the border of the lesion, but did not penetrate into it. Matrix metallopro-teinase-2 was not expressed in the TD lesion but was overexpressed in the rachitic lesion. We also studied the differentiation stage of the chondrocytes populating the lesions and found that the rachitic lesions were populated with proliferative chondrocytes, whereas the TD lesions were filled with chondrocytes that presented both proliferative and hypertrophic markers. These results suggest that MMP-3, 9, and 13 play a role in the vascularization and ossification processes, whereas MMP-2 is related to chondrocyte differentiation and may be involved in cartilage remodeling in the avian growth plate.

The epiphyseal plates of rachitic chicks fed a vitamin D-deficient and low-calcium diet were examined ultrastructurally at 4, 7, 14, 18 and 21 days old. On and after 14 days, changes were observed. Chondrocytes in resting, proliferating and maturing zones commonly showed a reduction in cell organelles, suggestive of a decrease in synthetic activity of matrix components. In addition, the resting zone cells had numerous intracytoplasmic microfilaments. Mitotic figures were present but not increased in number in the proliferating zone. Autolysosome-like dense bodies in the chondrocytes and clusters of degenerative and necrotic chondrocytes were observed in the proliferating and maturing zones. In the calcifying zone of the rachitic epiphyseal plate, initial calcification was observed characterized by the deposition of apatite crystals in matrix vesicles and the formation of spherical crystal clusters. The crystal clusters were separated from each other by collagen fibrils on which only small amounts of apatite crystals had been deposited. In this zone, hypertrophic chondrocytes responsible for the initial calcification were seen, but not the stellate chondrocytes responsible for the progression of matrix calcification after its initiation. These findings suggest that accumulation of the proliferating and maturing zone cells as a characteristic lesion of the epiphyseal plate in rachitic chicks is caused by a disturbance of cell maturation. Such disturbance in the calcifying zone may explain the progression of the defect in matrix calcification.

The morphology of the metaphyseal microvasculature at the epiphysis was examined at both the light and electron microscopic level in rickets and rachitic reversal. The animals studied were normal, rachitic, and rachitic reversed at 8, 24, and 96 hours post-vitamin D administration. The overall architecture of the metaphyseal vessels was significantly altered throughout the intervals examined. In the rachitic animal, arterioles, venules, and capillaries were found adjacent to the growth plate, either directly apposed to the hypertrophic chondrocytes or separated from them by bone-forming cells. These vessels are in many ways similar to the larger arterioles and venules that normally supply the metaphyseal capillary sprouts, but in the normal growing animal are usually located 350-500 microns from the epiphyseal cartilage. The rachitic capillaries appear relatively well differentiated with a partial basement membrane and a perivascular cell lining. In early rachitic reversal, small vascular projections are induced to grow from the large diameter venules that border upon the hypertrophic chondrocytes. These vascular sprouts that invade the epiphyseal cartilage are quite undifferentiated, with no basement membrane or pericyte lining at the sprout apex and occasional abluminal endothelial cell projections. Within 96 hours, the metaphyseal microvasculature has returned to an apparently normal state with only capillaries at the cartilage-vascular interface and larger vessels (arterioles and venules) located several hundred microns deeper into the metaphysis. The sequential processes of differentiation and cessation of capillary growth followed by dedifferentiation and reinitiation of microvascular growth make the rachitic system a unique one in which to study angiogenesis.

The effects of vitamin D metabolites on alkaline phosphatase [ALPase; orthophosphoric-monoester phosphohydrolase (alkaline optimum), EC 3.1.3.1] activity, a marker of terminal differentiation, in chondrocyte cultures and growth plates in vivo were examined. In cultures of pelleted rabbit growth-plate chondrocytes, 1,25-dihydroxycholecalciferol (1,25-dihydroxyvitamin D3) increased the contents of DNA and macromolecules containing uronic acid (proteoglycans). It also decreased ALPase activity with an ED50 of less than 1 nM. Other vitamin D3 metabolites, such as 24,25-dihydroxycholecalciferol and 25-hydroxycholecalciferol, had little effect on these biochemical parameters. In rachitic growth plates, the uronic acid content was half that in normal growth plates, whereas ALPase activity was 2.5 times that in normal growth plates. Administration of 1,25-dihydroxycholecalciferol at a low dose (0.1 micrograms per kg of body weight) to rachitic rats increased the uronic acid content 1.4-fold and decreased ALPase activity by 40%. This compound, like 24,25-dihydroxycholecalciferol (10 micrograms per kg of body weight), increased the calcium level of the blood. However, administration of 24,25-dihydroxycholecalciferol had little effect on the uronic acid and ALPase contents in growth plates. These observations suggest that 1,25-dihydroxycholecalciferol is a bioactive form of vitamin D that plays an important role in the control of chondrocyte terminal differentiation.

Calcifying matrix vesicles (MVs) are released from chondrocytes and osteoblasts in monolayer culture. In the present studies, we tested the ability of rachitic versus normal rat growth plate chondrocytes in micromass or monolayer primary cultures to produce MVs. Unlike earlier reports of in vitro MV biogenesis by chicken chondrocytes in which most MVs were released into the medium, we found that most of the released rat matrix vesicles were entrapped in a newly formed cartilaginous matrix enveloping the cells. These matrix-associated MVs could be isolated by mild collagenase treatment and concentrated by differential centrifugation. Vesicle production slowed in the older 2- to 4-week-old cultures and, unlike vesicle release from cultured chicken chondrocytes, active vesicle production did not show a second burst of activity at 3 to 4 weeks. Alkaline phosphatase (ALP) activity diminished with time in culture in cells and matrix vesicles, suggesting a decrease in differentiative expression. Protein profiles on SDS polyacrylamide gels of native matrix vesicles and culture-derived MVs from rachitic and normal cells were quite similar and showed a typical simplified protein pattern as compared to chondrocyte plasma membrane proteins. There were distinctive proteins migrating at 130, 80 to 95, 66, 43, 20, and 14 kd. Culture-derived MVs showed vigorous in vitro calcifying activity that was ALP related. We conclude that 1) rachitic chondrocytes are essentially normal in their matrix vesicle production; 2) matrix entrapment of MVs is a characteristic of rat chondrocyte cultures; and 3) culture-produced MVs are similar to native MVs in protein profile and calcifiability, and thus can be studied as a model for normal MV composition and calcification.

We examined collagens of both normal and vitamin D-deficient chick epiphyseal growth cartilage. Special emphasis was placed on the study of Type X collagen, a recently described product of hypertrophic chondrocytes. Scanning electron microscopy of the epiphyseal growth cartilage of vitamin D-deficient chickens showed an enlarged growth cartilage with a disorganized extracellular matrix. The cartilage collagens were solubilized by proteolytic digestion and disulfide bond reduction of both normal and rachitic growth tissues. Sequential extraction with neutral salt and acetic acid buffers followed by pepsin digestion at 4 degrees C solubilized about 12% of normal tissues and about 7% of collagen from rachitic growth cartilage. Treatment of the pepsin-resistant collagens with neutral salt-dithiothreitol buffer under nondenaturing conditions and a subsequent pepsin digestion increased the yield of solubilized collagen to greater than 95% of the total tissue collagen. Results of the biochemical studies showed a marked increase in the relative proportion of Type X collagen (from 5.6 to 27.9%), a corresponding decrease in the proportions of Types II and IX collagens, and a moderate increase in Type XI collagen in rachitic cartilage. Amino acid analysis indicated that there were no differences in the Types II and X collagens of normal and rachitic cartilage. However, an abnormality in the relative proportions of the CNBr peptides of Type X collagen was detected in the rachitic cartilage. We suggest that the increase in collagen in the rachitic state may reflect increased levels of Type X collagen synthesis by cells in the hypertrophic region. It is likely that in rickets the overproduction of Type X collagen may be a compensatory mechanism by which the hypertrophic chondrocyte attempts to provide a maximum area of calcifiable matrix for the calcium-depleted serum.

The preparation of mineralizing tissues for electron microscopy by methods that involve the use of aqueous solvents may result in translocation and/or removal of mineral, leading to the possible erroneous interpretation of initial calcifying loci. The failure of a recent cryoultramicrotomy study to detect a mineral phase associated with matrix vesicles raised doubts about the role of the matrix vesicle in initial mineralization and prompted us to reexamine the ultrastructure of calcifying cartilage using three different anhydrous preparatory techniques. Matrix vesicles or structures similar in profile and location were identified in rat epiphyseal growth plate cartilages that had been subjected to freeze-substitution, cryoultramicrotomy, or ethylene glycol fixation. Furthermore, mineral deposits, as confirmed by energy dispersive x-ray analysis, were observed within or in close association with the matrix vesicles in the extracellular matrix of the growth plate cartilage by all three methods. No evidence of mineral deposition at other sites preceding matrix vesicle calcification was found. Our studies support the conclusion of many investigations that matrix vesicles are the initial site of growth plate calcification.

Responses of cultured cartilage cells to metabolites of vitamin D3 were studied. Cells were obtained from the epiphyseal growth plate of rachitic chicks and were exposed to physiological and pharmacological concentrations of three metabolites of vitamin D3, 25 hydroxyvitamin D3 (25(OH)D3), 24,25-dihydroxyvitamin D3 (24,25(OH)2D3) and 1,25-dihydroxyvitamin D3 (1,25(OH)2D3). 1,25(OH)2D3 was found to reduce L-[U-14C]leucine incorporation into proteins and Na2 35SO4 incorporation into proteoglycans. The synthesis of 24,25(OH)2D3 from 25(OH)D3 was stimulated upon addition of 1,25(OH)2D3 to the cultures. Physiological concentrations of 24,25(OH)2D3 stimulated protein and proteoglycan synthesis. These findings support the notion that vitamin D3, through its active dihydroxylated metabolites, is directly involved in cartilage cells metabolism and healing of rickets.

Oral administration of manganese to young rats results in poorly mineralized primary spongiosa and an irregularly thickened growth plate with a histologic resemblance to that in vitamin D-deficiency rickets. In the present study, the rachitic lesions were characterized by stereologic methods at the light microscopic level. With increasing doses of Mn in the diet, the animals developed rachitic lesions of increasing severity, i.e., the total height of the growth plate and the relative volume of the hypertrophic zone increased. The experimental animals developed hypophosphatemia, which was dependent on the Mn dose. The observed serum concentrations of Mn and phosphorus are compatible with the idea that MnHPO4 is precipitated in the gut, leaving only small amounts of Mn and phosphate available for absorption. Furthermore, the severity of the rachitic lesions were inversely correlated to the concentration of phosphate in serum. The most important pathomechanism in Mn rickets is phosphate depletion, which per se causes similar rachitic changes, even though Mn also seems to have other effects. Starvation caused a decrease in the height of the growth plate and in the volume fraction of the hypertrophic zone, thus changes contrary to the rachitic lesions.

A model of the healing phase of low phosphate, vitamin D deficiency was used to investigate the initial stages of mineralization. The matrix vesicle distribution between the zones of the growth plate was found to be bimodal with high volume densities in the resting and hypertrophic zones and low volume densities in the proliferative and calcifying zones. Healing of the rachitic lesion was associated with a decrease in matrix vesicle volume density in the calcifying zone, compared with the lower hypertrophic zone in florid rickets. The volume density differences were due to differences in the number of vesicles, as the variation in mean caliper diameter was rather small. The findings are compatible with the dynamic cell debris theory for matrix vesicle origin and distribution presented earlier, which favours the view that a major part of matrix vesicles are formed from cell debris. A role of matrix vesicles in the mineralization process is indicated by the finding of an association between mineralization and matrix vesicle degradation.

In low phosphate, vitamin D-deficiency rickets normal mineralization is reversibly arrested, and rickets is thus a suitable model for studying factors influencing the mineralization process. Partly on the basis of their distribution within the growth cartilage, the so-called matrix vesicles are considered to play an important role in the process of mineralization and it has been claimed that the distribution pattern is the same in rickets as in normal animals. With the use of modern stereological techniques, our group recently demonstrated a matrix vesicle distribution between the zones of the epiphyseal growth plate in normal rats different from that described earlier. The same bimodal distribution pattern was observed in rachitic rats in the present study, the highest volume density being found in the resting and upper hypertrophic zones and the lowest in the proliferative zone. The volume density differences are explained by differences in the number of vesicles between zones, the variation in mean caliper diameter being small. Our findings are discussed in relation to the proposed theories on matrix vesicle origin. The results seem to support the dynamic cell debris theory for matrix vesicle origin presented earlier, but the existence of subpopulations of matrix vesicles with a specialized function and origin cannot be ruled out.

The levels of type X collagen in mineralizing normal chicken epiphyses and nonmineralizing rachitic chicken tibial epiphyses were measured and compared. Qualitative immunoperoxidase studies with anti-chick type X collagen monoclonal antibodies on sections from normal and rachitic cartilage demonstrated that the type X collagen levels in rachitic growth plates are reduced. Northern hybridization of mRNA and biosynthetic studies have confirmed that type X collagen synthesis in rickets is also decreased. In hypocalcemic rickets, the level of type X collagen mRNA is reduced by 80% whereas the level of type X collagen mRNA is only reduced by 50% in normocalcemic rickets. These observations provide additional evidence that type X collagen is involved in the process of cartilage mineralization and also suggest that the partial recovery of type X collagen synthesis in normocalcemic rickets may be related to the elevated plasma concentration of calcium. Calcium concentration may therefore play an important role in the control of type X collagen synthesis.

The mandibular condylar cartilages from normal and rachitic weanling rats were stained with ruthenium red to demonstrate the distribution of proteoglycan. In the rachitic animal there was an increased concentration of ruthenium-red positive granules in the pericellular region of chondrocytes from the extended hypertrophic zone when compared with the controls. These aggregates may represent unmodified proteoglycan which excludes calcium and thereby prevents matrix calcification.

In the transition from proliferating to hypertrophic cell zones in the growth plate, there is an increased in chondrocyte cell volume and a corresponding decrease in collagen content to allow for cell enlargement. To substantiate our hypothesis that collagenase is responsible for these changes, growth plates from rats treated with bisphosphonate (HEBP) were compared histologically and biochemically with growth plates from normal and vitamin D and phosphate deficient (-VDP) rats. HEBP-treated rats developed an expanded hypertrophic cell zone (HCZ) characterized by the presence of two distinct populations of hypertrophic cells. The proximal hypertrophic cells were only 2-fold enlarged compared to the proliferating cells, whereas 1/6 of the distal hypertrophic cells were enlarged almost 5-fold and appeared morphologically identical with hypertrophic cells from normal and -VDP rats. The HEBP growth plates were divided into cross-sectional thirds and analyzed for active and latent collagenase. The juxta-metaphyseal (lower 1/3) cartilage contained 100% of the fully enlarged hypertrophic cells and appeared identical to those found in normal and -VDP growth plates, along with 81% of the active and 77% of the total collagenase. Collagenase and tissue inhibitor of metalloproteinases (TIMP) were measured in extracts of similarly divided tissues. The presence of true collagenas was confirmed by using [3H]-telopeptide-free collagen. TIMP levels were inversely related to the presence of active collagenase and cellular hypertrophy. Substantial levels of latent collagenase were found in the extracellular fluid at sites of active collagenolysis, but not in the fluid phase surrounding the 2-fold enlarged hypertrophic cells. It is postulated that increased amounts of active collagenase and insufficient levels of TIMP may account for the reduced collagen content seen in the lower HCZ of both -VDP and HEBP rickets. Unlike active collagenase, which remains localized by binding to collagen, latent enzyme is probably restricted in its mobility throughout the extracellular space by diffusion, itself, or the interstices of the extracellular matrix.

A new mouse osteopetrotic mutation, osteosclerosis, has been examined with respect to rickets. These osteopetrotic mice were hypocalcemic and hypophosphatemic, and had greatly thickened epiphyseal plates with abnormalities in matrix vesicles when compared with normal littermates. Such biochemic and morphologic manifestations of rickets in this osteopetrotic mutation may explain the failure of osteosclerotic mice to be cured by transplantation of bone marrow from normal littermates and indicate that vitamin D metabolism and matrix vesicle biochemistry warrant further study.

When HEBP (1-hydroxyethylidene-1, 1-biphosphonic acid) was administered to young rats in large doses over a short period rickets was consistently produced. When HEBP was administered concomitantly with 1,25 (OH)2D3 or calcitonin (CT), calcification appeared in the growth-plate cartilage where there had been an increase in thickness due to the inhibition of calcification. This experiment was done in an attempt to clarify differences in the calcification-promoting mechanisms of 1,25 (OH)2D3 and CT. The serum alkaline phosphatase level was reduced in rats with an accelerated calcification following the administration of 1,25 (OH)2D3, but there was no reduction in the serum alkaline phosphatase level in rats in which the calcification was accelerated by the administration of CT. The mode of appearance of calcification in the growth-plate cartilage by 1,25 (OH)2D3 or CT differed, depending on the time of administration. These results suggest that mechanisms involved in the enhancement of calcification by 1,25 (OH)2D3 and CT differ in cases where rickets are induced by HEBP.

This paper reviews recent work in the authors' laboratories that has led to new observations and thoughts concerning the mineralization of normal and rachitic chick growth cartilage. The proximal tibial growth cartilages of normal and rachitic chicks were rapidly frozen and prepared for SEM and biochemical studies. Using a scanning microfluorimetric technique we showed that at the mineralization front of normal and rachitic cartilage there is an abrupt change in chondrocyte metabolism. Thus cells in this region exhibited an increase in NADH and oxidative metabolism. In rickets, there was a decrease in the reduced pyridine nucleotide content of each of the zones. The reversal in chondrocyte metabolism was not due to low oxygen tension. SEM observations indicated that this region of cartilage was well supplied with vascular channels; moreover, mineral was first seen deposited in matrix in close proximity to the blood supply. Indeed these vascular channels appeared to be a basic architectural feature of normal cartilage, although disorganized in the rachitic state. The morphological studies also showed that gaps existed in the continuity of the mineral phase in normal cartilage. Although the rachitic cartilage does mineralize, discontinuities in the mineral distribution are much more severe, with the general failure of fusion of adjacent mineral clusters. These structures would serve as pathways for transport of nutritional factors and gases to chondrocytes that are distant from the vascular channels. Observation of hypertrophic cells reinforced the view that some osteoblasts represented a terminal stage in the maturation of chondrocytes.

To estimate matrix vesicle distribution between longitudinal and transversal septal matrix in the proliferative, hypertrophic and calcifying zones of normal epiphyseal cartilage, the volume density of matrix vesicles in the longitudinal septal matrix was compared to that of total extralacunar matrix. The results confirm the qualitative observation by Anderson that matrix vesicles are located mainly in the longitudinal septa. To elucidate whether cartilage mineralization can be related to the disappearance of matrix vesicles of particular size classes, epiphyseal growth cartilage from three groups of animals were studied: normal rats, rats with florid rickets and rats with early healing rickets. The study was focused on the proliferative, hypertrophic and calcifying zones and in each zone the matrix vesicles were classified into four size classes: 1, less than or equal to 50 nm; 2, 51-67 nm; 3, 68-84 nm; 4, greater than or equal to 85 nm. The results show that the decrease in volume density previously demonstrated in normal rats to a large extent is due to a decreased number of larger vesicles. In florid rickets the decrease in this size group is much smaller while the values for healing rachitic animals fall between those of florid rachitic rats and those of controls. The data indicate that the decrease in the number of larger vesicles, which represents a considerable vesicle volume, is of particular importance. The heterogenous change in the concentration of matrix vesicles of different size classes during cartilage mineralization as well as under conditions of arrested calcification, is compatible with the existence of a matrix vesicle subpopulation of larger size.