Localization of specific binding sites for 125I-TGF-beta1 to fenestrated endothelium in bone and anastomosing capillary networks in enamel organ suggests a role for TGF-beta1 in angiogenesis.

Previous studies have shown endothelial cells to be a major target for endocrine TGF-beta in several soft tissues in the normal growing rat. The potent effect of TGF-beta1 on bone formation prompted us to analyze in detail the localization of specific binding sites for endocrine TGF-beta in hard tissues. At 2.5 minutes after injection of 125I-TGF-beta1, specific binding, as demonstrated by quantitative radioautography, was localized to fenestrated endothelium participating in angiogenesis in the vascular invasion region of the growth plate in bone as well as to anatomizing capillary networks in the maturation zone of the enamel organ. At 15 minutes after injection, the bound ligand was internalized into endocytic vesicles of endothelial cells. In bone, quantitation revealed significant differences in receptor density between endothelia undergoing proliferation vs those in a state of elongation and anastomosis with neighboring endothelial cells. In the rat incisor, specific binding of 125I-TGF-beta1 to endothelium correlated with increased formation of anastomotic capillary networks. These studies identify differential specific binding sites of 125I-TGF-beta1 in angiogenically active endothelium, providing an important link between TGF-beta1, the endothelium, and hard tissue development.

Cell-specific expression of the parathyroid hormone (PTH)/PTH-related peptide receptor gene in kidney from kidney-specific and ubiquitous promoters.

The kidney is the major site of expression of the PTH/PTH-related peptide receptor (PTHR) gene. Previously we have shown that the PTHR gene is expressed from two promoters in kidney, an upstream kidney-specific promoter (P1) and a downstream promoter (P2) that is active in a wide variety of tissues. Here, we have used immunohistochemical and transcript-specific in situ hybridization techniques to map the expression of the PTHR gene and protein and to determine the distribution of P1- and P2-driven messenger RNAs in renal tissue. Immunohistochemical and immunoelectron microscopic analysis showed that PTHR protein is expressed on both basolateral and luminal membranes of proximal tubular epithelial cells, strongly suggesting a bipolar mode of action of PTH. Receptor protein also was detected on the surface of glomerular podocytes. Strikingly, immunoelectron microscopic analysis showed that endothelial cells of the peritubular vasculature, but not the glomerular vasculature, contain high levels of PTHR protein. We found that both P1 and P2 are expressed at moderate levels in both cortical and medullary epithelial cells of nephrons, correlating well with the immunohistochemical localization of PTHR protein. However, although abundant transcripts were detected in peritubular endothelial cells with P1-specific and coding sequence probes, P2-specific expression was not observed in these cells. These results provide evidence that the physiological effects of PTH- and/or PTH-related peptide on renal tubular function may be mediated not only through direct effects on epithelial cells but also indirectly through endothelial cell-based signaling. In addition to expression in vascular endothelial cells, high levels of P1-specific, but not P2-specific, PTHR messenger RNA were detected in vascular smooth muscle. Taken together, these experiments provide evidence for strong PTHR gene expression in renal vascular tissues. Moreover, given that previous studies have shown that P2, but not P1, is active in other tissues with an abundant vasculature, our results suggest that regulation of PTHR gene expression in renal vascular tissue is distinct from that of other organs.

Programmed cell death of chondrocytes and aberrant chondrogenesis in mice homozygous for parathyroid hormone-related peptide gene deletion.

In previous work we showed that the chondrodysplastic phenotype of mice homozygous for a null mutation of the PTH-related peptide (PTHrP) gene was due in part to reduced proliferation and aberrant differentiation of growth plate chondrocytes. In the present study we have extended those observations by examining chondrocytes for evidence of PTH/PTHrP receptor expression, proliferation, and programmed cell death. Receptor messenger RNA and protein were expressed in chondrocytes in the resting and proliferative zones of both wild-type and mutant mice. In normal animals, expression was abundant in the area of transition between proliferative and hypertrophic chondrocytes and absent from cells in the lower hypertrophic region. On the other hand, the hypertrophic zone in mutant mice contained nonhypertrophic chondrocytes, which exhibited characteristics of proliferating cells, including PTH/PTHrP receptor expression, [3H]thymidine incorporation, and expression of proliferating cell nuclear antigen. In contrast to the situation in normal animals, some cells adjacent to the zone of vascular invasion in mutant growth plates showed biochemical and morphological evidence of programmed cell death. In addition to these alterations in the maturation of growth plate chondrocytes, homozygous mutants demonstrated signs of aberrant differentiation of periosteal precursor cells. In some specimens, clusters of chondrocytes embedded in a cartilaginous matrix were observed between the layers of periosteal osteoblasts and the bony collar in the sterna and tibiae of mice homozygous for a null mutation of the PTHrP gene. Taken together, these observations indicate that PTHrP plays a pivotal role in the orderly progression of chondrocytes through stages of proliferation, differentiation, and programmed cell death in the epiphyseal growth plate and may also facilitate the commitment of precursors to cells of the chondrocytic or osteoblastic lineages.

Lattice fringe continuity in the absence of crystal continuity in enamel.

Since high-resolution transmission electron microscopy (HRTEM) provides information on a nearly atomic level, the confidence level with this method is very high. Thus, when lattice fringe continuity is found between two enamel crystals in proximity, such continuity has been taken as evidence of crystal fusion (Daculsi and Kerebel, 1977). Similarly, selected-area dark-field (SADF) electron microscopic imaging has been used to study the axial and spatial orientation of crystals. These studies have shown that there is apparent continuity between enamel and dentin crystals (Arsenault and Robinson, 1989). This observation supported the hypothesis that enamel crystallites are initiated by crystallites in dentin. We have used both HRTEM and SADF methods to identify instances of spatial relationship between crystallites in sections of rat incisor enamel and shark enameloid. In each instance of apparent continuity, goniometric tilting was used to examine the continuous interface. All instances where two crystallites seemed to come into contact, and where HRTEM imaging showed the lattice fringes to be directly continuous, were separated into individual crystallites when the specimen was tilted a few degrees. Thus, adjacent crystallites can show lattice fringe continuity in the absence of real crystallite contact. When instances of overlapping crystallites were examined by SADF imaging, the overlapping crystallites gave a single bright image. Goniometric tilting revealed separate crystallites. Thus, neither lattice fringe continuity nor image continuity under SADF can be used as evidence of crystal continuity unless goniometric rotation and tilting are applied when spatial relationships are suspected.

Immunolocalization of the cation-independent mannose 6-phosphate receptor and cathepsin B in the enamel organ and alveolar bone of the rat incisor.

In order to examine our hypothesis that maturation ameloblasts could degrade the enamel matrix in a manner analogous to bone resorption mediated by osteoclasts, we have assessed the distribution of lysosomal enzymes in the enamel organ by immunolocalizing the cation-in-independent mannose 6-phosphate receptor (MPR) and the lysosomal enzyme cathepsin B at all stages of amelogenesis. Secretory ameloblasts showed strong immunoreactivity for MPR in the supranuclear Golgi region and in the cytoplasm between the Golgi region and the distal junctional complexes. However, cathepsin B immunoreactivity was mainly seen in the distal portion of Tomes' process, which was unreactive for MPR immunogenicity. In maturation ameloblasts, the MPR was observed on the ruffled border of the ruffle-ended ameloblast (RA) but not on the distal cell membrane of the smooth-ended ameloblast (SA), although both cell types demonstrated strong immunoreactivity for MPR in the Golgi region. Immunoreactive cathepsin B was seen at the distal ends of both RA and SA. It is postulated that the nascent lysosomal enzymes bind to the mannose 6-phosphate receptors which target them not only to intracellular lysosomes, but also to the ruffled border of maturation ameloblasts where these enzymes are secreted into the enamel. Since MPR and lysosomal enzymes were also detected on the ruffled border of osteoclasts (Ocl) adjacent to alveolar bone, our immunocytochemical approach provides strong evidence for a similarity between the maturation process in enamel, as mediated by the ruffle-ended maturation ameloblasts, and bone resorption mediated by osteoclasts. This study has established that a common mechanism, based on MPR-targeted lysosomal secretion and matrix degradation, is basic to the maturation process involved in calcified tissues as different as bone and enamel.

Haploinsufficiency of parathyroid hormone-related peptide (PTHrP) results in abnormal postnatal bone development.

Although apparently phenotypically normal at birth, mice heterozygous for inactivation of the gene encoding parathyroid hormone-related peptide (PTHrP) develop haplotype insufficiency by 3 months of age. In addition to histologic and morphologic abnormalities similar to those seen in homozygous mutants, heterozygous animals demonstrated alterations in trabecular bone and bone marrow. These included metaphyseal bone spicules which were diminished in volume, irregularly distributed, and less well developed than those seen in age-matched controls as well as bone marrow, which contained an inordinate number of adipocytes. A substantial reduction in PTHrP mRNA was detected in heterozygous tissue, while circulating parathyroid hormone (PTH) and calcium concentrations were normal. Thus, while a physiologic concentration of PTH was capable of maintaining calcium homeostasis, it was incapable of compensating for PTHrP haploinsufficiency in developing bone. In normal animals, both PTHrP and the PTH/PTHrP receptor were expressed predominantly in chondrocytes situated throughout the proliferative zone of the tibial growth plate. In the metaphysis, the PTH/PTHrP receptor was identified on osteoblasts and preosteoblastic cells situated in the bone marrow, while PTHrP was expressed only by osteoblasts. These observations indicate that postnatal bone development involves susceptible pathways that display exquisite sensitivity to critical levels of PTHrP and imply that the skeletal effects of PTH are influenced by locally produced PTHrP. Moreover, identification of both the ligand and its N-terminal receptor in metaphyseal osteoblasts and their progenitors suggests an autocrine/paracrine role for the protein in osteoblast differentiation and/or function. Impairment in this function as a consequence of PTHrP haploinsufficiency may critically influence the course of bone formation, resulting in altered trabecular architecture and perhaps low bone mass and increased bone fragility.

Nucleolar localization of parathyroid hormone-related peptide enhances survival of chondrocytes under conditions that promote apoptotic cell death.

Parathyroid hormone-related peptide (PTHrP) is a mediator of cellular growth and differentiation as well as a cause of malignancy-induced hypercalcemia. Most of the actions of PTHrP have been attributed to its interaction with a specific cell surface receptor that binds the N-terminal domain of the protein. Here we present evidence that PTHrP promotes some of its cellular effects by translocating to the nucleolus. Localization of transiently expressed PTHrP to the nucleolus was dependent on the presence of a highly basic region at the carboxyl terminus of the molecule that bears homology to nucleolar targeting sequences identified within human retroviral (human immunodeficiency virus type 1 and human T-cell leukemia virus type 1) regulatory proteins. Endogenous PTHrP also localized to the nucleolus in osseous cells in vitro and in vivo. Moreover, expression of PTHrP in chondrocytic cells (CFK2) delayed apoptosis induced by serum deprivation, and this effect depended on the presence of an intact nucleolar targeting signal. The present findings demonstrate a unique intracellular mode of PTHrP action and a novel mechanism by which this peptide growth factor may modulate programmed cell death.

Specific binding of endocrine transforming growth factor-beta 1 to vascular endothelium.

The presentation of recombinant biologically active 125I-TGF-beta 1 via the bloodstream to potential target cells in mice and rats was evaluated by quantitative light and electron microscope radioautography. Specificity was evaluated by in vivo competition with excess unlabeled TGF-beta 1, and integrity of the ligand at the binding site was demonstrated by trichloroacetic acid precipitation after extraction from tissues. The distribution of radiolabel at 2.5, 15, 30, 45, and 60 min after 125I-TGF-beta 1 injection revealed radiolabel principally over microvasculature endothelium but at times > 2.5 min over endothelial endocytic components indicative of internalization. Nonspecific binding of 125I-TGF-beta 1 to the apex of the proximal convoluted tubule of the kidney indicated it as the likely site of rapid clearance of TGF-beta 1 from the circulation, while a comparison of the binding of 125I-TGF-beta 1 (endothelial) to that of 125I-TGF-beta 1 complexed with alpha 2-macroglobulin-methylamine (liver parenchyma) indicated that clearance of TGF-beta 1 complexed alpha 2-macroglobulin was likely via the hepatic alpha 2-macroglobulin receptor. The endothelial TGF-beta receptors uncovered here are likely involved in the local regulatory mechanism of leukocyte and monocyte adhesion and tissue infiltration regulated by endocrine TGF-beta 1.

Failure to demonstrate a protein coat on enamel crystallites by morphological means.

Phosphotungstic acid (PTA) treatment of section of Epon-embedded enamel dissolves the crystallites and stains material postulated to be crystal-bound proteins. Alternative, capillarity forces within the channels left after crystallite removal may draw in PTA. This prediction was tested on three systems. (1) Protein free synthetic hydroxyapatite was embedded in Epon; treatment of thin sections with PTA removed most crystals, leaving empty holes outlined by stain that could not represent protein. (2) Sections of rat incisor enamel were treated with PTA and then re-embedded in Epon and sectioned at 90 degrees to the original plane. In these sections-of-section the cut ends of dissolved crystallite profiles were coated with stain. To determine if stained protein coats can be detected in the absence of the crystallite profiles, Epon sections were partially demineralized with formic acid, re-embedded in Epon and sections-of section were PTA treated. Previously extracted crystallites left no stained coats, and only the crystallites that were not removed by formic acid left PTA-stained outlines. (3) PTA-treated sections of dogfish shark enameloid were flooded with 5-nm colloidal gold particles and sections-of-section were prepared. The presence of gold particles on the section surface and in holes previously occupied by crystallites suggested that PTA solution could also be sucked into similar holes. It is concluded that PTA outlines are not crystal-bound proteins but artefacts caused by stain lining holes left in the section when the crystallites have been extracted.

Morphologic and radioautographic studies of bone formation in relation to titanium implants using the rat tibia as a model.

A rat tibia model was developed to analyze bone formation leading to osseointegration with threaded titanium implants. Miniaturized titanium implants were placed in the anterior aspect of the upper tibia of rats weighing 350 g. Twenty-four rats were involved; 12 rats were sacrificed at 6 weeks, and another two rats were sacrificed weekly for 6 weeks following implantation. Four days prior to sacrifice, the animals were injected intraperitoneally with 3H-proline (10 microCi/g body weight). The rats were fixed by perfusion with 5% glutaraldehyde, and the tibia were decalcified and embedded in Epon. The implants were removed from the embedded bone by fracturing, and the specimens re-embedded. Sections were prepared for light and electron microscopy and radioautography. Morphologic observations indicated that implant placement resulted in bone necrosis and resorption. This was followed by bone growth from the old bone surfaces filling the threads with vascular channels and new lamellar bone. Osseointegration was complete at 6 weeks in all animals examined. Electron microscopy at various places along the bone-titanium interface showed an amorphous layer, a granular electron-dense layer, or a layer of uncalcified collagen fibrils. At each week after surgery, radioautography showed the position and thickness of new bone labeled with 3H-proline during the last 4 days. Radiolabeled new bone was deposited only on previously existing bone and extended toward the available space. Since there was no apparent relationship between the implant and the new bone, it was suggested that titanium is biocompatible, but not necessarily osteoinductive.

Gene expression of epimorphin in rat incisor ameloblasts.

Epimorphin has been recently identified as an important factor in the morphogenesis of epithelial cells. A cDNA encoding epimorphin from skin of newborn mice was cloned by the polymerase chain-reaction technique before the preparation of digoxigenin-labelled cRNA probes. In situ hybridization of longitudinal sections of rat incisors revealed a distinct pattern of expression of epimorphin mRNA in the ameloblast layer. Epimorphin mRNA was detected from the presecretory stage up to the beginning of the maturation stage of amelogenesis. With the identification of this expression by epithelial-derived cells, i.e. ameloblasts, it is thought likely that epimorphin is one of the factors that modulate the differentiation cascade of ameloblasts in the course of amelogenesis.

A compilation of partial sequences of randomly selected cDNA clones from the rat incisor.

The formation of tooth organs is regulated by a series of developmental programs. We have initiated a genome project with the ultimate goal of identifying novel genes important for tooth development. As an initial approach, we constructed a unidirectional cDNA library from the non-calcified portion of incisors of 3- to 4-week-old rats, sequenced cDNA clones, and classified their sequences by homology search through the GenBank data base and the PIR protein data base. Here, we report partial DNA sequences obtained by automated DNA sequencing on 400 cDNA clones randomly selected from the library. Of the sequences determined, 51% represented sequences of new genes that were not related to any previously reported gene. Twenty-six percent of the clones strongly matched genes and proteins in the data bases, including amelogenin, alpha 1(I) and alpha 2(I) collagen chains, osteonectin, and decorin. Nine percent of clones revealed partial sequence homology to known genes such as transcription factors and cell surface receptors. A significant number of the previously identified genes were expressed redundantly and were found to encode extracellular matrix proteins. Identification and cataloging of cDNA clones in these tissues are the first step toward identification of markers expressed in a tissue- or stage-specific manner, as well as the genetic linkage study of tooth anomalies. Further characterization of the clones described in this paper should lead to the discovery of novel genes important for tooth development.

Development of a rat tibia model for morphological studies of the interface between bone and a titanium implant.

Over the last decade, osseointegrated dental implants have become an integral part of the restorative dental armamentarium. Reproducible success rates approaching 100% further emphasize the importance and value of this treatment modality. Still, a significant waiting period is required between implant placement and prosthesis delivery, which necessitates a two-surgery approach for implant protection during healing. This article discusses the development of an animal model that is being used to investigate methods of manipulating the healing process of bone next to dental implants. Osseointegration has been shown to occur at 6 weeks in the rat tibia model, demonstrating many of the same characteristics seen in humans. The successful manipulation of bone next to the implant surface may ultimately lead to a surgical protocol for the placement of dental implants requiring only one surgery, significantly reducing the overall healing time.

Parathyroid hormone-related peptide-depleted mice show abnormal epiphyseal cartilage development and altered endochondral bone formation.

To elucidate the role of PTHrP in skeletal development, we examined the proximal tibial epiphysis and metaphysis of wild-type (PTHrP-normal) 18-19-d-old fetal mice and of chondrodystrophic litter mates homozygous for a disrupted PTHrP allele generated via homologous recombination in embryonic stem cells (PTHrP-depleted). In the PTHrP-normal epiphysis, immunocytochemistry showed PTHrP to be localized in chondrocytes within the resting zone and at the junction between proliferative and hypertrophic zones. In PTHrP-depleted epiphyses, a diminished [3H]thymidine-labeling index was observed in the resting and proliferative zones accounting for reduced numbers of epiphyseal chondrocytes and for a thinner epiphyseal plate. In the mutant hypertrophic zone, enlarged chondrocytes were interspersed with clusters of cells that did not hypertrophy, but resembled resting or proliferative chondrocytes. Although the overall content of type II collagen in the epiphyseal plate was diminished, the lacunae of these non-hypertrophic chondrocytes did react for type II collagen. Moreover, cell membrane-associated chondroitin sulfate immunoreactivity was evident on these cells. Despite the presence of alkaline phosphatase activity on these nonhypertrophic chondrocytes, the adjacent cartilage matrix did not calcify and their persistence accounted for distorted chondrocyte columns and sporadic distribution of calcified cartilage. Consequently, in the metaphysis, bone deposited on the irregular and sparse scaffold of calcified cartilage and resulted in mixed spicules that did not parallel the longitudinal axis of the tibia and were, therefore, inappropriate for bone elongation. Thus, PTHrP appears to modulate both the proliferation and differentiation of chondrocytes and its absence alters the temporal and spatial sequence of epiphyseal cartilage development and of subsequent endochondral bone formation necessary for normal elongation of long bones.

Zinc iodide-osmium tetroxide impregnation of the "tubulo-vesicular system" in Tomes' process of the rat incisor ameloblast.

Zinc iodide-osmium tetroxide (ZIO) is a nonspecific but selective impregnation method that visualizes a tubulo-vesicular system in cells. The detailed structure and three-dimensional distribution of this ZIO-impregnated system was studied in the Tomes' process of secretory ameloblasts in the rat incisor. The ZIO-impregnated system consisted of an extensive array of smooth membrane-bound thick and thin tubules and vesicles. The interconnected thick and thin tubules formed a complex "core network" in the central cytoplasm of Tomes' process that enmeshed and often surrounded individual secretory granules. From the core network, radial branches extended toward the smooth cell membrane of the interdigitating portion of Tomes' process. Although the core network and branches frequently appeared connected to the secretory granules and the cell membrane, stereo-pair electron microscopy failed to show conclusive evidence of such continuity. However, many coated vesiclelike structures were attached to the core network and its branches. No special relationship was found between interrod and rod secretory sites and the tubulo-vesicular network. In thick sections, the ZIO-impregnated tubulo-vesicular network occupied a considerable volume of cytoplasm. The vinblastine-labile nature of this network as demonstrated previously (Nanci et al., 1987) indicated that the system undergoes rapid and extensive turnover. Considering the dynamic nature and sheer volume of the tubulo-vesicular system, we propose that it be regarded as a major cell organelle.

Localization of epidermal growth factor receptors in cells of the enamel organ of the rat incisor.

Epidermal growth factor (EGF) is a peptide shown to effect precocious incisor tooth eruption in rat pups. Binding sites for EGF were visualized in the continuously erupting adult rat incisor by light and electron microscope radioautography after in vivo injection of 125I-EGF. These binding sites represented EGF receptors because of (i) competition between 125I-EGF binding at 2 min after injection and a coinjected excess of unlabeled EGF; (ii) the receptor-mediated endocytosis of 125I-EGF at 15 and 30 min after injection; and (iii) the demonstration of EGF receptor kinase activation in vivo. The stem and the mitotic cells in the epithelial odontogenic organ at the growing end of the tooth develop into two nondividing layers of the enamel organ: (i) ameloblasts which secrete enamel and are subsequently involved in the enamel maturation process, and (ii) papillary layer cells situated between the blood supply and the ameloblasts. Although few EGF receptors were present at the mitotic end, receptor density was highest at the mature end of the enamel organ. High levels of 125I-EGF binding were found on papillary layer cells and ruffle-ended, but not smooth-ended, ameloblasts. This implies a cyclical exteriorization and internalization of receptors during modulations between the two cell types. These data suggest that the EGF receptor mediates a major function of the enamel organ in the formation of enamel.

Organization of crystals in enamel.

It has been variously suggested that the organic matrix associated with the mineral phase of enamel is present as either calcified fibrils, central dark lines, peripheral sheaths around hexagonal crystals, or organic ghosts apparently contained within crystal profiles. The most consistent findings confirm the crystal ghost conception. Grid decalcification of nearly mature sectioned enamel and staining revealed hollow, noncrystalline structures whose external measurements were statistically identical to those of the dissolved crystallites, but with internal measurements too small to accommodate the crystallites. To explain these apparent ghosts in view of the incompatibility of ghosts with crystal structure, it has been proposed that the crystallites are not hexagonal in cross-sections and the hexagonal appearance is due to projections of parallelepiped-shaped crystallite segments with cut surfaces that are rhomboidal in shape. Material on the surface of such profiles would project as if it were contained within the profile. Hexagonal forms could not be demonstrated in isolated crystallites examined by transmission electron microscopy, high-resolution scanning electron microscopy, and replicas made of the isolated crystallite preparations examined by transmission electron microscopy. Existing evidence does not rule out the possibility that the noncrystalline profiles represent stain drawn into the holes left by the dissolved crystallites as a result of high capillarity forces.

Banding patterns in rat incisor enamel stained by histochemical complexing methods for calcium.

A characteristic banding pattern can be visualized at the surface of the rat incisor in the maturation zone of amelogenesis by staining with glyoxal bis(2-hydroxyanil) (GBHA). Other banding patterns can be obtained with certain histological and fluorochrome stains and by radioautography following 45Ca injection. In this study, several histochemical reagents known to complex with different states of calcium were used to stain the surface of enamel. Rat incisors were quickly dissected and immediately immersed in solutions containing the following calcium-binding reagents: arsenazo III, calmagite, murexide, N,N-naphthaloylhydroxylamine, and calcein. Routinely, one contralateral lower incisor from each pair was counterstained with GBHA in order to relate each of the staining patterns to the banded distribution of maturation ameloblasts that is reflected by the characteristic GBHA staining pattern in the enamel. Each of the reagents used in this study demonstrated a staining pattern consisting of a series of broad bands running transversely and obliquely across the enamel. In all cases, the dyes stained predominantly that enamel associated with ruffle-ended ameloblasts, i.e. enamel left unstained by GBHA. Some of the reagents also stained enamel in the secretion zone. The appearance and distribution of the staining patterns reflect the banded distribution of maturation ameloblasts and appear to be controlled on a time scale related to the rapid modulation of these cells.