ERRα promotes breast cancer cell dissemination to bone by increasing RANK expression in primary breast tumors

Bone is the most common metastatic site for breast cancer. Estrogen-related-receptor alpha (ERRα) has been implicated in cancer cell invasiveness. Here, we established that ERRα promotes spontaneous metastatic dissemination of breast cancer cells from primary mammary tumors to the skeleton. We carried out cohort studies, pharmacological inhibition, gain-of-function analyses in vivo and cellular and molecular studies in vitro to identify new biomarkers in breast cancer metastases. Meta-analysis of human primary breast tumors revealed that high ERRα expression levels were associated with bone but not lung metastases. ERRα expression was also detected in circulating tumor cells from metastatic breast cancer patients. ERRα overexpression in murine 4T1 breast cancer cells promoted spontaneous bone micro-metastases formation when tumor cells were inoculated orthotopically, whereas lung metastases occurred irrespective of ERRα expression level. In vivo, Rank was identified as a target for ERRα. That was confirmed in vitro in Rankl stimulated tumor cell invasion, in mTOR/pS6K phosphorylation, by transactivation assay, ChIP and bioinformatics analyses. Moreover, pharmacological inhibition of ERRα reduced primary tumor growth, bone micro-metastases formation and Rank expression in vitro and in vivo. Transcriptomic studies and meta-analysis confirmed a positive association between metastases and ERRα/RANK in breast cancer patients and also revealed a positive correlation between ERRα and BRCA1mut carriers. Taken together, our results reveal a novel ERRα/RANK axis by which ERRα in primary breast cancer promotes early dissemination of cancer cells to bone. These findings suggest that ERRα may be a useful therapeutic target to prevent bone metastases.

Craniofacial and Dental Defects in the Col1a1Jrt/+ Mouse Model of Osteogenesis Imperfecta

Certain mutations in the COL1A1 and COL1A2 genes produce clinical symptoms of both osteogenesis imperfecta (OI) and Ehlers-Danlos syndrome (EDS) that include abnormal craniofacial growth, dental malocclusion, and dentinogenesis imperfecta. A mouse model (Col1a1(Jrt)/+) was recently developed that had a skeletal phenotype and other features consistent with moderate-to-severe OI and also with EDS. The craniofacial phenotype of 4- and 20-wk-old Col1a1(Jrt)/+ mice and wild-type littermates was assessed by micro-computed tomography (µCT) and morphometry. Teeth and the periodontal ligament compartment were analyzed by µCT, light microscopy/histomorphometry, and electron microscopy. Over time, at 20 wk, Col1a1(Jrt)/+ mice developed smaller heads, a shortened anterior cranial base, class III occlusion, and a mandibular side shift with shorter morphology in the masticatory region (maxilla and mandible). Col1a1(Jrt)/+ mice also had changes in the periodontal compartment and abnormalities in the dentin matrix and mineralization. These findings validate Col1a1(Jrt)/+ mice as a model for OI and EDS in humans.

Divergent regulation of the Osteopontin promoter by the estrogen receptor-related receptors is isoform- and cell context dependent

We sought to determine whether the estrogen receptor-related receptor gamma (mEsrrg) regulated the Osteopontin (Opn) promoter through the same AP1/CAAT box element that we have previously described for mEsrra. In HeLa cells mEsrrg used an additional site present in the 5'UTR, while in ROS17/2.8 cells the AP1/CAAT site was not used, but a completely novel site surrounding the transcription start site was used. We also find that in ROS17/2.8 cells mEsrra repressed, while mEsrrg activated the Opn promoter. None of the sites identified conform to established Esrr response elements (ERREs). Additionally, the two reported mEsrrg protein isoforms showed differences in their activation potential. Mutations in the activation function 2 (AF2) of mEsrra, predicted to abolish activation, surprisingly turned mEsrra into a better activator. In contrast, similar AF2 mutations in Esrrg2 abolished its ability to activate the Opn promoter. Mutation of the DNA binding domain of mEsrra/g2 abolished transcriptional activity in HeLa and ROS17/2.8 cells. Our data indicate, first, that the two Esrr isoforms regulate Opn in a cell context-dependent manner. Second, they suggest that although the DNA binding domains of mEsrra and mEsrrg are 93% identical and required for regulation, the receptors bind to distinct Opn promoter elements, suggesting that the two isoforms may co-regulate Opn, and perhaps other genes, without competing for the same site in the promoter. Finally, the results suggest that each isoform interacts differently with co-activators and co-repressors, as highlighted by the AF2 mutation that turns mEsrra into a better activator but abolishes activity of Esrrg2.

Deficiency in acellular cementum and periodontal attachment in bsp null mice

Bone sialoprotein (BSP) is an extracellular matrix protein found in mineralized tissues of the skeleton and dentition. BSP is multifunctional, affecting cell attachment and signaling through an RGD integrin-binding region, and acting as a positive regulator for mineral precipitation by nucleating hydroxyapatite crystals. BSP is present in cementum, the hard tissue covering the tooth root that anchors periodontal ligament (PDL) attachment. To test our hypothesis that BSP plays an important role in cementogenesis, we analyzed tooth development in a Bsp null ((-/-)) mouse model. Developmental analysis by histology, histochemistry, and SEM revealed a significant reduction in acellular cementum formation on Bsp (-/-) mouse molar and incisor roots, and the cementum deposited appeared hypomineralized. Structural defects in cementum-PDL interfaces in Bsp (-/-) mice caused PDL detachment, likely contributing to the high incidence of incisor malocclusion. Loss of BSP caused progressively disorganized PDL and significantly increased epithelial down-growth with aging. Bsp (-/-) mice displayed extensive root and alveolar bone resorption, mediated by increased RANKL and the presence of osteoclasts. Results collected here suggest that BSP plays a non-redundant role in acellular cementum formation, likely involved in initiating mineralization on the root surface. Through its importance to cementum integrity, BSP is essential for periodontal function.

Estrogen receptor-related receptor-alpha (ERR-alpha) is dysregulated in inflammatory arthritis

OBJECTIVES

Subchondral bone loss is a characteristic feature of inflammatory arthritis. Recently, estrogen receptor-related receptor-alpha (ERR-alpha), an orphan nuclear receptor, has been found to be involved in activation of macrophages. We hypothesized that ERR-alpha which is expressed and also functional in articular chondrocytes, osteoblasts and osteoclasts, may be involved in rodent models of inflammatory arthritis.

METHODS

Erosive arthritis was induced in DBA/1 mice by injection of type II collagen in Freund's complete adjuvant. RNA was isolated from the bone and joints and expression of ERR-alpha and cartilage (GDF5 and Col2a1) and bone [bone sialoprotein (BSP) and osteocalcin (OCN)] markers was analysed by semi-quantitative PCR.

RESULTS

We report for the first time that the expression of ERR-alpha is dysregulated in bones and joints in a mouse model of inflammatory arthritis. Specifically, we show that ERR-alpha expression is down-regulated early in bone and later in joints of mice with type II CIA. Concomitantly, temporal changes were observed in GDF-5 and Col2a1 expression in joints following both initial injection and booster injection of type II collagen. Similarly, down-regulation of ERR-alpha mRNA expression in subchondral bone in mice with induced joint inflammation was also paralleled by down-regulation of markers of bone formation (BSP, OCN).

CONCLUSIONS

These data suggest that dysregulation of ERR-alpha expression may precede and contribute to the destruction of cartilage and bone accompanying inflammatory arthritis.

The orphan nuclear estrogen receptor-related receptor-alpha regulates cartilage formation in vitro: implication of Sox9

We report for the first time the expression of estrogen receptor-related receptor (ERR)-alpha in fetal and adult rat chondrocytes in growth plate and articular cartilage and the rat chondrogenic cell line C5.18 cells in vitro. ERRalpha mRNA and protein were expressed from proliferating chondrocyte to mature chondrocyte stages. We show that overexpressing ERRalpha in C5.18 cell cultures induces an increase in Sry-type high-mobility-group box transcription factor (Sox)-9 expression, a master gene in cartilage formation. In parallel, we report Sox9 promoter regulation by ERRalpha in C5.18 cells. To assess a functional role for ERRalpha in chondrogenesis, its expression was blocked by antisense oligonucleotides in C5.18 cell cultures, and this led to inhibition of cartilage formation associated with down-regulation of Sox9 and Indian hedgehog expression and maturation of proliferating chondrocytes into hypertrophic chondrocytes in vitro. Together these results implicate ERRalpha in the formation and maintenance of cartilage and also suggest that agonists and antagonists of ERRalpha may be useful as therapeutic agents in a wide variety of diseases affecting cartilage and joints.

Estrogen receptor-related receptor alpha: a mediator of estrogen response in bone

Estrogen receptor-related receptor-alpha (ERRalpha) is an orphan nuclear receptor with sequence homology to the estrogen receptors, ERalpha/beta, but it does not bind estrogen. However, several recent studies suggest that ERRalpha not only plays a functional role in osteoblasts but also impinges on the estrogen axis in bone, as it does in at least certain other estrogen target tissues. We summarize here data on ERRalpha and its cellular and molecular modes of action that have broad implications for considering the potential role of this orphan receptor as a new therapeutic target in osteopenic disorders such as osteoporosis as well as other estrogen-responsive conditions.

Expression of leukemia inhibitory factor (LIF)/interleukin-6 family cytokines and receptors during in vitro osteogenesis: differential regulation by dexamethasone and LIF

The leukemia inhibitory factor/interleukin-6 (LIF/IL-6) family of cytokines is known to play a major role in bone physiology. Although much work has focused on the regulation of bone resorption by IL-6 and related cytokines, their effects on osteoblast development and bone formation have not been as well studied. Previously, we reported that LIF inhibits, in a non-IL-6-dependent manner, osteoblast differentiation and bone nodule formation in the rat calvaria (RC) model, an effect that is antagonized by dexamethasone (Dex). The culture time-sensitive window suggested that LIF targets late preosteoblasts or early osteoblasts, and that this stage-specific effect coincided with a period of low endogenous production of LIF and IL-6. To detect potential crosstalk between members of this family, we have extended these observations by assessing the expression levels of other LIF/IL-6 cytokines (CNTF, OSM, IL-11, CT-1) and their receptors in the same RC cell model treated with or without LIF or Dex. Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis showed that IL-11 and its receptor, CNTF and its receptor, LIFR, and gp130 were constitutively expressed throughout the culture period. Expression of CT-1 and OSM increased with culture time - that is, with osteoblast differentiation - whereas the specific receptor for OSM (OSMR) was highly expressed at early timepoints and either plateaued or decreased thereafter. Continuous treatment with Dex (10(-8) mol/L) inhibited the endogenous production of IL-6, LIF, OSM, IL-11R, and OSMR, but had no detectable effect on the expression of IL-11, CT-1, CNTF, CNTFR, LIFR, or gp130. Finally, treatment with exogenously added LIF stimulated IL-6, LIF, LIFR, and OSMR, but had no other detectable effects. These data indicate that multiple members of the LIF/IL-6 family and their receptors are expressed in RC cell cultures, and are differentially regulated by Dex and LIF, suggesting that these cytokines play a complex and interdependent role, further modulated by glucocorticoid levels, in osteoprogenitor differentiation and bone nodule formation.

Biphasic effects of leukemia inhibitory factor on osteoblastic differentiation

Leukemia inhibitory factor (LIF) is a cytokine produced by multiple cell types including osteoblasts and which is active on bone metabolism. We have previously shown that in a bone nodule forming in vitro model of osteogenesis, the fetal rat calvaria (RC) cell model, LIF inhibits osteoblast differentiation, acting on late osteoprogenitors and/or early osteoblasts. These results are in contrast to in vivo experiments, in which LIF has been found to increase bone formation. To resolve this discrepancy, we have tested the effect of LIF on rat bone marrow (RBM) stromal cell cultures, an in vitro model encompassing earlier osteoprogenitor stages. LIF inhibited cell growth in early, proliferating RBM cultures, but increased the culture saturation density. The effect of LIF on bone nodule formation in this model was cell density dependent and biphasic. Continuous treatment with LIF reduced the number of bone nodules present in confluent, more mature cultures, and the inhibitory effect was strongest when cells were plated at higher cell density than lower. In contrast, during the early stages of RBM culture, nodule numbers were higher in LIF-treated dishes than in controls, and this effect was greater in lower density cultures. Acute LIF treatment restricted to early time points increased the final number of bone nodules formed in mature RBM cell cultures, but not in RC cell cultures. Our results indicate that LIF exerts complex, stage-specific effects on osteoprogenitor recruitment, differentiation, and bone formation, and that the effects are cell nonautonomous, in the rat bone marrow stromal cell model. J. Cell. Biochem. Suppl. 36: 63-70, 2001.

Individual osteoblasts in the developing calvaria express different gene repertoires

Several studies in vitro and a few in vivo have suggested that mature osteoblasts heterogeneously express osteoblast markers. In one recent study of the osteoblasts associated with bone nodules formed in vitro in rat calvaria cell populations, extensive diversity was documented in the overall gene repertoires expressed. To address whether comparable heterogeneity is evident in vivo, we investigated the expression of nine osteoblast lineage markers by both in situ hybridization and immunohistochemistry. At 21 days of fetal rat development, the calvaria is a rapidly growing bone with distinct maturational zones that are readily observed in coronal sections; that is, an osteogenic front emerging at sagittal and coronal sutures is adjacent to areas of growing trabeculae of bone, followed by more mature areas of remodeling bone. Based on expression patterns, markers can be divided into two categories. One category comprises markers that are globally expressed by all osteoblasts irrespective of their position in the calvaria. Of those tested, only two, alkaline phosphatase and the pth/pthrp receptor, fit into this category. All other markers analyzed, including transcription factors (c-fos and msx-2), matrix molecules (bone sialoprotein, osteopontin, and osteocalcin), and a hormone (pthrp), were differentially expressed only in subpopulations of osteoblasts, based on cell maturational status, environment (ectocranial vs. endocranial surfaces), and microenvironment (adjacent osteoblasts). Preosteoblasts and osteocytes in different regions of the calvaria also expressed different subsets of the lineage markers. Mechanisms responsible for generating differential gene expression profiles appear to be both transcriptional and posttranscriptional. These results indicate that postproliferative, morphologically indistinguishable osteoblasts are not a homogeneous class of cells, but instead are molecularly diverse. The present results also raise the possibility that lineage progression and/or maintenance of the differentiated state may be adaptable in the calvaria.

Developmental expression and tissue distribution of Phex protein: effect of the Hyp mutation and relationship to bone markers

Mutations in PHEX, a phosphate-regulating gene with homology to endopeptidases on the X chromosome, are responsible for X-linked hypophosphatemia (XLH). The murine Hyp homologue has the phenotypic features of XLH and harbors a large deletion in the 3' region of the Phex gene. We characterized the developmental expression and tissue distribution of Phex protein, using a monoclonal antibody against human PHEX, examined the effect of the Hyp mutation on Phex expression, and compared neprilysin (NEP), osteocalcin, and parathyroid hormone/parathyroid hormone-related protein (PTH/PTHrP) receptor gene expression in bone of normal and Hyp mice. Phex encodes a 100- to 105-kDa glycoprotein, which is present in bones and teeth of normal mice but not Hyp animals. These results were confirmed by in situ hybridization (ISH) and ribonuclease protection assay. Phex protein expression in femur and calvaria decreases with age, suggesting a correlation between Phex expression and bone formation. Immunohistochemical studies detected Phex protein in osteoblasts, osteocytes, and odontoblasts, but not in osteoblast precursors. In contrast to Phex, the abundance of NEP messenger RNA (mRNA) and protein is not significantly altered in Hyp bone. Similarly, osteocalcin and PTH/PTHrP receptor gene expression are not compromised in bone of Hyp mice. Our results are consistent with the hypothesis that loss of Phex function affects the mineralizing activity of osteoblasts rather than their differentiation.

Osteoprotegerin and its ligand: A new paradigm for regulation of osteoclastogenesis and bone resorption

In just 3 years, striking new advances have been made in understanding the molecular mechanisms that govern the crosstalk between osteoblasts/stromal cells and hemopoietic osteoclast precursor cells that leads to osteoclastogenesis. Led first by the discovery of osteoprotegerin (OPG), a naturally occurring protein with potent osteoclastogenesis inhibitory activity, rapid progress was made to the isolation of RANKL, a transmembrane ligand expressed on osteoblasts/stromal cells, that binds to RANK, a transmembrane receptor on hemopoietic osteoclast precursor cells. The interaction of RANK and RANKL initiates a signaling and gene expression cascade that results in differentiation and maturation of osteoclast precursor cells to active osteoclasts capable of resorbing bone. Osteoprotegerin acts as a decoy receptor; it binds to RANKL and blocks its interaction with RANK, thus inhibiting osteoclast development. Many of the calciotropic hormones and cytokines, including vitamin D3, parathyroid hormone, prostaglandin E2 and interleukin-11, appear to stimulate osteoclastogenesis through the dual action of inhibiting production of OPG and stimulating production of RANKL. Estrogen, on the other hand, appears to inhibit production of RANKL and RANKL-stimulated osteoclastogenesis. Recently, the results of the first clinical trial with OPG supported its potential as a therapeutic agent for osteoporosis. The new understanding provided by the RANK/RANKL/OPG paradigm for both differentiation and activation of osteoclasts has had tremendous impact on the field of bone biology and has opened new avenues for development of possible treatments of diseases characterized by excessive bone resorption.

Osteoprotegerin and its ligand: a new paradigm for regulation of osteoclastogenesis and bone resorption

In just 3 years, striking new advances have been made in understanding the molecular mechanisms that govern the crosstalk between osteoblasts/stromal cells and hematopoietic osteoclast precursor cells that leads to osteoclastogenesis. Led first by the discovery of osteoprotegerin (OPG), a naturally occurring protein with potent osteoclastogenesis inhibitory activity, rapid progress was made to the isolation of RANKL, a transmembrane ligand expressed on osteoblasts/stromal cells that binds to RANK, a transmembrane receptor on hematopoietic osteoclast precursor cells. The interaction of RANK and RANKL initiates a signaling and gene expression cascade that results in differentiation and maturation of osteoclast precursor cells to active osteoclasts capable of resorbing bone. OPG acts as a decoy receptor, binding to RANKL and blocking its interaction with RANK, inhibiting osteoclast development. Many of the calciotropic hormones and cytokines, including 1,25(OH)2D3, PTH, PGE2 and IL-11, appear to act through a dual capacity to inhibit production of OPG and stimulate production of RANKL. Estrogen, on the other hand, appears to inhibit production of RANKL and RANKL-stimulated osteoclastogenesis. Recently, the results of the first clinical trial with OPG supported its potential as a therapeutic agent for diseases such as osteoporosis. The new understanding provided by the RANK/RANKL/OPG paradigm for both differentiation of osteoclasts and their activation has had tremendous impact on the field and opened new avenues for development of possible treatments of diseases characterized by excessive bone resorption.

Kinetics of osteoprogenitor proliferation and osteoblast differentiation in vitro

Fetal rat calvaria cells plated at very low density generate discrete colonies, some of which are bone colonies (nodules) from individual osteoprogenitors that divide and differentiate. We have analyzed the relationship between cell proliferation and acquisition of tissue-specific differentiation markers in bone colonies followed individually from the original single cell to the fully mineralized state. The size distribution of fully formed nodules is unimodal, suggesting that the coupling between proliferation and differentiation of osteoprogenitor cells is governed by a stochastic element, but distributed around an optimum, corresponding to the peak colony size/division potential. Kinetic analysis of colony growth showed that osteoprogenitors undergo 9-10 population doublings before the appearance of the first morphologically differentiated osteoblasts in the developing colony. Double immunolabeling showed that these proliferating cells express a gradient of bone markers, from proliferative alkaline phosphatase-negative cells at the periphery of colonies, to postmitotic, osteocalcin-producing osteoblasts at the centers. An inverse relationship exists between cell division and expression of osteocalcin, the latter being restricted to late-stage, BrdU-negative osteoblasts, while the expression of all other markers is acquired before the cessation of proliferation, but not concomitantly. Bone sialoprotein expression is biphasic, detectable in some of the early, alkaline phosphatase-negative cells, and again later in both late preosteoblast (BrdU-positive) and osteoblast (BrdU-negative, osteocalcin-positive) cells. In late-stage, heavily mineralized nodules, staining for osteocalcin and bone sialoprotein is not detectable in the oldest/most mature cells. Our observations support the view that the bone nodule "tissue-like" structure, originating from a single osteoprogenitor and finally encompassing mineralized matrix production, recapitulates successive stages of the osteoblast differentiation pathway, in a proliferation/maturation sequence. Understanding the complexity of the proliferation/differentiation kinetics that occurs within bone nodules will aid in the qualitative and/or quantitative interpretation of tissue-specific marker expression during osteoblastic differentiation.

Advances in the osteoblast lineage

Osteoblasts are the skeletal cells responsible for synthesis, deposition and mineralization of the extracellular matrix of bone. By mechanisms that are only beginning to be understood, stem and primitive osteoprogenitors and related mesenchymal precursors arise in the embryo and at least some appear to persist in the adult organism, where they contribute to replacement of osteoblasts in bone turnover and in fracture healing. In this review, we describe the morphological, molecular, and biochemical criteria by which osteoblasts are defined and cell culture approaches that have helped to clarify transitional stages in osteoblast differentiation. Current understanding of differential expression of osteoblast-associated genes during osteoprogenitor proliferation and differentiation to mature matrix synthesizing osteoblasts is summarized. Evidence is provided to support the hypothesis that the mature osteoblast phenotype is heterogeneous with subpopulations of osteoblasts expressing only subsets of the known osteoblast markers. Throughout this paper, outstanding uncertainties and areas for future investigation are also identified.

Aluminum accelerates osteoblastic differentiation but is cytotoxic in long-term rat calvaria cell cultures

We have examined the effects of aluminum (Al) on osteoprogenitor proliferation and differentiation, cell survival, and bone formation in long-term rat calvaria (RC) cell cultures. RC cells were grown in alpha minimal essential medium containing 10% fetal bovine serum, 50 microg/ml ascorbic acid, and 10 mM beta-glycerophosphate with or without Al added to final concentrations of 1 microM-1 mM. Al caused a dose-dependent increase in the number of bone nodules present at early times (day 11) but had no significant effect on nodule numbers at later times (day 17). Time course experiments showed that Al increased nodule number beginning from day 7. Alkaline phosphatase activity, assessed at four stages during the differentiation sequence of RC cell cultures (from 4 to 13 days) was stimulated by Al at all times. However, Al decreased colony formation, inhibited cell growth in late log phase, and decreased saturation density of the treated cultures. Al concentrations of 30 microM and above resulted in degeneration of the cell layer and an increasing fibrillar appearance of the matrix present in between or adjacent to nodules when cultures were maintained for more than 15 days. The presence of Al significantly decreased the viability of cells obtained from 13-17 days cultures, as determined by plating efficiency and trypan blue exclusion. We frequently observed cellular toxicity (in 8 of 10 experiments) in cultures containing 300 microM Al, and by days 17-19, cells, nodules, and matrix were disintegrating in these cultures. We conclude that Al accelerates the rate of osteoprogenitor cell differentiation and the formation of bone nodules while concomitantly inhibiting nodule mineralization. However, concentrations that accelerate differentiation appear to be cytotoxic in long-term cultures.

Cloning of a 2.5 kb murine bone sialoprotein promoter fragment and functional analysis of putative Osf2 binding sites

Bone sialoprotein (BSP) is an extracellular matrix protein that is intimately associated with the process of biomineralization. Osf2, a member of the Cbf/runt family of transcription factors, is required for the development of osteoblasts in vivo and has been reported to stimulate the transcription of BSP when overexpressed in mesenchymal cell lines. To investigate the role of Osf2 in BSP expression, we cloned a 2.5 kb fragment of a 5' untranscribed sequence from the murine BSP gene and evaluated it for putative Osf2 binding sites. This promoter, which was able to direct 5- to 10-fold higher levels of luciferase reporter expression in osteoblastic cells than in nonbone cell lines, contains two consensus core binding sites for members of the Cbf/runt family. One, at -61 relative to the start of transcription, is within a region having 75% overall sequence identity with the rat and human BSP promoters. The other is located at -1335, outside this highly conserved region. Neither site is completely conserved in the rat or human sequences. Only the -1335 site was able to bind a protein in nuclear extracts of osteoblastic cells, and this protein was identified as Osf2. Despite this in vitro binding ability, we detected no significant enhancer activity in the -1335 element when placed in front of a minimal osteocalcin promoter driving a luciferase reporter gene in osteoblastic cells nor any loss in transcriptional activity of a 5' promoter deletion which eliminated this element as compared with the full-length 2.5 kb promoter. These results suggest that Osf2 binding to the BSP promoter is not essential for its osteoblast-selective expression.

Osteoprogenitor cell frequency in rat bone marrow stromal populations: role for heterotypic cell-cell interactions in osteoblast differentiation

Glucocorticoids, notably dexamethasone (Dex), have been reported to be a requirement for osteoprogenitor cell differentiation in young adult rat bone marrow stromal cell populations. We have reinvestigated the requirement for Dex and analyzed the frequency of osteoprogenitor cells present. Stromal cells were grown as primary or first subcultures in the presence or absence of Dex and their expression of osteogenic markers (alkaline phosphatase activity, hormone responsiveness, and matrix molecules, including type I collagen, osteopontin, bone sialoprotein, and osteocalcin), as well as their functional capacity to differentiate to form a mineralized bone nodule, were assessed. Dex increased, but was not an absolute requirement for, the expression of osteogenic markers. Bone nodule formation was plating cell density dependent and occurred under all combinations of treatment with or without Dex but was maximal when Dex was present in both the primary and secondary cultures. Dex increased CFU-F by approximately 2-fold, but increased CFU-O (osteoprogenitor cells; bone nodule forming cells) by 5- to 50-fold depending on the cell density and duration of treatment. Neither CFU-F nor CFU-O expression followed a linear relationship in limiting dilution analysis until very high cell densities were reached, suggesting cooperativity of cell types within the population and a multitarget phenomenon leading to osteoprogenitor differentiation. When a large number of nonadherent bone marrow cells or their conditioned medium was added to the stromal cells, osteoprogenitors comprised approximately 1/100 of plated adherent cells and their expression followed a linear, single-hit relationship. By contrast, rat skin fibroblasts or their conditioned medium totally inhibited bone nodule formation. These data support the hypothesis that in marrow stroma, as in other bone cell populations such as those from calvaria, there are at least two classes of osteoprogenitor cells: those differentiating in the absence of added glucocorticoid and those requiring glucocorticoid to differentiate, that more than one cell type is limiting for stromal osteoprogenitor differentiation suggesting a role for heterotypic cell-cell interactions in osteogenesis in this tissue, and that Dex may be acting directly and/or indirectly through accessory cells in the bone marrow to alter osteoprogenitor cell expression.

cDNA fingerprinting of osteoprogenitor cells to isolate differentiation stage-specific genes

A cDNA fingerprinting strategy was developed to identify genes based on their differential expression pattern during osteoblast development. Preliminary biological and molecular staging of cDNA pools prepared by global amplification PCR allowed discrim-inating choices to be made in selection of expressed sequence tags (ESTs) to be isolated. Sequencing of selected ESTs confirmed that both known and novel genes can be isolated from any developmental stage of interest, e.g. from primitive progenitors, intermediate precursors or mature osteoblasts. EST expression provides insight into possible interrelated physiological functions and putative interacting molecules during differentiation. This method offers a functional genomics approach to isolate differentiation stage-specific genes in samples as small as a single cell.

Bone stem cells

Osteoblasts are the skeletal cells responsible for synthesis, deposition, and mineralization of the extracellular matrix of bone. By mechanisms that are only beginning to be understood, stem and primitive osteoprogenitors and related mesenchymal precursors arise in the embryo and at least some appear to persist in the adult organism, where they contribute to replacement of osteoblasts in bone turnover and in fracture healing. In this paper, the nature of these cells, whether they constitute a stem cell pool or a committed progenitor pool, and aspects of their apparent plasticity are discussed. Current understanding of differential expression of osteoblast-associated genes during osteoprogenitor proliferation and differentiation to mature matrix synthesizing osteoblasts is summarized. Finally, evidence is discussed that supports the hypothesis that the mature osteoblast phenotype is heterogeneous with subpopulations of osteoblasts expressing only subsets of the known osteoblast markers, raising also the possibility of multiple parallel differentiation pathways and perhaps even different progenitor pools.

LIF, but not IL-6, regulates osteoprogenitor differentiation in rat calvaria cell cultures: modulation by dexamethasone

Cytokines of the interleukin 6 (IL-6) subfamily are a group of factors produced by osteoblasts and acting through the same transducing element, membrane protein gp130. We have previously shown that exogenous (added to the culture medium) leukemia inhibitory factor (LIF) inhibits bone nodule formation and expression of osteoblast-associated genes in fetal rat calvaria (RC) cell cultures and that dexamethasone (Dex) increases the ID50 of LIF. To investigate the respective roles of IL-6-related cytokines and receptors in osteprogenitor differentiation, and their regulatory interplay with Dex, we used reverse transcribed polymerase chain reaction, bioassay, and blocking antibody techniques to assess the time courses of LIF, IL-6, LIF transmembrane receptor, IL-6 receptor, and gp130 expression in RC cell cultures grown with and without Dex. The levels of the mRNAs for IL-6, LIF, and gp130 decreased concomitantly with the formation of bone nodules. Dex treatment, which stimulates bone nodule formation, reduced the expression of LIF and IL-6 mRNAs and IL-6 bioactivity in the culture medium. LIF treatment strongly stimulated the expression of IL-6. Incubation with anti-LIF antibodies increased the number of nodules, while an antibody blocking IL-6 activity had little or no effect on nodule numbers and did not antagonize the action of exogenous LIF, indicating that IL-6 does not mediate the action of LIF in this system. Moreover, although exogenously added IL-6 was active in the cultures as noted by a reduction of nodule mineralization, it had no effect on nodule numbers, i.e., on osteoprogenitor differentiation, in the presence or absence of Dex. In conclusion, IL-6, LIF, and their receptors are expressed throughout the time-course of osteogenesis in RC cell cultures. However, only LIF, but not IL-6, appears to play a significant role in autocrine regulation of osteoblastic differentiation in this system. The antagonist action of Dex on the effects of exogenously added LIF, as well as the bone-promoting action of Dex in RC cell cultures, could be exerted partly through the down-regulation of the expression of endogenous LIF.

Cell sorting enriches osteogenic populations in rat bone marrow stromal cell cultures

The presence of multiple cell types in bone marrow stromal populations complicates interpretation of cytokine and hormone effects on the osteoprogenitors present, indicating a need for a method for purification of the osteoprogenitor population. Flow cytometric sorting of 7 day primary rat bone marrow stromal cell cultures was performed on the basis of alkaline phosphatase (AP) expression with an antibody against AP (RBM 211.13). The resultant AP(high), AP(low), or control cells were plated to determine osteoprogenitor, macrophage, and adipocyte distribution and frequency. Approximately 50% of osteoprogenitor/bone nodule-forming cells were lost during processing/sorting when compared with unsorted controls. Nevertheless, within the AP(high) fraction, the numbers of AP-positive colonies and osteoprogenitors (bone nodules) were significantly enriched compared with the unfractionated control; the increase in osteoprogenitor frequency ranged from approximately 2 to 100-fold. There were few assayable osteoprogenitors in either the AP(high) or AP(low) fractions in the absence of dexamethasone (dex), suggesting that RBM stroma contains largely dex-dependent osteoprogenitor populations, and that dex may regulate osteoprogenitors subsequent to the upregulation of AP. Osteoprogenitor/bone nodule numbers in either the AP(high) or AP(low) fraction did not follow a linear relationship with decreasing plating density. The AP(high) fraction of cells was depleted for adipocyte and macrophage colonies. In contrast, within the AP(low) fraction of cells, adipocyte and macrophage colonies were consistently enriched. We conclude that flow-cytometric sorting of RBM stromal populations according to high or low AP expression is an effective technique for enrichment of AP-positive colonies and osteoprogenitors/bone nodule-forming cells.

Osteogenic protein-1 up-regulation of the collagen X promoter activity is mediated by a MEF-2-like sequence and requires an adjacent AP-1 sequence

Bone morphogenetic proteins induce chondrogenesis and osteogenesis in vivo. To investigate molecular mechanisms involved in chondrocyte induction, we examined the effect of osteogenic protein (OP)-1/bone morphogenetic protein-7 on the collagen X promoter. In rat calvaria-derived chondrogenic C5.18 cells, OP-1 up-regulates collagen X mRNA levels and its promoter activity in a cell type- specific manner. Deletion analysis localizes the OP-1 response region to 33 bp (-310/-278), which confers OP-1 responsiveness to both the minimal homologous and heterologous Rous sarcoma virus promoter. Transforming growth factor-beta2 or activin, which up-regulates the expression of a transforming growth factor-beta-inducible p3TP-Lux construct, has little effect on collagen X mRNA and on this 33-bp region. Mutational analysis shows that both an AP-1 like sequence (-294/-285, TGAATCATCA) and an A/T-rich myocyte enhancer factor (MEF)-2 like sequence (-310/-298, TTAAAAATAAAAA) in the 33-bp region are necessary for the OP-1 effect. Gel shift assays show interaction of distinct nuclear proteins from C5.18 cells with the AP-1-like and the MEF-2-like sequences. OP-1 rapidly induces nuclear protein interaction with the MEF-2-like sequence but not with the AP-1 like sequence. MEF-2-like binding activity induced by OP-1 is distinct from the MEF-2 family proteins present in C2C12 myoblasts, in which OP-1 does not induce collagen X mRNA or up-regulate its promoter activity. In conclusion, we identified a specific response region for OP-1 in the mouse collagen X promoter. Mutational and gel shift analyses suggest that OP-1 induces nuclear protein interaction with an A/T-rich MEF-2 like sequence, distinct from the MEF-2 present in myoblasts, and up-regulates collagen X promoter activity, which also requires an AP-1 like sequence.

Characterization of a monoclonal antibody recognizing mast cells

Immunohistochemical screening for monoclonal antibodies prepared by immunization of mice with a rat osteoblastic cell population led to identification of one antibody that reacted against a small population of cells present in the soft connective tissue compartment of 21 days fetal rat calvaria. The morphology of the cells and the immunohistochemical staining characteristics (a distinct intracellular granular pattern) suggested that the antibody might be reacting specifically against mast cells. We used combined histochemistry and immunohistochemistry to further characterize this antibody, designated RCJ102. Cryosections containing calvaria bone, soft connective tissues and skin were prepared from the top of the head of 21 days fetal rats, and from adult rats cryosections of lung, muscle, adipose tissue and small intestine were prepared. Some sections were labelled by indirect immunofluorescence with RCJ102; corresponding sections were labelled histochemically with toluidine blue. There was a direct correspondence between mast cells identified histochemically and cells labelling with RCJ102 in all tissues except intestine, in which the mast cell detectable by histochemistry were not labelled by RCJ102. These results suggest that the RCJ102 antibody will be a valuable new reagent for further elucidation of the heterogeneity described between connective tissue and intestinal mucosal mast cells.

The ERR-1 orphan receptor is a transcriptional activator expressed during bone development

We studied the expression of estrogen-related receptor ERR-1 during mouse embryonic development. ERR-1 mRNA is present in bones formed by both the endochondral and intramembranous routes, and the onset of its expression coincides with bone formation. By RT-PCR experiments, we found that ERR-1, but not the related receptor ERR-2, is expressed in osteoblastic osteosarcoma cell lines as well as in primary osteoblastic cell populations derived from normal human bone. By gel shift analysis we found that ERR-1 binds as a monomer specifically to the SFRE sequence (SF-1-responsive-element; TCAAGGTCA). Mutation analysis revealed that both the core AGGTCA motif and the TCA 5'-extension are required for efficient ERR-1 binding. In transient transfection assays, ERR-1 acts as a potent transactivator through the SFRE sequence. This effect is cell-specific since ERR-1 activates transcription in the rat osteosarcoma cell line ROS 17.2/8 as well as in HeLa, NB-E, and FREJ4 cells but not in COS1 and HepG2 cells. Notably, the osteopontin (a protein expressed by osteoblasts and released in the bone matrix) gene promoter is a target for ERR-1 transcriptional regulation. Our findings suggest a role for ERR-1 in bone development and metabolism.

The mature osteoblast phenotype is characterized by extensive plasticity

While both morphological and biochemical-molecular attributes demarcate differentiation stages in specific cell and tissue types, what constitutes necessary and sufficient expression to define particular cell types is not always known. For example, mature osteoblasts (OBs) are defined morphologically as the cuboidal, biosynthetically active, basophilic cells residing on bone surfaces and responsible for the deposition of osteoid matrix. However, several recent observations suggest that not all mature OBs are identical. To explore further the validity of the hypothesis that heterogeneity of phenotype exists among mature OBs, we grew fetal rat calvaria cells in vitro at low density under conditions in which bone nodules form and mineralize in isolation of other contaminating cell and colony types. Cells resident in mature OB colonies, i.e., those comprising mainly cuboidal cells associated with an osteoid matrix that had begun to mineralize, were analyzed in situ for protein expression by immunocytochemistry with antibodies against collagen type I, alkaline phosphatase, osteopontin, bone sialoprotein, and osteocalcin. Consistent with the expected phenotype of mature OBs, many OBs expressed high levels of all of these markers, but strikingly even adjacent morphologically indistinguishable cuboidal OBs had differences in protein expression, especially in relation to osteopontin, bone sialoprotein, and osteocalcin expression. Double-labeling with Hoechst 33258 and osteocalcin indicated that the variation in antibody labeling intensity/protein expression appeared independent of a variation in cell cycle. To further ascertain the extent of this heterogeneity, 20 single cells were micromanipulated from colonies and subjected to poly(A)-PCR to analyze the simultaneous coexpression profiles of the same five markers analyzed by immunocytochemistry and two other markers, the OB-osteocyte transition marker E11 and the parathyroid hormone/parathyroid hormone-related protein receptor. Notably, the repertoire of genes expressed and their levels of expression varied markedly in individual OBs. The observed heterogeneity suggests that the mature OB phenotype is not a single unique phenotype but rather encompasses a flexible pattern of expression from the repertoire of OB-associated markers.

Expression and regulation of galectin 3 in rat osteoblastic cells

Galectin 3 is an endogenous soluble beta-galactoside-specific lectin originally identified and termed epsilon BP or IgE-binding protein in rat basophilic leukemia cells, but its wide tissue distribution and the multiple contexts in which it has been isolated have suggested that its function may not be limited to IgE binding but may include a role in cell growth regulation and differentiation, neoplastic transformation, and cell adhesion (Liu, 1990, Crit. Rev. Immunol., 10:289-306; Barondes et al., 1994, J. Biol. Chem., 269:20807-20810). After immunoscreening of a lambda gt11 cDNA expression library made from bone-nodule forming cultures of fetal rat calvaria (RC) cells with an antibody raised against osteoblastic cells (Turksen et al., 1992, J. Histochem. Cytochem., 40:1339-1352), three cDNA clones were isolated and sequenced; the sequence matched that of rat galectin 3. Galectin 3 mRNA was detected in various fetal and adult rat tissues, including calvaria and cultured RC cells. In RC cells and the rat osteosarcoma cell line ROS 17/2.8, galectin 3 mRNA expression increased with time in culture, in contrast to its behavior in fetal rat skin fibroblasts (RSF) in which its expression decreased with time in culture. In a second rat osteosarcoma line, UMR 106.01, galectin 3 mRNA was almost nondetectable. The synthetic glucocorticoid dexamethasone (Dex) enhanced galectin 3 expression in RSF cell cultures, while 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) had no significant effect. In contrast, Dex downregulated and 1,25(OH)2D3 upregulated galectin 3 expression in RC and ROS 17/2.8 cells, especially at later time points in culture when expression of osteoblast-associated differentiation markers by these cell types is most marked. Immunolabeling with an antibody against rat galectin 3 to identify galectin 3 protein showed that cells labelled within both the ROS 17/2.8 and RC populations but with marked intercellular heterogeneity of intensity. Our data support the conclusion that galectin 3 is a previously unrecognized product of osteoblastic cells, that galectin 3 mRNA and protein expression increases with time in vitro concomitant with other markers of osteogenesis, including formation of bone nodules and expression of osteoblast-associated markers such as alkaline phosphatase, bone sialo-protein, and osteocalcin, and that its expression is regulated by hormones such as glucocorticoids and 1,25(OH)2D3 that modulate other aspects of the osteoblast phenotype.

Analysis of chondroprogenitor frequency and cartilage differentiation in a novel family of clonal chondrogenic rat cell lines

We have isolated through sequential steps of subcloning a series of normal clonal cell lines enriched for chondroprogenitors that undergo differentiation in vitro from progenitors to mature chondroblasts and chondrocytes forming three-dimensional cartilage nodules. In the parental chondroblast clone RCJ 3.1C5 (C5), differentiation and cartilage formation occurred without added hormones or growth factors, but chondrogenesis could be stimulated markedly in the presence of the glucocorticoid steroid Dexamethasone (Dex). Limiting dilution analysis indicated that greater than one in ten C5 cells plated was a chondroprogenitor capable of differentiating and forming a cartilage nodule in low density cultures, but chondrogenesis was down-regulated in higher density cultures. Dex elicited a greater stimulatory effect on cartilage nodule formation when C5 cells were plated at higher rather than lower densities. Since Dex also maintained the chondrogenic potential of C5 cells passaged repeatedly, we subcloned C5 in the presence of Dex. Eight of eleven subclones were chondrogenic and the frequency of chondroprogenitors capable of cartilage formation in isolated subclones ranged from lower to much higher than in the parental C5 clone. Both Dex-independent as well as Dex-dependent clones were identified, although long-term maintenance of the chondrocyte phenotype in all subclones required Dex. These data suggest that there are Dex-dependent and Dex-independent chondroprogenitor cells, that cell-cell interactions and/or local factors can modulate cartilage nodule formation and that Dex-responsive steps are involved in long-term maintenance of chondroprogenitors in vitro. Thus, this unique family of non-transformed, clonal chondrogenic cell lines provides a quantifiable, readily manipulatable system in which cartilage differentiation and metabolism can be assessed.

CD44 expression in fetal rat bone: in vivo and in vitro analysis

Knowledge of the expression pattern of the cell surface glycoprotein CD44 in the development and differentiation of bone is limited. We investigated CD44 expression (a) in bone sections of 21-day-old fetal rat calvaria (RC), metatarsals, and tibiae, (b) in primary cultures of RC cells undergoing differentiation in vitro, and (c) in three rat osteosarcoma cell lines: ROS 17/2.8, UMR 106.01, and UMR 106.06. By immunocytochemistry, Western, Northern and reverse transcription polymerase chain reaction analyses, we found that osteoblastic cells express the 'hematopoietic' or 'standard' CD44 (CD44s) isoform. Osteoblastic cells in vivo and in vitro stained at all detectable stages of differentiation, but intercellular heterogeneity of CD44s staining was evident, with lesser staining in preosteoblastic cells and greater staining in mature osteoblasts and osteocytes. As cells in RC cultures differentiated and formed bone in vitro, CD44s mRNA and protein levels as measured on immunoblots were invariant. All three osteosarcoma cell lines expressed CD44s mRNA and protein. The synthetic glucocorticoid dexamethasone, which stimulates osteogenesis in RC cells in vitro and regulates a number of osteoblast-associated genes, had no apparent effect on either CD44s protein or mRNA levels. The widespread presence of CD44s in osteoblastic cells at various maturational stages suggests that further analyses will be required to determine what role CD44s may play in osteogenesis and in bone tissue organization.

Monoclonal antibodies as tools for studying the osteoblast lineage

Knowledge of the number and kinds of differentiation steps characterizing cells of the osteoblast lineage is inadequate. To analyze further osteoblast differentiation, a number of labs have generated monoclonal antibodies to osteogenic cells, derived from both normal bone and osteosarcomas. A variety of immunolabelling patterns on primary cell cultures, cell lines, and tissue sections has been reported, including cell surface, cytoplasmic, and extracellular matrix-associated patterns. Most of the antibodies selected recognize predominantly the mature osteoblast and osteocyte; in addition, however, antibodies have been generated that recognize pre-osteoblasts. Some recognize cells of both the osteoblast and chondroblast lineages and may contribute to a better understanding of the lineage and phenotypic relationships between these two cell types. In addition to recognition in vivo of cell subpopulations of discrete maturational stages, changes in the immunolabelling patterns in vitro have also documented a differentiation sequence in cells undergoing osteogenesis in cell and tissue cultures. In at least two cases, the antibodies have been used to isolate subpopulations of cells from bone, including relatively pure populations of osteocytes. With the exception of several antibodies that are against alkaline phosphatase or known matrix proteins including osteocalcin, the nature of the macromolecular species recognized by most of the antibodies generated to date are unknown. Recently, however, one antibody was used to clone the cDNA for the beta-galactoside-binding lectin, galectin 3 or epsilon binding protein (epsilon BP; IgE-binding protein; Mac-2), from a lambda gt11 osteoblast expression library; another was used to clone from an ROS 17/2.8-COS cell expression library the cDNA for OTS-8, a putative target gene of early response genes stimulated in response to phorbol esters in MC3T3-E1 cells. Neither of these macromolecules had previously been identified in bone cells, but the recent molecular and cellular analyses have shown them to be developmentally and/or hormonally regulated in osteoblastic cells. These antibodies extend the available markers and support earlier observations that a variety of molecules are differentially expressed by cells at different stages of the osteoblast lineage. This chapter will not be an exhaustive survey of all immunocytochemical and immunohistochemical analyses of osteogenic cells and tissues but will focus on the approach of eliciting novel monoclonal antibodies by the injection of osteogenic cells or crude bone extracts and its potential for establishing new markers of the osteoblast lineage. We have not included a large number of studies documenting the use of antibodies raised against several known bone matrix proteins; while these have been crucial in developing our current understanding of osteogenic differentiation, we sought rather to highlight the potential of the "random" injection approach.

Aluminum inhibits both initiation and progression of mineralization of osteoid nodules formed in differentiating rat calvaria cell cultures

Osteoid nodules form in cultures of fetal rat calvaria (RC) cells grown in medium containing 10% fetal bovine serum (FBS) and 50 microns/ml of ascorbic acid. When 10 mM beta-glycerophosphate (beta-GP) is added, the nodules mineralize in two phases: an initiation phase that is dependent upon alkaline phosphatase activity for cleavage of beta-GP to inorganic phosphate (P(i)) and a progression phase that proceeds independently of the activity of alkaline phosphatase and does not require exogenous phosphate. We have used this system to investigate the effects of aluminum (Al3+)on mineralization. When AlCl3 was added to culture medium at concentrations of 0, 3, 10, 30, 100, and 300 muM, the total concentrations of aluminum were 0.98, 6.07, 16.82, 40.19, 88.45, and 284.52 muM, respectively. The corresponding free Al3+ concentrations, assessed after ultrafiltration, were found to be 1.11, 1.75, 3.40, 6.22, 5.38, and 12.11 muM. In cultures in which osteoid was formed and mineralization initiated in the presence of added Al+ (3-300 muM), a dose-dependent inhibition of mineralization occurred. Osteoid formed in the presence of added Al3+ mineralized normally when Al3+ was removed from cultures at the time of initiation of mineralization with beta-GP. In osteoid nodules grown in the absence of Al3+, addition of Al3+ (3-300 muM) at the start of the initiation phase of mineralization resulted in a dose-dependent inhibition of mineralization. Addition of Al3+ to cultures after mineralization had been initiated in the absence of Al3+ inhibited progression of mineralization at added Al3+ concentrations of 10 muM and above. Al3+ did not decrease the conversion of beta-GP to P(i) and caused a small but significant increase in alkaline phosphatase activity at added concentrations of 100 muM or greater. The data show that Al3+ inhibits both the initiation and progression phases of mineralization starting at added concentrations of 3-10 muM (approximately 1.7-3.4 muM free Al3+) and that mineralization of osteoid formed in the presence of Al3+ is unaffected if Al3+ is removed prior to the initiation of mineralization.

Osteoblast and chondroblast differentiation

Recognition of discrete commitment and differentiation stages requires characterization of changes in proliferative capacity together with the temporal acquisition or loss of expression of molecular and morphological traits. Both cell lines and primary cultures have been useful for analysis of transitional steps in the chondroblast (CB) and osteoblast (OB) lineages. One striking feature is that OBs and CBs share expression of some molecules, including newer markers such as epsilon BP (galectin-3), while also having unique markers. The fact that hypertrophic chondrocytes appear able to downregulate cartilage markers and upregulate OB markers also points to an interesting lineage relationship that needs to be explored further. Recently, we have focused on the osteoprogenitors that divide and differentiate into mature OBs forming bone nodules in fetal rat calvaria cell cultures. We use cellular, immunocytochemical, and molecular approaches, including PCR on small numbers of cells, to discriminate stages. Nodule formation is characterized by loss of proliferative capacity and sequential increased marker expression, that is, alkaline phosphatase (AP), followed by bone sialoprotein (BSP), and osteocalcin. Upregulation of collagen type I and biphasic expression of osteopontin, with two peaks corresponding to proliferation and differentiation stages, also occurs. A variety of other molecules are also upregulated in the mature OB, including epsilon BP and CD44s. By replica plating and PCR, we have begun to study the expression of the messenger RNAs (mRNAs) for potential regulatory molecules (e.g., PTHrP) and their receptors (e.g., PTHR, FGFR-1, and PDGFR alpha) and have found all to be modulated during the progression from committed osteoprogenitor to mature OB.(ABSTRACT TRUNCATED AT 250 WORDS)

Tri-iodothyronine (T3) and dexamethasone interact to modulate osteoprogenitor cell differentiation in fetal rat calvaria cell cultures

We investigated the role of 3,5,3'-tri-iodothyronine (T3) in regulating differentiation of osteoprogenitor cells and also studied the effects of the glucocorticoid hormone dexamethasone (Dex) on the T3-induced effects on osteoprogenitor populations. This was done by determining the effects of either hormone alone, or of combinations of the two hormones, on the number of bone nodules formed in long-term cultures of rat calvaria cells. In this system, Dex has been shown to increase bone nodule formation, the maximal effective dose being 10 nM (Bellows et al. Endocrinology 121: 1985-1992; 1987). In standard culture medium containing 15% fetal bovine serum FBS), low concentrations of T3 (0.001-0.1 nM) had no effect on the number of bone nodules, while higher concentrations of 1-100 nM inhibited. However, in culture medium containing 10 nM Dex, the lower concentrations of T3 markedly increased the number of nodules. Short term pulse experiments with these low concentrations of T3 in the presence of Dex indicated that stimulation of nodule formation occurred only when T3 was present prior to confluency. Higher concentrations of T3 (1-100 nM) decreased nodule number whether or not Dex was added. We then cultured cells in medium containing FBS from which T3 and T4 were removed by treatment with AG-1 chi-10 resin. In both + or - Dex conditions, bone nodule formulation was increased 1.5 to 2-fold in T3, T4-depleted medium when compared with cultures maintained in standard culture medium.(ABSTRACT TRUNCATED AT 250 WORDS)

Leukemia inhibitory factor inhibits osteogenic differentiation in rat calvaria cell cultures

Leukemia inhibitory factor (LIF) is a pleiotropic cytokine with both anabolic and catabolic effects on bone tissue. To investigate the effect of LIF on bone formation in the absence of a resorption cycle, we used fetal rat calvaria cell cultures and quantified bone nodule production, which provides a colony assay to analyze the effects of factors on osteoprogenitor differentiation and bone formation. In these cultures, dexamethasone (Dex) stimulates bone nodule formation. In dose-response experiments, LIF inhibited bone nodule formation by cells cultured with (+Dex; ID50 = 250 U/ml) or without (-Dex; ID50 = 30 U/ml) 10(-8) M Dex. Residual nodules were small and poorly mineralized. Continuous exposure to LIF (500 U/ml) up to day 25 did not affect either the growth rate or saturation density of the cultures, but decreased alkaline phosphatase activity and bone nodule production, with greater inhibition in -Dex cultures. Exposure to LIF (500 U/ml) for 3 days early during nodule formation (about day 10) reduced bone nodule numbers to the same extent as continuous treatment in -Dex cultures and significantly, but less markedly, in +Dex cultures; earlier and later pulses had no effect. Northern blot analysis of expression of messenger RNAs of bone related proteins in cultures pulsed (-Dex) at various stages of development showed marked inhibition of alkaline phosphatase, bone sialoprotein, and osteocalcin; slight inhibition of type I collagen; early stimulation of osteopontin; and no effect on Secreted Protein, Acidic and Rich in Cysteine/osteonectin. These results suggest that LIF is an inhibitor of bone nodule formation in these cultures, acting at a stage when late osteoprogenitors and/or early osteoblasts are present, and that Dex may modulate the effects of LIF by shifting effective doses to higher concentrations.

Bone matrix proteins in osteogenesis and remodelling in the neonatal rat mandible as studied by immunolocalization of osteopontin, bone sialoprotein, alpha 2HS-glycoprotein and alkaline phosphatase

The neonatal rat mandible was used as a model to study bone formation, mineralization, quiescence, and resorption, using immunolocalization and a variety of tissue-processing techniques. Monospecific antibodies for osteopontin (OPN), bone sialoprotein (BSP), alkaline phosphatase (AP) and alpha 2HS-glycoprotein (alpha 2HS-GP) were used on fixed paraffin-embedded tissue, fixed frozen tissue and unfixed frozen tissue. Immunostaining was correlated with mineral content by two procedures, the von Kossa and the morin techniques. Morin fluorescence was used with secondary immunostaining to provide a way of closely correlating bone matrix proteins and matrix mineralization. Co-immunolocalization procedures were used to compare the sites of bone proteins in the matrix. AP was found earliest during osteogenic cell differentiation, appearing in the preosteoblasts, followed by OPN and BSP, which first appeared in osteoblasts. alpha 2HS-GP expression was not observed in cells. The results provide clear evidence for the presence of OPN in osteoid, while BSP and alpha 2HS-GP were confined to the mineralized matrix. Immunostaining of bone proteins is highly technique-dependent: immunolocalization investigations required several methods of approach to ensure adequate demonstration of these proteins in cells and matrix. The results support the contention that osteopontin is multifunctional in bone metabolism, and that alpha 2HS-GP, though produced in the liver, is abundant in bone matrix and may also have a function in bone metabolism.

Dexamethasone alters the subpopulation make-up of rat bone marrow stromal cell cultures

Bone marrow stromal cells comprise a heterogeneous population including fibroblastic, adipocytic, hemopoietic, and osteogenic cells. Although the conditions under which different lineages are regulated have not been fully elucidated, dexamethasone clearly stimulates osteogenic expression in stromal cultures. The purpose of this study was to begin to elucidate and quantify some of the subpopulations present when rat bone marrow stromal cells are grown with or without dexamethasone under conditions favoring bone formation. Bone marrow stromal cells from young adult rats were cultured with ascorbic acid, beta-glycerophosphate, and with or without dexamethasone for various periods of time. Culture dishes were then analyzed for cell counts, or stained with either histochemical or immunohistochemical stains, and colony types were quantitated, or cells were processed for flow cytometry. Dexamethasone significantly increased the number of alkaline phosphatase (AP) positive colonies, von Kossa positive bone nodules, alpha-naphthylbutyrate esterase positive colonies, and ED2 positive (macrophage) colonies. The number of adipocytic foci was largely unaffected in these experiments. Flow cytometry confirmed colony counts and showed stimulation by dexamethasone of AP positive cells and macrophages, and in addition, the reduction of hemopoietic cells expressing leukocyte common antigen. These data show conclusively that when rat bone marrow stromal populations are grown under conditions stimulating osteoprogenitor differentiation and bone formation, the stromal subpopulation make-up, including expression of hemopoietic lineages, is markedly altered.

Simultaneous detection of multiple bone-related mRNAs and protein expression during osteoblast differentiation: polymerase chain reaction and immunocytochemical studies at the single cell level

Messenger RNA expression analyzed by in situ hybridization and Northern analysis and protein expression analyzed biochemically or immunocytochemically have been used to study the developmental expression of various osteoblast (OB)-associated molecules. These approaches have shown that over a time course of OB differentiation in vivo and in vitro, the expression of macromolecules associated with OB cells changes. However, ambiguities in data from different approaches and in populations representative of cells at different developmental stages are extant. To begin to discriminate differentiation stages with more precision and to address intercellular heterogeneity, fetal rat calvaria cells were grown at low densities under conditions in which bone nodules form and mineralize and colonies were classified morphologically as fibroblastic or osteoblastic (early, intermediate, or mature). Whole discrete colonies and single cells from individual colonies were analyzed molecularly by a random amplification poly(A)-polymerase chain reaction (PCR) and for protein expression by immunocytochemistry; we analyzed the expression of known bone-related macromolecules (collagen type I, alkaline phosphatase, osteopontin, bone sialoprotein, and osteocalcin). Both PCR and immunocytochemistry revealed that different colony types were reproducibly distinguishable in their expression of either general (collagen type I) or bone-associated (alkaline phosphatase, osteopontin, bone sialoprotein, and osteocalcin) macromolecules, such that fibroblastic colonies were distinguishable from osteoblastic colonies and the latter could be subdivided into less mature or more mature osteoblastic colonies. While some aspects of the temporal differentiation sequence defined earlier were confirmed, several additional features were evident from these single cell-single colony studies. First, different repertoires of OB-associated markers were expressed in different cells, suggesting variation in the switch-on of the OB differentiation program and heterogeneity in the OB phenotype. Second, among colonies classified as fibroblastic on the basis of morphology heterogeneity was also evident and there were some cells expressing features consistent with their being osteoprogenitor cells. Our data support the hypothesis that individual fibroblastic and osteoblastic cells are heterogeneous in expression of marker molecules. We also conclude that individual cells and colonies analyzed by poly(A)-PCR will be useful in lieu of mass populations to extend investigation of stages in the progression of OB differentiation.

Displacement and translocation of osteoblast-like cells by osteoclasts

Rabbit osteoclasts and rabbit osteoblast-like stroma cells (OB cells) were placed onto plastic surfaces and the migration patterns of individual osteoclasts and osteoclast-OB interactions were analyzed with time-lapse recording. To induce directed migration, the cultures were exposed to an electrical field of 0.01 or 0.1 V/mm. At 0.1 V/mm, osteoclasts moved directly toward the anode in some cases, clearing OB cells from their path of migration. In other cases, osteoclasts migrated toward the anode for part of the time but then changed direction and moved toward groups of OB cells. Observations were made on osteoclasts interacting with single OB cells or small colonies and on osteoclasts interacting with OB monolayers, at both field strengths; the results were independent of field strength. There were several characteristic behaviors. With single OB cells and small OB colonies, retraction of OB cells upon contact with the osteoclast was the predominant mechanism whereby these cells begin to move out of the path of the osteoclast. A pronounced ruffling or blebbing of the OB cell membrane often followed retraction. When osteoclasts displaced OB cells that were part of a monolayer, extension of an osteoclast lamellipodium underneath the edge of the OB cell layer generally preceded partial retraction of the OB cells involved. It sometimes appeared as if the detached or partially detached OB cells were "pushed" by the osteoclast, which in some cases resulted in OB cells being moved hundreds of microns in a period of a few hours, at rates comparable to the normal speed for osteoclast migration (congruent to 100 microns/h), much faster than the normal speed for OB cells (congruent to 10 microns/h).(ABSTRACT TRUNCATED AT 250 WORDS)

Carbonic anhydrase II mRNA expression in individual osteoclasts under "resorbing" and "nonresorbing" conditions

Rabbit osteoclasts can be transformed from a nonresorbing state to a resorbing state by transferring them from culture medium at pH 7.5 to one at pH 6.5. We evaluated whether expression of mRNA for carbonic anhydrase (CA-II) could be used as an indicator of the state of activity of individual osteoclasts. A cDNA probe to rabbit carbonic anhydrase II (CA-II) was prepared and used for in situ hybridization analysis of osteoclasts isolated from neonatal rabbit long bones. Quantitation by grain counting revealed heterogeneity within the osteoclast population: osteoclasts with a "compact" (rounded, less spread) morphology expressed higher levels of CA-II mRNA than "spread" osteoclasts with similar numbers of nuclei. When maintained at pH 6.5 for 6 h, the level of CA-II mRNA was increased significantly in osteoclasts of both morphologies compared with those in parallel cultures maintained at pH 7.5. These results were confirmed by quantitating CA-II mRNA using the polymerase chain reaction (PCR). Oligonucleotide primers specific for rabbit CA-II were synthesized and used to amplify CA-II cDNA transcribed from mRNA prepared from single or small numbers (one to eight cells) of osteoclasts that were collected with a micromanipulator. This generated a approximately 510 bp PCR product, corresponding to the predicted size of the CA-II fragment encompassed by the primers. For quantitation, CA-II mRNA levels were compared with the levels of a approximately 900 bp actin fragment that was coamplified in the same reaction mixture or amplified separately in a duplicate sample of the reaction mixture.(ABSTRACT TRUNCATED AT 250 WORDS)

1,25-dihydroxyvitamin D3 stimulates adipocyte differentiation in cultures of fetal rat calvaria cells: comparison with the effects of dexamethasone

Progenitor cells for several mesenchymally derived cell types exist within freshly isolated fetal rat calvaria (RC) cell populations. We have characterized the effects of 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3] on the differentiation of adipocytes from primary RC cells and compared these effects with those of dexamethasone (Dex). RC cells were plated at 3 x 10(4)/35-mm dish, and cultures were maintained for 14-19 days in alpha-Minimum Essential Medium containing 10% fetal bovine serum, 50 micrograms/ml ascorbic acid, 10 mM Na beta-glycerophosphate, and 0.1-100 nM 1,25-(OH)2D3 or 1-1000 nM Dex. Morphological (quantitation of adipocyte foci number and area after staining cultures with Sudan IV) and biochemical (glycerol-3-phosphate dehydrogenase activity) methods of assessing adipogenesis were used. In the presence of 1,25-(OH)2D3, adipocyte foci developed about 3 days after confluency as clusters of rounded or stellate cells. Stimulation of adipocyte foci development was dose dependent from 0.1-100 nM and was maximal with 10 nM 1,25-(OH)2D3; half-maximal stimulation occurred at about 1 nM. The presence of ascorbic acid and beta-glycerophosphate was not required for 1,25-(OH)2D3-induced stimulation of adipocytes, but both significantly increased the number of adipocyte foci in the presence of 1,25-(OH)2D3. The critical period for initiation of adipocyte differentiation with 1,25-(OH)2D3 was between 1-9 days, and once committed along the adipogenic pathway, adipocytes maintained their differentiated state in the absence of 1,25-(OH)2D3. Short term (48-h) pulses of 1,25-(OH)2D3 resulted in slight, but significant, increases in adipocyte formation. Other vitamin D3 metabolites were less effective than 1,25-(OH)2D3 in stimulating adipocyte differentiation. Dex (1-100 nM) also caused a dose-dependent increase in the differentiation of adipocyte foci in RC cell cultures. The adipocyte foci that developed in the presence of Dex frequently appeared earlier in culture, i.e. when cells reached confluency on days 6-7, and were more diffuse than those forming with 1,25-(OH)2D3. The stimulation of adipocyte differentiation by 1,25-(OH)2D3, however, was greater than that by Dex in mixed RC II-V cells. The combined effects of 1,25-(OH)2D3 and Dex were additive at low concentrations and synergistic at higher concentrations of either 1,25-(OH)2D3 or Dex. The data show that bone cell populations isolated from fetal RC contain adipocyte progenitors and that 1,25-(OH)2D3 as well as Dex are potent regulators of adipocyte differentiation within these bone cell populations.

Cellular expression of bone-related proteins during in vitro osteogenesis in rat bone marrow stromal cell cultures

Rat bone marrow stromal cells comprise a heterogeneous mixture of cell lineages including osteoblastic cells. When grown in the presence of ascorbic acid, beta-glycerophosphate and 10(-8) M dexamethasone, osteoprogenitor cells within the population divide and differentiate to form bone nodules (Maniatopoulos et al., 1988, Cell Tissue Res., 254:317-330; Aubin et al., 1990, J. Bone Miner. Res., 5:S81) providing a useful model to investigate temporal and spatial changes in expression of osteoblastic markers. Immunocytochemistry was combined with Northern blotting, enzymatic assay, and radioimmunoassay to analyze the expression of bone-related proteins during the growth and differentiation sequence. By mRNA levels, protein production and/or enzymatic activity, expression of osteocalcin, bone sialoprotein, and alkaline phosphatase increased concomitantly with the development of bone nodules, while osteopontin mRNA levels decreased and those of SPARC/osteonectin did not change significantly. In older cultures with mineralizing nodules, mRNA levels for alkaline phosphatase and bone sialoprotein, but not osteocalcin, declined. Immunolabeling revealed that cells in early cultures stained poorly for SPARC/osteonectin and strongly for thrombospondin. Later, SPARC/osteonectin staining increased in most cells, while thrombospondin staining could be seen in both matrix and in cells, but with marked intercellular variability in intensity. At all time points studied, osteoblasts within bone nodules stained homogeneously for thrombospondin and alkaline phosphatase, and with marked heterogeneity of intensity amongst cells for SPARC/osteonectin and osteocalcin. Labelling with RCC455.4, a monoclonal antibody raised against rat calvaria cells which intensely labels osteoblasts and osteocytes (Turksen et al., 1992, J. Histochem. Cytochem., 40:1339-1352), co-localized with osteocalcin. Alkaline phosphatase activity and the amount of osteocalcin determined by both radioimmunoassay and immunolabelling decreased in very late cultures, a time corresponding to appearance of fully mineralized nodules. These studies indicate that the bone marrow stromal cell system is a useful model to study the temporal and spatial expression of bone-related proteins during osteogenesis and formation, mineralization, and maturation of bone nodules. Further, immunolabelling at the individual cell and single bone nodule level allowed discrimination of marked variability of expression of osteoblast markers during the differentiation sequence.

Osteoclast differentiation in cocultures of a clonal chondrogenic cell line and mouse bone marrow cells

Previous reports have demonstrated that hemopoietic progenitor cells derived from mouse bone marrow can form osteoclast-like cells when cultured in the presence of stromal cells and 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3]. We show here that in cocultures of mouse bone marrow cells and a clonal chondrogenic cell line (C5.18), a stimulation of the number of tartrate-resistant acid phosphatase-positive (TRAP+) colonies is seen with or without the addition of 1,25-(OH)2D3 to the cultures. A large proportion of the TRAP+ cells had calcitonin receptors. In addition resorption lacunae were observed on bone slices on which cocultures were maintained, thus these cells had the characteristics of osteoclasts. The number of osteoclast-containing colonies that formed in cocultures varied with the plating density of the C5.18 cells and the length of time the C5.18 cells were cultured before adding mouse bone marrow. These results suggested that osteoclast differentiation decreased with increasing cartilage differentiation. C5.18 cells treated with 1,25-(OH)2D3 before coculture stimulated TRAP+ osteoclast colony formation to a greater extent than untreated C5.18 cells, whereas C5.18 cells cultured in the presence of dexamethasone before coculture inhibited TRAP+ osteoclast colony formation relative to untreated C5.18 cells. Since 1,25-(OH)2D3 inhibits and dexamethasone stimulates cartilage differentiation in C5.18 cells, these results agree with the view that chondroprogenitor cells stimulate osteoclast colony formation, whereas cultures containing predominantly mature chondrocytes do not. Osteoclast-containing colonies were frequently associated with colonies of alkaline phosphatase-positive (AP+) cells. This raised the possibility that C5.18 cells stimulated osteoclast differentiation indirectly by increasing the numbers of AP+ stromal cells from the marrow population, which in turn could stimulate osteoclast differentiation from marrow hemopoietic progenitors. In cocultures in which the C5.18 cells were physically separated from the marrow cells, we also observed increased numbers of TRAP+ colonies growing in association with large colonies of AP+ cells, suggesting that C5.18 cells release a soluble factor that mediates these effects.

Differential effects of fluoride during initiation and progression of mineralization of osteoid nodules formed in vitro

Osteoid nodules form in cultures of fetal rat calvarial (RC) cells grown in medium containing 10% FBS and 50 micrograms/ml of ascorbic acid. When 10 mM beta-glycerophosphate (beta-GP) is added, osteoid nodules mineralize in two phases: an initiation phase, which is dependent upon alkaline phosphatase activity for conversion of beta-GP to P(i), and a progression phase that proceeds independently of alkaline phosphatase activity and does not require exogenous phosphate. We have now used this system to investigate the effects of fluoride (F-) on mineralization. In cultures in which osteoid was formed and mineralization initiated in the presence of F-, a dose-dependent inhibition of the initiation of mineralization occurred over a concentration range of 25-500 microM F- (p < 0.001 in all cases). The initiation of mineralization was not inhibited if F- was removed from the cultures at the time when mineralization was initiated with beta-GP. In osteoid nodules grown in the absence of F-, addition of F- resulted in a dose-dependent inhibition of the initiation of mineralization, with significant decreases in 45Ca uptake occurring at F- concentrations of 3 microM (p < 0.01) and higher. However, if F- was added to cultures after mineralization was initiated in the absence of F-, a stimulation of 45Ca uptake was observed at F- concentrations of 250 microM and above (p < 0.001). F- (1-1000 microM) did not affect the conversion of beta-GP to P(i) or alkaline phosphatase activity in the cultures.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of the 1,25-(OH)2D3-induced inhibition of bone nodule formation in long-term cultures of fetal rat calvaria cells

We investigated the effects of 1,25-dihydroxyvitamin D3[1,25-(OH)2D3], on osteoprogenitor cell differentiation and bone nodule formation at various stages of differentiation by evaluating the effects on long term cultures of fetal rat calvaria (RC) cells. RC cells were plated at 3 x 10(4) cells/35-mm dish in alpha-minimal essential medium containing 15% fetal bovine serum, ascorbic acid, and beta-glycerophosphate (beta-GP), conditions under which bone nodules form. 1,25-(OH)2D3 inhibited bone nodule formation in a dose-dependent manner with total inhibition occurring at 1-10 nM and half-maximal inhibition occurring at approximately 0.06 nM. 1,25-(OH)2D3 also significantly stimulated RC cell growth in a dose-dependent manner in both the presence and absence of ascorbic acid. Addition of 1 nM 1,25-(OH)2D3 at different times after the start of culture inhibited nodule formation when added before and up to the early multilayering stage (up to day 11 of culture), but had no effect on nodule number when added later. When 1,25-(OH)2D3 was added at the start of the culture period and removed at the early multilayering stage, nodule formation was also inhibited. Pulses of 48-h duration also inhibited nodule formation, with maximal effect occurring between days 3 and 11. Thus, 1,25-(OH)2D3 inhibited osteoprogenitor cell differentiation during the earlier stages of culture before visible bone nodule formation occurred and the effect was not reversible upon removal of 1,25-(OH)2D3. In cultures grown to the multilayering stage in medium without ascorbic acid and beta-GP and then changed to medium with ascorbic acid and beta-GP, 1,25-(OH)2D3 inhibited when present before, but not after, the addition of ascorbic acid and beta-GP. Two other vitamin D3 metabolites, 24,25-dihydroxyvitamin D3 [24,25-(OH)2D3] and 1,24,25-trihydroxyvitamin D3 [1,24,25-(OH)3D3] had inhibitory effects similar to 1,25-(OH)2D3. The effects were dose dependent for each metabolite tested and correlated with the biological effectiveness of these metabolites in other systems: i.e. 1,25-(OH)2D3 was more effective than 1,24,25-(OH)3D3 which in turn was more effective than 24,25-(OH)2D3. The data show that 1,25-(OH)2D3 inhibits osteoprogenitor cell differentiation at an early stage and at a time during which cell growth is stimulated.

The role of phospholipase A2 in interleukin-1 alpha-mediated inhibition of mineralization of the osteoid formed by fetal rat calvaria cells in vitro

Interleukin-1 (IL-1) may be an important mediator of bone remodeling, since it is a potent stimulator of bone resorption and has biphasic effects on bone formation. Continuous exposure of fetal rat calvaria (RC) cells to IL-1 alpha or IL-1 beta results in a dose-dependent inhibition of both bone nodule formation and mineralization of the organic matrix. In this study, the effects of recombinant human IL-1 alpha on the mineralization process were examined by the addition of IL-1 alpha late in the culture period, after osteoid nodules had formed and when they were induced to mineralize by the addition of organic phosphate. By means of a quantitative 45calcium radiolabeling assay, it was shown that short-duration exposures of fully-formed bone nodules to IL-1 alpha also inhibited mineralization, and that the duration of treatment directly correlated with the degree of inhibition. Because our earlier studies had demonstrated that IL-1 stimulated the release of PLA2 and PGE2 from RC cells, the effects of PLA2 and of inhibition of PGE2 synthesis on mineralization were investigated. Exogenous Naja naja group I PLA2 had little effect on the mineralization of bone nodules; however, Crotalus adamanteus group II PLA2 inhibited mineralization at concentrations similar to those found in the media from IL-1 alpha-treated cultures. Although PLA2 is thought to stimulate PGE2 synthesis by releasing arachidonic acid from membrane phospholipids, PGE2 release by RC cells accounted for only part of the IL-1 alpha-mediated inhibition, suggesting the presence of other mechanisms of exogenous PLA2 action in inhibiting mineralization.(ABSTRACT TRUNCATED AT 250 WORDS)

Differential regulation of phospholipase A2 by cytokines inhibiting bone formation and mineralization

Treatment of fetal rat calvarial cells with interleukin-1 alpha, tumor necrosis factor-alpha, transforming growth factor-beta 1, or group II phospholipase A2 inhibits the number of bone nodules formed in long-term cultures. These same mediators also inhibit the mineralization of fully developed bone nodules in a time and dose-dependent fashion. The pro-inflammatory cytokines interleukin-1 alpha and tumor necrosis factor-alpha cause a dose-dependent induction of rat calvarial cell phospholipase A2-II mRNA levels, suggesting that their effects on bone formation may be mediated indirectly by activation of this enzyme. In contrast, transforming growth factor-beta 1, which has more potent effects on bone formation than interleukin-1 alpha or tumor necrosis factor-alpha, suppresses basal levels of phospholipase A2-II mRNA, indicating a different mechanism of action for this cytokine.