Search

Osteoclast formation from the hemopoietic stem cell line FDCP-mix C2GM was shown to be strongly dependent on osteoblast density. In cocultures of C2GM cells with fetal mouse osteoblasts seeded at high density (i.e., 2.5 x 10(4) cells/cm2), we found a significantly lower osteoclast formation compared with cocultures with osteoblasts seeded at low density (i.e., 1 x 10(4) cells/cm2). The differentiation state of osteoblasts in high-density cultures resembled more than that of osteoblasts in low-density cultures, the differentiation state of mature osteoblasts, since the cells in the former cultures showed higher alkaline phosphatase (APase) activity than the cells in the latter cultures, and nodules were formed in high-density cultures but not in low-density cultures. Endogenous interleukin-6 (IL-6) production was found to be significantly lower in high-density cultures, which may partly explain the impaired osteoclast formation in high-density cocultures. Addition of IL-6 to the high-density cocultures indeed restored osteoclast formation. There appeared to be no overt difference in IL-6 receptor mRNA expression between high-density and low-density cultures. In conclusion, this paper suggests that mature, highly differentiated osteoblasts are not directly involved in osteoclastogenesis. In contrast, osteoblast-like cells lacking mature osteoblast markers induce osteoclast formation. Whether these low-density osteoblast-like cells represent an immature differentiation state or the lining cell phenotype is unclear.

Stem cell factor (SCF) is a polypeptide growth factor active on multiple cell types, mainly of hematopoietic origin. We studied the effects of avian SCF on the differentiation of chicken osteoclasts from their putative progenitors as well as on the bone-resorbing activity of terminally differentiated osteoclasts. Osteoclast formation was analyzed in long-term cocultures of osteoblasts and nonadherent, osteoclast-depleted bone marrow cells. Osteoclast activity was studied in short-term (48 h) cultures of bone marrow cell populations enriched for osteoclasts, on dentine slices. SCF strongly enhanced osteoclast differentiation. The IL-6-related chicken myelomonocytic growth factor (cMGF) had a similar effect, and the effects of SCF and cMGF were additive. SCF, but not cMGF, also stimulated the bone-resorbing activity of existing osteoclasts. As osteoblasts have been found to regulate osteoclast activity and formation, chicken osteoblasts were tested for their ability to express and secrete SCF. Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis showed that osteoblasts express SCF mRNA and that parathyroid hormone increases expression levels about fourfold. SCF did not accumulate in the culture medium, but remained cell (osteoblasts) surface associated.

A voltage-dependent, fast, transient inward current was characterized in embryonic chicken osteoclasts using the permeabilized patch configuration of the patch-clamp technique. The current was activated by depolarizations to higher than -28 +/- 4 mV from a holding potential of -80 mV. It peaked within 1-1.5 ms, and inactivated within 3.3-6.9 ms. The 50% inactivation voltage was -59 +/- 6 mV with a steepness factor of 0.11 +/- 0.06. The current disappeared with the removal of extracellular Na+ and was reversibly blocked by tetrodotoxin (K0.5 < 15 nM) but not by verapamil (< or = 100 microM). We conclude that this new current in embryonic chicken osteoclasts is a sodium current known from excitable cells.

It has been established, that the osteoclast is derived from the haemopoietic stem cell, but its exact lineage is still controversial. It is sometimes suggested, that osteoclasts and monocytes/macrophages are related cells. It has also been suggested that osteoclast differentiation is regulated by osteoblasts and 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3). In the present paper we addressed the question whether avian monocytes can differentiate into osteoclasts in vitro, using an array of immunocytochemical, enzyme cytochemical and function markers. We have also determined the effects of osteoblasts, osteoblast conditioned medium and 1,25-(OH)2D3 on the expression of osteoclastic features on monocytes during culture. Monocytes developed tartrate resistant acid phosphatase (TRAcP) enzyme activity and antigens for all anti-osteoclast antibodies tested, during culture. However, they did not acquire the ability to resorb dentine and still showed phagocytosis of latex spheres. This indicates that the monocytes developed into cells resembling osteoclasts but lacking their function while retaining the function of macrophages. Osteoblast conditioned medium stimulated TRAcP enzyme activity and proliferation of monocytes in cultures. Addition of osteoblasts or osteoblast conditioned medium to monocyte cultures on dentine in the presence or absence of 1,25-(OH)2D3 did not result in the generation of genuine osteoclasts, nor in pit formation. 1,25-(OH)2D3 appeared to be cytotoxic to the avian monocytes in concentrations considered optimal for mouse osteoclast formation. These results suggest that avian monocytes do not readily differentiate into osteoclasts under in vitro conditions that stimulate osteoclast differentiation from bone marrow derived haemopoietic cells. Furthermore, labelling with anti-osteoclast antibodies and TRAcP as osteoclast-markers should be used only with great caution in the identification of osteoclasts formed in vitro.

Many cell types respond to mechanical membrane perturbation with intracellular Ca2+ responses. Stretch-activated (SA) ion channels may be involved in such responses. We studied the occurrence as well as the underlying mechanisms of cell membrane stretch-evoked responses in fetal chicken osteoclasts using separate and simultaneous patch-clamp and Ca2+ imaging measurements. In the present paper, evidence is presented showing that such responses involve a self-reinforcing mechanism including SA channel activity, Ca(2+)-activated K+ (KCa) channel activity, membrane potential changes and local and general intracellular Ca2+ ([Ca2+]i) increases. The model we propose is that during membrane stretch, both SA channels and KCa channels open at membrane potential values near the resting membrane potential. SA channel characterization showed that these SA channels are permeable to Ca2+. During membrane stretch, Ca2+ influx through SA channels and hyperpolarization due to KCa channel activity serve as positive feedback, leading ultimately to a Ca2+ wave and cell membrane hyperpolarization. This self-reinforcing mechanism is turned off upon SA channel closure after cessation of membrane stretch. We suggest that this Ca2+ entry mechanism plays a role in regulation of osteoclast activity.

We succeeded in the isolation of osteocytes from parietal bones of 16-day-old chick embryos. Isolated osteocytes showed a typical stellate morphology. More than 95% of these cells reacted with the osteocyte-specific antibody OB 7.3. In culture osteocytes formed gap junctions with each other, as could be established by ACAS. Sixteen-millimeter time-lapse microcinematography of the cells also demonstrated the formation of intercellular connections and gap junctions, and portrayed the interaction between osteocytes and osteoclasts: osteocytes seemed to inhibit osteoclast activity. This cinematography also showed the ability of osteocytes to proliferate after they had been disconnected from each other. Thereafter these cells redifferentiated into osteoblasts that became embedded in bone matrix produced by themselves. These findings suggest that osteocytes might be involved in bone formation during remodeling.

Investigating the potentiality of cord monocytes to differentiate toward osteoclast-like cells (OCL) in vitro, we previously reported that in the presence of 1,25(OH)2 vitamin D3 (1,25-(OH)2D3), multinucleated-cells generated by cord monocyte cultures though displaying morphological features of OCL failed to resorb devitalized bones. We thus hypothesized that full differentiation of cord monocytes toward bone-resorbing cells may require the presence of factors released from and/or direct interactions with living osteogenic cells. In the present study, we tested these hypotheses using two culture systems supporting the development of bone-resorbing cells in the presence of bone matrix. First, cord mononuclear cells were co-cultured with murine fetal metatarsals depleted of osteoclast progenitor cells (stripped metatarsals) in the presence of 1,25-(OH)2D3. We found that cord mononuclear cells failed to differentiate toward OCL as indicated by the absence of the release of 45Ca previously incorporated in fetal bones and by the absence of formation of TRAP-positive (TRAP[+]) multinucleated cells which have invaded mineralized cartilage during the co-culture period. In the same model, we then investigated the effect of some soluble factors known as stimulators of osteoclast differentiation. Whereas exogenous rhIL6 and rhIL3 were ineffective in this assay, rhM-CSF consistently increased both the number of TRAP(+) multinucleated cells inside the mineralized cartilage and the release of 45Ca into the culture media. The effects of rhM-CSF were time-dependent reaching the maximum after 3 weeks of culture.(ABSTRACT TRUNCATED AT 250 WORDS)

Six murine macrophage precursor cell lines, thought to be arrested around the CFU-GM stage of the myeloid differentiation and shown to be negative for acid phosphatase, F4/80 antigen expression and phagocytosis capacity, were tested for their ability to differentiate into osteoclasts. Their differentiation potential was compared with that of the haemopoietic stem cell line FDCP-mix C2GM. None of the macrophage precursor cell lines could be induced to differentiate into osteoclasts when the cells were cocultured with either periosteum-free metatarsal bones of fetal mice, or monolayers of osteoblast-like cells. In contrast, when the haemopoietic stem cell line FDCP-mix C2GM, murine fetal liver cells or murine spleen cells were used as a source of haemopoietic precursor cells, numerous osteoclasts were formed in both culture systems. During cell culture a small percentage of the macrophage precursor cells attached to the bottom of the culture well. These firmly attached cells acquired acid phosphatase activity, F4/80 antigen expression and phagocytosis capacity. Furthermore, when the cell lines were cultured for 2 or 4 days with 1% DMSO, up to 30% of the precursor cells differentiated into metamyelocytes. These results suggest that the macrophage precursor cell lines are able to acquire macrophage and granulocyte characteristics, but are unable to differentiate into osteoclasts. In contrast, the haemopoietic stem cell line FDCP-mix C2GM is able to differentiate into both macrophages and osteoclasts. We therefore suggest that the osteoclast lineage branches off at an early stage of the myeloid differentiation pathway.

In the coculture system of periostless metatarsal bones of 17-day-old fetal mice and osteoclast progenitors, osteoclasts will develop. Our goal in the present report was to provide further evidence that in the coculture system of fetal metatarsal bone rudiments with hemopoietic cells, the osteoclasts developing inside the bone rudiments are exclusively derived from the cells suspended in the plasma clot and not from endogenous precursor cells of the bone explants themselves, by using the technique of in situ hybridization with a probe for the mouse Y chromosome. Osteoclast formation in unstripped male metatarsal rudiments, occurring after 3-4 days of culture, was compared with osteoclast formation in cocultures of female metatarsal rudiments and male bone marrow cells, occurring after 5-6 days of culture. Osteoclasts were recognized by their tartrate-resistant acid phosphatase activity. In paraffin sections of cultured male metatarsals, the mean percentage of microscopically identifiable osteoclast nuclei, in which the Y chromosome could be detected, was 43.1 +/- 4.2% (n = 12). For cocultures of female metatarsal bones and male bone marrow cells this mean percentage was 40.9 +/- 5.7% (n = 17). Statistical comparison by means of the two sample t-test indicated no significant difference in the percentages of osteoclast nuclei containing the Y chromosome for both groups. We concluded that the osteoclasts do derive from cocultured cells and not from precursor cells in the bone explant itself.(ABSTRACT TRUNCATED AT 250 WORDS)

Calcium-activated potassium channels were found in embryonic chick osteoclasts using the patch-clamp technique. The activity of the channel was increased by both membrane depolarisation and an increase in intracellular Ca2+ concentration in the range 10(-5) to 10(-3) M. In the cell-attached-patch configuration the channel was only active at extreme depolarising potentials. Ca2+ addition to the cytoplasm via ionomycin increased channel activity at the resting membrane potential of the osteoclast. The channel had a single-channel conductance of 150 pS in the inside-out patch under symmetrical K+ conditions (150 mM) and was selective for potassium ions. During sustained application of increased [Ca2+] at the cytoplasmic side of inside-out patches, channel activity sometimes decreased again after the initial increases (desensitization). The results established the properties of the single channels underlying an outward rectifying K+ conductance in chick osteoclasts described previously by us.

Embryonic chick bone cells express various types of ionic channels in their plasma membranes for as yet unresolved functions. Chick osteoclasts (OCL) have the richest spectrum of channel types. Specific for OCL is a K+ channel, which activates (opens) when the inside negative membrane potential (Vm) becomes more negative (hyperpolarization). This is consistent with findings of others on rat OCL. The membrane conductance constituted by these channels is called the inward rectifying K+ conductance (GKi), or inward rectifier, because the hyperpolarization-activated channels cause cell-inward K+ current to pass more easily through the membrane than outward K+ current. Besides GKi channels, OCL may express two other types of voltage-activated K+ channels. One constitutes the transient outward rectifying K+ conductance (GKto), which is activated upon making the membrane potential less negative (depolarization) but has a transient nature. This conductance favors transient K+ conduction in the cell-outward direction. The GKto also occurs in a small percentage of cells in osteoblast (OBL) and periosteal fibroblast (PFB) cultures. The other OCL K+ conductance, the GKCa, is activated by both membrane depolarization and a rise in [Ca2+]i. GKCa channels are also present in the other chick bone cell types, that is, OBL, osteocytes (OCY), and PFB. Furthermore, in excised patches of all bone cell types, channels have been found that conduct anions, including Cl- and phosphate ions. These channels are only active around Vm = 0 mV. While searching for a membrane mechanism for adaptation of bone to mechanical loading, we found stretch-activated channels in chick osteoclasts; other investigators have found stretch-activated cation channels (K+ or aselective) in rat and human osteogenic cell lines. In contrast to other studies on cell lines or OBL from other species, we have not found any of the classic macroscopic voltage-activated calcium conductances (GCa) in any of the chick bone cells under our experimental conditions. However, our fluorescence measurements of [Ca2+]i in single cells indicate the presence of Ca2+ conductive pathways through the plasma membrane of osteoblastic cells and osteoclasts, consistent with other studies. We discuss possible roles for GKi, GKCa, and anion channels in acid secretion by OCL and for stretch-activated channels in OCL locomotion.

Bone remodeling is a complex process coupling bone formation and resorption. Osteoblasts, the bone-forming cells, are known to produce various bone matrix proteins and cytokines; however, little is known about protein factors produced by osteoclasts or bone-resorbing cells. A method utilizing the high affinity of osteoclasts for tissue culture dishes was developed to isolate a large number of pure osteoclasts from rabbit long bones. A cDNA library was then constructed from these isolated osteoclasts, and differential cDNA screening was performed between osteoclasts and spleen cells. Two clones representing osteoclast-specific clones, named OC-1 and OC-2, were isolated. By Northern blot analysis, OC-1 was expressed in osteoclasts and in kidneys, whereas OC-2 was specific for osteoclasts. OC-1 was found to encode osteopontin from its nucleotide sequence, and therefore, osteopontin may have other functions for osteoclastic bone resorption besides osteoclast attachment to bone.

Recently we reported that the osteoclast originates from the pluripotent hematopoietic stem cell. However, a detailed analysis of the progenitor and precursor stages of the osteoclast lineage is hard to perform with primary cultures of stem cells. In the present investigation interleukin-3 (IL-3)-dependent multipotent hematopoietic stem cell lines (FDCP-mix), which have many characteristics in common with freshly isolated hematopoietic stem cell lines (FDCP-mix), which have many characteristics in common with freshly isolated hematopoietic stem cells, were assayed for their osteoclast formation capacity. FDCP-mix cell lines A4, C2GM, and 15S were cocultured with periosteum-free 17-day-old fetal metatarsal bones. The effects of culture time, medium composition, and addition of WEHI-3b-conditioned medium (an unpurified IL-3 preparation) on osteoclast formation were studied. 15S cells never differentiated into osteoclasts. Both A4 and C2GM cells were able to generate osteoclasts. Osteoclast formation was visualized by staining for tartrate-resistant acid phosphatase activity and confirmed by 45Ca release assays and electron microscopic studies. Medium supplemented with fetal calf serum clearly supported osteoclast formation from A4 cells better than medium supplemented with cock serum. The difference between fetal calf serum and horse serum is generally less pronounced. C2GM cells formed osteoclasts more readily and, generally, earlier than A4 under all culture conditions. WEHI-3b-conditioned medium addition increased the numbers of osteoclasts and their resorption activity. The coculture of stripped metatarsal bones with FDCP-mix cell lines therefore offers a model system with many possibilities for the study of osteoclastogenesis and its regulation.

The method of Barka and Anderson was used for the demonstration of tartrate-resistant acid phosphatase (TRAcP) in cultures of bone marrow, spleen, lung, and peritoneal cells of the mouse. The staining was performed either in the usual way by adding both substrate (naphthol-AS-BI-phosphate) and coupler (hexazonium pararosanilin) together (the simultaneous-coupling technique) or by adding first the substrate and then the coupler (the post-coupling technique). We measured TRAcP-activity fluorometrically after extraction of the product naphthol-AS-BI, using the same staining solution as in cytochemical method, but without the coupler. In bone marrow, spleen, lung, and peritoneal cell cultures a biochemically measurable TRAcP-activity was detected. Post-coupling generally gave a higher level of staining and larger numbers of TRAcP-positive cells than simultaneous-coupling. In bone marrow cultures macrophages, identifiable by their ability to phagocytose microspheres, became TRAcP-positive during culture. In lung cell cultures cells capable of phagocytosis of bacteria were shown to be TRAcP-positive. Peritoneal macrophages remained TRAcP-negative in the simultaneous-coupling technique. Using the post-coupling technique a small number stained TRAcP-positive. In spleen cell cultures TRAcP-positive cells containing hemosiderin were visible. In cultures of all four cell types, F4/80 positive cells staining also for TRAcP were present. F4/80 is a well known marker for macrophages, whereas osteoclasts are negative. In conclusion, mouse macrophages originating from various tissues can become TRAcP-positive in vitro. TRAcP activity alone is not a reliable marker for osteoclasts in bone marrow cultures.

Osteoclast progenitors are seeded via the blood stream in the mesenchyme surrounding embryonic long bone models long before the appearance of multinucleated osteoclasts. The proliferation and differentiation of these progenitors in embryonic mouse metatarsal bones was studied with acid phosphatase (AcP) histochemistry and 3H-thymidine autoradiography. In vivo, tartrate-resistant, acid phosphatase-positive, mononuclear cells appear in the periosteum (AcPP-P cells) at the age of 17 days (after conception). On day 18, AcP-positive, multinucleated osteoclasts invade the bone rudiment and start resorbing the calcified cartilage matrix, resulting in the formation of the marrow cavity. The kinetics of osteoclast formation in vitro was studied in metatarsal bones of embryonic mice of different ages cultured in the continuous presence of 3H-thymidine. In young bones (15 days), mainly proliferating, 3H-thymidine-incorporating progenitors gave rise to AcPP-P cell and osteoclast formation. In older bones (16 and 17 days) osteoclasts were progressively more derived from postmitotic, unlabeled precursors. Irradiation of the metatarsal bones with a radiation dose of 5.0 Gy prior to culture resulted in a selective elimination of the proliferating progenitors, whereas the contribution of postmitotic precursors in AcPP-P cell and osteoclast formation remained unchanged. The results demonstrate that in the periosteum of embryonic metatarsal bones a shift occurs from a population composed of proliferating osteoclast progenitors (15 days) to a population composed of postmitotic precursors (17 days) before multinucleated osteoclasts are formed (18 days). Obviously, postmitotic AcP-negative precursors, already present in 16-day-old bones, differentiate into precursors characterized by tartrate-resistant AcP activity, the preosteoclasts (17 days), which in their turn fuse into osteoclasts.

Patch-clamp measurements on freshly isolated embryonic chicken osteoclasts revealed three distinct types of voltage-dependent K+ conductance. The first type of conductance, present in 72% of the cells, activated at membrane potentials less negative than -30 to -20 mV and reached full activation at +40 mV. It activated with a delay, reached a peak value, and then inactivated with a time constant of approximately 1.5 s. Inactivation was complete or almost so. Recovery from inactivation, at -70 mV, had a time constant of roughly 1 s. The conductance could be blocked, at least partly, by 4 mM 4-aminopyridine. The second type of conductance (present in all cells) activated at membrane potentials more negative than -40 to -80 mV and reached full activation at -130 mV. Activation potential and maximal conductance were dependent on the extracellular K+ concentration. Inactivation of the conductance first became apparent at membrane potentials more negative than -100 mV and was a two-exponential process. The conductance could be blocked by external 5 mM Cs+ ions. The third type of conductance (present in all cells) activated at membrane potentials more positive than +30 mV. Generally, the conductance did not inactivate.

The effects of ionizing irradiation on the differentiation and activity of the osteoclast were investigated. Embryonic mouse metatarsal bones of different ages (14, 15, 16, 17 days) in which no osteoclasts had as yet been formed were irradiated with various x-ray doses and cultured until a marrow cavity became visible in the nonirradiated paired control bones. Bone growth and calcification were followed microscopically during culture. Irradiation caused a dose-dependent stunting of the longitudinal growth. Calcification was inhibited by high radiation doses (10 to 20 Gray (Gy), whereas a dose of 2.5 Gy stimulated the process in the early stages of long bone development. Histologic examination revealed complete inhibition of osteoclast formation in the 14- and 15-day-old bones after irradiation with 2.5 Gy or more. The number of osteoclasts in cultured older bones (16 days) was significantly reduced by irradiation, but osteoclast formation could not be completely prevented even by high dosages. Irradiation of explanted bone rudiments which were in a stage 1 day prior to the appearance of osteoclasts in vivo (17 days) did not significantly influence the formation of osteoclasts. Autoradiographic experiments using young bones showed that differentiation of osteoclast precursors into multinucleated osteoclasts is preceded by one or more divisions of the precursors in the periosteum. Furthermore, it was established from continuous 3H-thymidine-labeling experiments that in older bones (16 days) a part of the osteoclast nuclei originated from postmitotic osteoclast precursors. Irradiation mainly inhibited the appearance of labeled osteoclast nuclei in these bones. The results indicate that the osteoclast precursor, already present in the periosteum at an early stage of embryonic development, first proliferates and then differentiates into a mononuclear postmitotic preosteoclast. The proliferation is probably highly radiosensitive. Subsequently, the preosteoclasts fuse into multinucleated osteoclasts and invade the calcified hypertrophic cartilage zone. The resorbing activity of the osteoclast is less radiosensitive but can be inhibited by 5.0 Gy or more, as was established by morphometric and biochemical methods.

An in vitro co-culture system was applied to study the direct and indirect effects of irradiation on osteoclast formation. Osteoclast precursor-free fetal mouse metatarsal bones were employed as osteoclast-forming inductor and periostea dissected from fetal calvaria as source of proliferating progenitor cells. Direct radiation effects on the formation of osteoclasts were assessed in co-cultures of irradiated periostea and non-irradiated bone rudiments. The results showed that the (blood-borne) periosteal progenitors were rather radiosensitive. A radiation 'survival' curve of osteoclast formation in relation to various doses could be constructed yielding a mean lethal dose (Do value) of 0.94 +/- 0.02 Gy and an extrapolation number of 1.67 +/- 0.01. Irradiation of the fetal long bones by low doses, effective for direct elimination of osteoclast progenitor cells, did not indirectly affect osteoclast development from the non-irradiated periosteal progenitor population. However, at relatively high radiation levels, though not lethal for the long bone rudiments, a significant inhibition of osteoclast formation became evident. The results indicate that radiation primarily affects osteoclast formation via a direct action on radiosensitive, proliferating progenitor cells. Injury to long bone models by relatively high radiation doses may also lead to severe disturbance of osteoclast formation kinetics.

The properties of five monoclonal antibodies raised against isolated osteoclasts are described. Osteoclasts were isolated from medullary bone of egg-laying female quails. Mice were immunized with cell preparations consisting for about 10% of multinucleated osteoclasts. A large number of monoclonal antibodies against cell surface antigens were obtained, five of which were extensively characterized by their interactions with different tissues of the quail and their cross-reactivity with other species. Two monoclonals (OC 5.3 and OC 6.8), recognize surface antigens present on osteoclasts, monocytes, granulocytes and endothelial cells, but not on osteoblasts, osteocytes, fibroblasts, lymphocytes, erythrocytes and others. The three other monoclonal antibodies are specific for multinucleated osteoclasts in bone tissue but recognize some cell surface structures in other tissues. Antibody OC 6.9, which in bone tissue stains primarily the surface area of the osteoclast that is adjacent to the resorbing bone surface, also interacts with bile capillaries in the liver and with specific, but not yet identified parts of the nephron. The antibodies OC 6.1 and OC 6.3 interact with Kupffer cells in the liver and tissue macrophages of small intestine. In view of the possible fallacies inherent to the use of cell surface markers for the demonstration of cell relationship and origin, definite conclusions can not yet be made. The fact that the osteoclast, the Kupffer cell and the intestine macrophage are the only cells in bone, bone marrow, liver, kidney and intestine, that share the same surface antigen recognized by monoclonals OC 6.1 and OC 6.3, suggests, however, a common origin for osteoclasts and a number of well described tissue macrophages.

In the present report osteoclast formation from cloned pluripotent hemopoietic stem cells (PHSC) is described. Populations enriched in hemopoietic stem cells were cloned (1 cell/well) and cultured in the presence of different colony-stimulating factors, or combinations of these growth factors. In cultures containing interleukin-3 (Il-3) or pregnant mouse uterus extract (PMUE) alone, cloning efficiency was low. Cultures containing Il-3 and Il-1 or Il-3 and PMUE showed a somewhat higher cloning efficiency, whereas cultures containing Il-3, Il-1 and PMUE had the highest cloning efficiency. All colonies of cloned PHSC, tested for their osteoclast formation capability in cocultures with periosteum-free metatarsal bones of fetal mice, gave rise to osteoclast formation. Other hemopoietic cells could also be demonstrated. In control cultures in which the bones were kept without stem cells, no osteoclast formation was observed. In conclusion, we have demonstrated that the osteoclast is derived from the pluripotent hemopoietic stem cell. A combination of various growth factors is important for stem cell proliferation in vitro.

Bone cells obtained by digestion of fetal mouse or chicken calvaria were tested for their ability to form or resorb bone in vitro. The isolated cells were precultured for 6 days and subsequently cocultured for 11 days with periosteum-free noninvaded fetal mouse long bone rudiments. Bone formation and resorption during coculture were evaluated by histology and 45Ca release from prelabeled bones. The calvarial origin of cells in cocultures was traced by labeling the cells with 3H-thymidine before coculture, followed by autoradiography. Many osteoblasts and osteoclasts as well as fibroblasts developed from mouse periosteal cells released late in the sequential digestion procedure and previously denoted as "osteoblastlike" (BL). No or few osteoblasts and osteoclasts but many fibroblasts developed from early released cell fractions that have previously been denoted as "osteoclastlike" (CL). Only osteoblasts and fibroblasts but not osteoclasts developed from chicken calvarial cell fractions. The osteoblasts developed primarily from cell fractions from the inner layer of the periosteum, previously denoted as "osteoblastlike" (OB). Cells obtained from the outer layer of the periosteum (PF) gave rise mainly to fibroblasts. These studies show that osteoblast and osteoclast precursor cells are maintained in monolayer cultures of periosteal cell fractions. However, sequential digestion of mouse calvaria does not lead to separation of the two types of bone cells. Rather, osteoclast and osteoblast precursors are released jointly, from the periosteal cell layers closest to the bone surface. In the chicken cell fractions osteoclast precursors are absent after preculture, resulting in a more homogeneous population of osteoblast and fibroblast but not osteoclast precursors.

It is well established that the osteoclast is formed by fusion of post-mitotic, mononuclear precursors derived from circulating progenitor cells. However, the precise haematopoietic origin of the osteoclast is unknown. We have investigated this here by fractionating mouse bone marrow and isolating haematopoietic stem cells using a three-step method combining equilibrium density centrifugation and two fluorescence-activated cell sortings (FACS), and have tested the ability of each bone marrow fraction, including highly purified haematopoietic stem cells, to generate osteoclasts during co-culture with preosteoclast-free embryonic long bones. The osteoclast-forming capacity was found to increase with increasing stem cell purity. On the other hand, the culture time needed for osteoclast formation also increased with purification, suggesting the presence of progressively more immature progenitor cells. The pluripotent haematopoietic stem cell fractions with the highest purity needed preincubation with a stem cell-activating factor (interleukin-3) to activate the predominantly quiescent stem cells in vitro.

In a previous study, using co-cultures of embryonic bone rudiments stripped of periosteum, and mononuclear phagocytes of various sources, we found that multinucleated mineral-resorbing osteoclasts developed in vitro from radiosensitive mouse bone marrow mononuclear phagocytes (BMMP). (Burger, E. H., J. W. M. van der Meer, J. S. van de Gevel, C. W. Thesingh, and R. van Furth, 1982, J. Exp. Med. 156:1604-1614). In the present study, this co-culture technique was used to analyze the influence of bone-forming cells on osteoclast formation and bone resorption by BMMP or peritoneal exudate cells (PEC). BMMP or PEC were co-cultured with liver or dead bone, i.e., in the presence or absence of liver bone-forming cells. Mineral resorption and osteoclast formation were monitored via 45Ca release from prelabeled live or dead bone followed by histology. Osteoclasts developed from precultured BMMP as indicated by [3H]thymidine labeling, but only in live and not in dead bone. They formed readily from BMMP but only erratically, and after a longer culture period, from PEC. Macrophages from BMMP and PEC invaded live and dead bone rudiments but did not resorb the intact mineralized matrix. In contrast, ground bone powder was resorbed avidly by both cell populations, without formation of osteoclasts. We conclude that live bone-forming cells are required for osteoclast formation from progenitors. Live bone is only resorbed by osteoclasts, and not by macrophages. Osteoclast progenitors are abundant in cultures of BMMP but scarce in PEC, which makes a direct descendance of osteoclasts from mature macrophages unlikely.

Hemopoietic stem and progenitor cells from different sources differ in radiosensitivity. Recently, we have demonstrated that the multinucleated cell responsible for bone resorption and marrow cavity formation, the osteoclast, is in fact of hemopoietic lineage. In this investigation we have studied the radiosensitivity of osteoclast formation from two different hemopoietic tissues: fetal liver and adult bone marrow. Development of osteoclasts from hemopoietic progenitors was induced by coculture of hemopoietic cell populations with fetal mouse long bones depleted of their own osteoclast precursor pool. During culture, osteoclasts developed from the exogenous cell population and invaded the calcified hypertrophic cartilage of the long bone model, thereby giving rise to the formation of a primitive marrow cavity. To analyze the radiosensitivity of osteoclast formation, either the hemopoietic cells or the bone rudiments were irradiated before coculture. Fetal liver cells were found to be less radiosensitive than bone marrow cells. The D0, Dq values and extrapolation numbers were 1.69 Gy, 5.30 Gy, and 24.40 for fetal liver cells and 1.01 Gy, 1.85 Gy, and 6.02 for bone marrow cells. Irradiation of the (pre)osteoclast-free long bone rudiments instead of the hemopoietic sources resulted in a significant inhibition of osteoclast formation at doses of 4 Gy or more. This indirect effect appeared to be more prominent in the cocultures with fetal than with adult hemopoietic cells. Furthermore, radiation doses of 8.0-10.0 Gy indirectly affected the appearance of other cell types (e.g., granulocytes) in the newly formed but underdeveloped marrow cavity. The results indicate that osteoclast progenitors from different hemopoietic sources exhibit a distinct sensitivity to ionizing irradiation. Radiation injury to long bone rudiments disturbs the osteoclast-forming capacity as well as the hemopoietic microenvironment.