Evidence that the rat osteopetrotic mutation toothless (tl) is not in the TNFSF11 (TRANCE, RANKL, ODF, OPGL) gene.

The toothless (tl) osteopetrotic mutation in the rat affects an osteoblast-derived factor that is required for normal osteoclast differentiation. Although the genetic locus remains unknown, the phenotypic impact of the tl mutation on multiple systems has been well characterized. Some of its actions are similar to tumornecrosis factor superfamily member 11(TNFSF11; also called TRANCE, RANKL, ODF and OPGL) null mice. TNFSF11 is a recently described member of the tumor necrosis factor superfamily which, when expressed by activated T cells, enhances the survival of antigen-presenting dendritic cells, and when expressed by osteoblasts, promotes the differentiation and activation of osteoclasts. The skeletal similarities between tl rats and TNFSF11(-/-) mice include 1) profound osteoclastopenia (TNFSF11-null mice, 0% and tl rats 0-1% of normal); 2) persistent, non-resolving osteopetrosis that results from 3) a defect not in the osteoclast lineage itself, but in an osteoblast-derived, osteoclastogenic signal; and 4) a severe chondrodysplasia of the growth plates of long bones not seen in other osteopetrotic mutations. The latter includes thickening of the growth plate with age, disorganization of chondrocyte columns, and disturbances of chondrocyte maturation. These striking similarities prompted us to undertake studies to rule in or out a TNFSF11 mutation in the tl rat. We looked for expression of TNFSF11 mRNA in tl long bones and found it to be over-expressed and of the correct size. We also tested TNFSF11 protein function in the tl rat. This was shown to be normal by flow cytometry experiments in which activated, spleen-derived T-cells from tl rats exhibited normal receptor binding competence, as measured by a recombinant receptor assay. We also found that tl rats develop histologically normal mesenteric and peripheral lymph nodes, which are absent from TNFSF11-null mice. Next, we found that injections of recombinant TNFSF11, which restores bone resorption in null mice, had no therapeutic effect in tl rats. Finally, gene mapping studies using co-segregation of polymorphic markers excluded the chromosomal region containing the TNFSF11 gene as harboring the mutation responsible for the tl phenotype. We conclude that, despite substantial phenotypic similarities to TNFSF11(-/-) mice, the tl rat mutation is not in the TNFSF11 locus, and that its identification must await the results of further studies.

Endochondral bone formation in toothless (osteopetrotic) rats: failures of chondrocyte patterning and type X collagen expression.

The pacemaker of endochondral bone growth is cell division and hypertrophy of chondrocytes. The developmental stages of chondrocytes, characterized by the expression of collagen types II and X, are arranged in arrays across the growth zone. Mutations in collagen II and X genes as well as the absence of their gene products lead to different, altered patterns of chondrocyte stages which remain aligned across the growth plate (GP). Here we analyze GP of rats bearing the mutation toothless (tl) which, apart from bone defects, develop a progressive, severe chondrodystrophy during postnatal weeks 3 to 6. Mutant GP exhibited disorganized, non-aligned chondrocytes and mineralized metaphyseal bone but without cartilage mineralization or cartilaginous extensions into the metaphysis. Expression of mRNA coding for collagen types II (Col II) and X (Col X) was examined in the tibial GP by in situ hybridization. Mutant rats at 2 weeks exhibited Col II RNA expression and some hypertrophied chondrocytes (HC) but no Col X RNA was detected. By 3rd week, HC had largely disappeared from the central part of the mutant GP and Col II RNA expression was present but weak and in 2 separate bands. Peripherally the GP contained HC but without Col X RNA expression. This abnormal pattern was exacerbated by the fourth week. Bone mineralized but cartilage in the GP did not. These data suggest that the tl mutation involves a regulatory function for chondrocyte maturation, including Col X RNA synthesis and mineralization, and that the GP abnormalities are related to the Col X deficiency. The differences in patterning in the tl rat GP compared to direct Col X mutations may be explained by compensatory effects.

Colony-stimulating factor-1 (CSF-1) rescues osteoblast attachment, survival and sorting of beta-actin mRNA in the toothless (tl-osteopetrotic) mutation in the rat.

We have shown that in the osteopetrotic rat mutation toothless (tl) osteoblasts are absent from older bone surfaces in mutants and that mutant osteoblasts in vivo lack the prominent stress fiber bundles polarized along bone surfaces in osteoblasts from normal littermates. Our recent data demonstrate that in normal osteoblasts in vitro beta- and gamma-actin mRNAs have different, characteristic intracellular distributions and that tl (mutant) osteoblasts fail to differentially sort these mRNAs. Because bone resorption and formation are highly interdependent and injections of CSF-1, a growth factor, increase bone resorption and growth in tl rats, we examined the effects of CSF-1 treatment on osteoblast survival and ultrastructure in vivo and ability to sort actin mRNAs in vitro. Neonatal CSF-1 treatment of mutants restores osteoblasts on older bone surfaces, normalizes the intracellular distribution of stress fibers in osteoblasts in vivo and promotes normal sorting of beta-actin mRNA in mutant osteoblasts in vitro without normalizing gamma-actin distribution. These data suggest the beta- and gamma-actin mRNAs in osteoblasts are sorted by different mechanisms and that the differential sorting of beta-actin mRNA is related to the characteristic polarization of stress fibers in osteoblasts and their survival on bone surfaces. This experimental system can be used to explore the relationships and regulation of these aspects of cell and tissue biology.

The toothless osteopetrotic rat has a normal vitamin D-binding protein-macrophage activating factor (DBP-MAF) cascade and chondrodysplasia resistant to treatments with colony stimulating factor-1 (CSF-1) and/or DBP-MAF.

The osteopetrotic rat mutation toothless (tl) is characterized by little or no bone resorption, few osteoclasts and macrophages, and chondrodysplasia at the growth plates. Short-term treatment of tl rats with colony-stimulating factor-1 (CSF-1) has been shown to increase the number of osteoclasts and macrophages, producing dramatic resolution of skeletal sclerosis at some, but not all, sites. Defects in production of vitamin D-binding protein-macrophage activating factor (DBP-MAF) have been identified in two other independent osteopetrotic mutations of the rat (op and ia), and two in the mouse (op and mi), in which macrophages and osteoclasts can be activated by the administration of exogenous DBP-MAF. The present studies were undertaken to examine the histology and residual growth defects in tl rats following longer CSF-1 treatments, to investigate the possibility that exogenous DBP-MAF might act synergistically with CSF-1 to improve the tl phenotype, and to assess the integrity of the endogenous DBP-MAF pathway in this mutation. CSF-1 treatment-with or without DBP-MAF-induced resorption of metaphyseal bone to the growth plate on the marrow side, improved slightly but did not normalize long bone growth, and caused no improvement in the abnormal histology of the growth plate. Injections of lysophosphatidylcholine (lyso-Pc) to prime macrophage activation via the DBP-MAF pathway raised superoxide production to similar levels in peritoneal macrophages from both normal and mutant animals, indicating no defect in the DBP-MAF pathway in tl rats. Interestingly, pretreatments with CSF-1 alone also increased superoxide production, although the mechanism for this remains unknown. In summary, we find that, unlike other osteopetrotic mutations investigated to date, the DBP-MAF pathway does not appear to be defective in the tl rat; that additional DBP-MAF does not augment the beneficial skeletal effects seen with CSF-1 alone; and that the growth plate chondrodystrophy seen in this mutation is unaffected by either molecule. Thus, the tl mutation intercepts the function of a gene required for both normal endochondral ossification and bone resorption, thereby uncoupling the coordination of skeletal metabolism required for normal long bone growth.

Facial development and type III collagen RNA expression: concurrent repression in the osteopetrotic (Toothless,tl) rat and rescue after treatment with colony-stimulating factor-1.

The toothless (osteopetrotic) mutation in the rat is characterized by retarded development of the anterior facial skeleton. Growth of the anterior face in rats occurs at the premaxillary-maxillary suture (PMMS). To identify potential mechanisms for stunted facial growth in this mutation we compared the temporospatial expression of collagen I (Col I) and collagen III (Col III) RNA around this suture in toothless (tl) rats and normal littermates by in situ hybridization of specific riboprobes in sagittal sections of the head. In normal rats, the suture is S shaped at birth and becomes highly convoluted by 10 days with cells in the center (fibroblasts and osteoblast progenitors) expressing Col III RNA and those at the periphery (osteoblasts) expressing no Col III RNA but high amounts of Col I RNA throughout the growth phase (the first 2 postnatal weeks). In the mutant PMMS, cells were reduced in number, less differentiated, and fewer osteoblasts were encountered. Expression of Col I RNA was at normal levels, but centrosutural cells expressed Col III RNA only after day 6 and then only weakly. A highly convoluted sutural shape was never achieved in mutants during the first 2 postnatal weeks. Treatment of tl rats with the cytokine CSF-1 improved facial growth and restored cellular diversity and Col III RNA expression in the PMMS to normal levels. Taken together, these data suggest that normal facial growth in rats is related to expression of Col III RNAby osteoblast precursors in the PMMS, that these cells are deficient in the tl mutation and are rescued following treatment with CSF-1.

Actin mRNA isoforms are differentially sorted in normal osteoblasts and sorting is altered in osteoblasts from a skeletal mutation in the rat.

Actin isoform sorting has been shown to occur in a variety of cell types in culture. To this list we add osteoblasts, in which we show by in situ hybridization that beta-actin is distributed primarily in cell processes and on one side of the nucleus and gamma-actin has a perinuclear distribution. Osteoblasts from the skeletal mutation toothless (tl), evaluated under identical conditions, fail to sort these actin isoforms differentially and exhibit diffuse labeling as their major manifestation. Northern analyses of actin mRNAs showed no differences between normal and mutant cultures. Shortened osteoblast life span and an inability to direct osteoclast-mediated bone resorption have recently been demonstrated in tl mutants. The present results suggest that a failure of osteoblasts to sort actin mRNAs may be related to one or both of these pathological manifestations in this mutation and represent, to our knowledge, the first correlation of an actin mRNA-sorting abnormality with a mammalian disease.

The effects of colony-stimulating factor-1 on the number and ultrastructure of osteoclasts in toothless (tl) rats and osteopetrotic (op) mice.

The role of colony-stimulating factor-1 (CSF-1 or M-CSF) in osteoclast development is illustrated by observations that administration of exogenous CSF-1 increases osteoclast number and improves the skeletal sclerosis of two osteopetrotic mutations, toothless (tl) in the rat and osteopetrotic (op) in the mouse. We examined the effects of CSF-1 treatment on the number and ultrastructure of osteoclasts in the tibial metaphysis of normal and mutant animals of both stocks to understand the similarities and differences between these two mutations. Osteoclasts from normal animals of both stocks were abundant and possessed the ultrastructural features of active cells. These included apical areas in contact with mineralized surfaces with tightly apposed clear zones, extensive ruffled borders, and a vacuolated cytoplasm with numerous mitochondria. In toothless rats osteoclasts were difficult to locate and those present had poorly defined ruffled borders, fewer cytoplasmic vacuoles, and a basal membrane with both smooth and ruffled areas. Large lipid accumulations were often found near tl osteoclasts. Osteoclasts in op mice were difficult to find, but more numerous than in tl rats. Unlike tl osteoclasts, those of op mice possessed very well developed ruffled borders, small clear zones, and large electron-dense cytoplasmic inclusions. These cells also had unusual basal membranes with both smooth and ruffled regions. CSF-1 treatment increased the number of osteoclasts in both mutant stocks, normalizing the numbers in op mice, but not tl rats. CSF-1 injections caused dramatic changes in the morphology of tl osteoclasts, including increased incidence and size of ruffled borders and cytoplasmic vacuolization. The growth factor had little effect on ruffled borders or clear zones in op mice. Interestingly, mutant osteoclasts of both stocks exhibited a ruffled basal membrane in response to CSF-1 treatment. This increase in membrane ruffling may reflect the ability of CSF-1 to promote rapid formation of osteoclasts from mononuclear precursors in a more permissive microenvironment. Our data indicate that CSF-1 is not required for the development of at least some osteoclasts. The differences in response to CSF-1 treatment which we report lead us to speculate that additional factors may be involved in osteoclastogenesis.

Ultrastructural evidence of abnormally short and maldistributed actin stress fibers in osteopetrotic (toothless) rat osteoblasts in situ after detergent perfusion.

Osteopetrosis is a heterogenous group of metabolic bone diseases characterized by a generalized increase in skeletal mass, the product of reduced bone resorption and interceptions in the development and/or function of osteoclasts. In one such mutation in the rat, toothless (tl), osteoblasts are absent from older bone surfaces and there is evidence for aberrant osteoblast gene expression and function. Given the emerging appreciation of the role of osteoblasts in the differentiation and activation of osteoclasts, we have examined ultrastructural features of the cytoskeleton of normal and mutant osteoblasts after perfusion fixation with the non-ionic detergent Triton X-100. This procedure reduces the electron density of the cytoplasm, rendering visible the microfilamentous network in osteoblasts. In normal osteoblasts a prominent system of stress fibers (bundles of actin microfilaments) run parallel to the cell membrane adjacent to osteoid surfaces, stretching for 75% of that distance. However, only 50% of mutant (tl) osteoblasts had stress fibers and in these cells stress fibers were either significantly shorter (18% of normal) or distributed intracellularly rather than along the osteoid surface. In mutant osteoblasts without stress fibers, 20% showed ultrastructural signs of cell degeneration. Given the role of stress fibers in cellular attachment ot extracellular matrices, these observations suggest that the reduced number of osteoblasts in tl rats may be related to their inability to organize actin filaments and adhesion plaques for attachment to bone surfaces. We propose that a feature of osteopetrosis in the tl rat is a disruption of the mechanisms that regulate the synthesis, sorting, and/or assembly of actin.

Thrombocytopenia in the toothless (osteopetrotic) rat and its rescue by treatment with colony-stimulating factor-1.

Osteopetrosis in toothless (tl) rats is characterized by reductions in bone resorption, osteoclasts, and macrophages, resistance to cure by bone marrow transplantation, and skeletal improvement after treatment with colony-stimulating factor-1 (CSF-1). Reductions in skeletal osteocalcin tl rats, together with the recent demonstration of osteocalcin expression in platelets and its possible role in bone turnover, prompted us to examine whether this rat mutation is associated with altered platelet numbers. Our prediction of a thrombocytopenia was confirmed by examination of tl rats, in which a profound reduction (32%) in platelets was accompanied by a significant elevation (62%) in megakaryocytes (MKC) compared to normal littermates. Light and transmission electron microscopy confirmed increases in both number and size of MKC in mutants without morphologic abnormalities of circulating platelets. CSF-1 treatment (10(6) U/48 hours for 10 days) of mutants restored platelet numbers to those found normal littermates and increased osteoclasts and the frequency of MKC in numbers. Preliminary studies of another mutation the rat, osteopetrosis (op), revealed a similar reduction (33%) in platelets. These data demonstrate the coexistence of osteopetrosis and thrombocytopenia in two osteopetrotic rat mutations and an increase in osteoclasts and platelets in one mutation after CSF-1 treatment. Together, these data suggest a potential functional interaction of MKC and osteoclasts in bone turnover.

Microvascular pattern in the metaphysis during bone growth.

BACKGROUND: Little is known about the three-dimensional micromorphology of vessels in the growth zone of long bones, where significant vasculogenesis occurs. Therefore, we examined the microvascular pattern of the femoral metaphysis. METHODS: Six-week-old normal rats of either sex were used. We cast the femurs of 14 rats with Mercox for scanning electron microscopy (SEM), and in 10 rats we prepared tissue sections of femurs for light (LM) and transmission electron microscopy (TEM). RESULTS: In the LM, calcified cartilage was found to define cylindrical compartments beneath the last row of hypertrophied chondrocytes of the metaphyseal growth plate. These compartments ran in the bone's longitudinal axis and contained a single capillary profile. Endothelial cells of these capillaries often showed increased cytoplasmic volume and loose texture of nuclear chromatin. Cast metaphyses by SEM showed numerous parallel vascular loops with nodular protrusions 10-12 microns in diameter at their tips. The loops had ascending and descending limbs with a luminal diameter of 10-14 microns. Small projections 4-5 microns in diameter and delicate crests were sometimes found on the tip of the larger nodes. In a 100 x 100 microns area, there were 14-17 large nodes. By TEM, capillary sprouts were identified at the level beneath the last row of hypertrophied chondrocytes. These capillaries had voluminous endothelial cells rich in free ribosomes and rough endoplasmic reticulum. Endothelial cell nuclei were rounded and showed loose chromatin texture. Endothelial cells were connected by intermediate junctions and there was no basal lamina. Deeper into the metaphysis, arterioles and sinusoids were present. CONCLUSIONS: We conclude that the metaphyseal plate of the growing rat offers an optimal model to study vasculogenesis. Capillary sprouts can be readily identified, measured, and counted because they are located within a plane bordering against avascular cartilage. In addition, by using microvascular corrosion casting in SEM not only capillary sprouting per se but also different stages of neovascularization, indicated by differently sized nodular projections at the tip of vascular loops, can be studied in the growing long bone.

CSF-1 treatment promotes angiogenesis in the metaphysis of osteopetrotic (toothless, tl) rats.

It has recently been shown that following treatment with colony-stimulating factor-1 (CSF-1) the osteopetrotic condition in toothless (tl) rats greatly improves and growth is accelerated. We have examined the effects of such treatment on the microvasculature of the distal femoral chondro-osseous junction, a site where bone growth in length is coordinated with angiogenesis. Vascular casts and ultrastructural analyses of this region showed that, compared to untreated normal rats, untreated mutants showed little bone growth or angiogenesis. When mutants were treated with CSF-1 angiogenesis was markedly accelerated. These data show a remarkable effect of this growth factor on angiogenesis in this osteopetrotic mutation. Whether this effect of CSF-1 on angiogenesis is direct or indirect is not known and indicates that its effects on the normal microvasculature deserve further study.

New aspects of microvascular corrosion casting: a scanning, transmission electron, and high-resolution intravital video microscopic study.

We used intravital microscopy of small intestine and pancreas in order to show dynamic interactions between vascular wall and undiluted Mercox, because previous studies of ours have shown that Mercox diluted with monomeric methylmethacrylate penetrates cells in the vascular wall. Scanning and transmission electron microscopy were used to show three-dimensional pathways and correlating tissue structures, which cannot be identified in vivo. The microvascular diameters were not altered when the vasculature was flushed with saline/dextran solution using perfusion pressures between 70 and 140 mm Hg, but, in circumscribed areas, contraction of vascular wall was observed immediately after Mercox injection. This phenomenon was carried out by endothelial cells; pericytes were never present at the site of constrictions. Extravasation, i.e., leakage of the resin into the surrounding tissue, occurred in circumscribed areas regardless of the applied perfusion pressure. The resin also filled routes, which were not perfused with blood before casting. Scanning microscopy of corresponding specimens showed flattened cast channels, with impressions of valves and endothelial cell nuclear imprints characteristic of lymphatics. These results show that undiluted Mercox is a stimulus for vascular cellular components and that it changes the vascular wall permeability, resulting in extravasation and filling of lymphatics. Transmission electron microscopy showed that large vessels were homogeneously filled with resin and that cellular structures were not infiltrated with Mercox. Cut sections of the gold-coated surface of casts showed grooves up to 20 nm wide, suggestive of minimal deformation, while the abluminal surface of the metal film was almost smooth. Another proof of minimal deformation of undiluted Mercox casts is that the diameter of vessels was not altered during and after polymerization. Obtained casts are not fragile, as are casts of diluted Mercox, and phase separation does not occur, which would result in penetration of the cells in the vascular wall. For these reasons, the use of undiluted Mercox is recommended. Mixing 10 ml Mercox with 1 g catalyst resulted in complete polymerization within 5.5-7 min. This mixture can be used for casting biological specimens.

Ribosomes in the skeletal muscle filament lattice.

The compartmentalization of myosin isoforms within a muscle cell (Gauthier: J. Cell Biol. 110:693-701, 1990) suggests that myosin might be assembled directly into thick filaments at sites where it is synthesized. We therefore examined myofibrils by immunoelectron microscopy to determine whether ribosomes are associated with thick filaments under conditions in which new myosin can be identified. We used the embryonic chick anterior latissimus dorsi (ALD), a slow muscle that is induced, by curare, to synthesize a fast myosin isoform that is not normally present. Myosin was localized in situ, using a gold-labeled monoclonal antibody that recognizes the new isoform. The gold marker, as expected, was localized preferentially to the A band. There was an overall increase of fivefold in the number of gold particles per micron2 of A band in the curare-treated compared to the normal ALD, indicating that the labeled isoform was largely newly formed. There was a corresponding preferential distribution of ribosomes at the A band, especially in the H-band region, and the number of ribosomes per micron2 of A band was nearly twice as high in the curare-treated as in the normal muscle. Ribosomes were located between thick filaments, often aligned in rows. We conclude that ribosomes are located within the filament lattice, and therefore that they are available for local myosin synthesis.

Intracytoplasmic eosinophilic inclusion bodies in breast cyst fluids are giant lysosomes.

Electron microscopic (EM) studies were performed on a benign breast cyst fluid to determine the nature of intracytoplasmic eosinophilic inclusion bodies (EIBs). EIBs were usually found in macrophages and had the ultrastructural appearance of giant lysosomes. EIBs contained cellular material and granular debris; when viewed by EM, the contents were somewhat variable, depending on the nature of phagocytosed material and the extent of enzymatic breakdown. Usually the phagocytosed material had a rather homogeneous, finely granular texture. Occasionally cell remnants, including cytoplasmic organelles, were preserved inside the EIBs. Since the "host" macrophages were in various states of degeneration, the formation of EIBs appears to be related to a severe disturbance in the macrophage cellular function. This study showed that EIBs in cells from breast cyst fluids are not viral particles.