Bone formation by isolated calvarial osteoblasts in syngeneic and allogeneic transplants: light microscopic observations

Retroviral-based gene therapy with cyclooxygenase-2 promotes the union of bony callus tissues and accelerates fracture healing in the rat

BACKGROUND

An in vivo gene therapy strategy was developed to accelerate bone fracture repair.

METHODS

Direct injection of a murine leukemia virus-based vector targeted transgene expression to the proliferating periosteal cells arising shortly after fracture. Cyclooxygenase-2 (Cox-2) was selected because the transgene for its prostaglandin products that promote angiogenesis, bone formation and bone resorption, are all required for fracture healing. The human (h) Cox-2 transgene was modified to remove AU-rich elements in the 3'-untranslated region and to improve protein translation.

RESULTS

In vitro studies revealed robust and sustained Cox-2 protein expression, prostaglandin E(2) and alkaline phosphatase production in rat bone marrow stromal cells and osteoblasts transgenic for the hCox-2 gene. In vivo studies in the rat femur fracture revealed that Cox-2 transgene expression produced bony union of the fracture by 21 days post-fracture, a time when cartilage persisted within the fracture tissues of control animals and approximately 1 week earlier than the healing normally observed in this model. None of the ectopic bone formation associated with bone morphogenetic protein gene therapy was observed.

CONCLUSIONS

This study represents the first demonstration that a single local application of a retroviral vector expressing a single osteoinductive transgene consistently accelerated fracture repair.

Stromal stem cells: marrow-derived osteogenic precursors

Evidence is discussed for the hypothesis that there are stromal stem cells present in the soft connective tissues associated with marrow and bone surfaces that are able to give rise to a number of different cell lines including the osteogenic line. Fibroblastic colonies, each derived from a single colony-forming unit fibroblastic (CFU-F), are formed when marrow cells are cultured in vitro. In vivo assays of CFU-F have demonstrated that some CFU-F have a high ability for self renewal and multipotentiality whereas some have more limited potential. In vitro studies also support the hypothesis and have shown that CFU-F are a heterogeneous population of stem and progenitor cells and that their differentiation in vitro can be modified at the colony level. Factors added to the medium can activate osteogenesis in a range of multipotential and more committed precursors. Different stromal cell lines can be promoted under different culture conditions. The number and hierarchy of cell lines belonging to the stromal fibroblastic system are not yet fully elucidated and more specific markers for the different lines are required before a better understanding can be achieved.

Sustained In Vitro Expansion of Bone Progenitors Is Cell Density Dependent

Osteogenic cells are an integral part of the dynamic tissue-remodeling process in bone and are potential tools for tissue engineering and cell-based therapies. We examined the role of glucocorticoids and cell density in the expansion of primary rat calvaria cell populations and osteoprogenitor subpopulations in adherent cell culture. Osteoprogenitor response to dexamethasone (dex, a synthetic glucocorticoid known to stimulate bone formation in vitro) supplementation and long-term osteoprogenitor cell proliferation and differentiation were quantified using functional (colony forming unit-osteoblast [CFU-O]) and phenotypic analyses. Although osteoprogenitor self-renewal occurred at both standard and high initiating cell densities, progenitor cell expansion (measured by changes in CFU-O number relative to input) was sustained and dramatically increased at high initiating cell densities (30-fold CFU-O expansion for standard-density cultures compared with a greater than 10,000-fold CFU-O expansion in high-density cultures). Cell density was also found to impact upon the potential of dex to recruit additional progenitors towards bone development. These multifaceted effects appeared to be independent of cell proliferation rates or population phenotypic expression. Together, our results emphasize a roll for cell-cell interactions and/or community effects in the control and maintenance of progenitor cells during in vitro culture.

Expression of colony-stimulating factor-1 in vivo during the formation of osteoclasts

Colony-stimulating factor-1 (CSF-1), originally described as a growth factor for macrophages, is essential for the proliferation and differentiation of the cells of the osteoclast lineage. The cytokine is synthesized either as a secreted or a membrane-bound protein, which are encoded by four transcripts. The aim of the present study was to investigate the expression of CSF-1 in vivo at the mRNA level. Transcripts encoding CSF-1 were determined in total RNA from fetal murine metatarsals of different ages by a quantitative reverse-transcription polymerase chain reaction assay. Within the investigated period of time, the bone rudiments contain cells of the osteoclastic lineage representing well-defined differentiation stages. We found that only low levels of transcripts encoding CSF-1 could be detected in metatarsals from 15-day-old fetuses. Transcript levels increased slowly during the following days to reach a maximum in the rudiments from 18-day-old fetuses. After birth, in newborn animals, transcript levels were lowered again. While in rudiments from 15-day-old fetuses a considerable portion of the transcripts encoded the membrane-bound molecule, a transcript encoding the secreted form of the cytokine was the predominant species during the following days. These results suggest that the maintenance of proliferating and postmitotic osteoclast precursors requires low levels of CSF-1 only. Highest levels of locally synthesized CSF-1 are required, however, during the initial recruitment and activation of osteoclasts. After birth, levels of CSF-1 transcripts decrease again, suggesting that newly synthesized CSF-1 may be replaced by protein released from the mineralized matrix during resorption. In conclusion, the present data further strengthen the notion that CSF-1 produced locally acts in a paracrine fashion during the formation of osteoclasts.

Osteogenic potential of cultured periosteal cells in a distracted bone gap in rabbits

Osteogenesis of cultured periosteal cells was investigated in a bone-defective gap that was created artificially by distraction in the tibia of 12-week-old rabbits. A 10-mm circumferential length of periosteum was stripped from each stump of the osteotomized tibia, and the tibia was distracted rapidly (2 mm/day), resulting in disturbance of callus formation. Periosteal-derived cells, which were isolated from the contralateral tibia, were introduced into cell culture, subcultured twice to a population of 5 x 10(7) cells, and then injected into the defective bone gap when distraction was complete. Following inoculation of the cultured cells, significant new bone formation in the bone gap was observed. The control group which did not undergo cell transplantation showed only slight new callus formation which is supposed to be formed by osteogenic cells from the bone marrow. The bone mineral content of newly formed bone between the distracted tibia was analyzed quantitatively on radiographs. Histologically, the transplanted cells initially formed a mass at the injected site and then gradually differentiated into bone tissue from the peripheral region. Bromodeoxyuridine immunohistochemical stain was utilized to investigate the localization of the transplanted cells. The present study confirms that the orthotopically implanted periosteum-derived cells facilitate osteogenesis in a bone defect created using distraction in rabbits.

Bone formation in vivo: comparison of osteogenesis by transplanted mouse and human marrow stromal fibroblasts

BACKGROUND

Marrow stromal fibroblasts (MSFs) are known to contain bone precursor cells. However, the osteogenic potential of human MSFs has been poorly characterized. The aim of this study was to compare the osteogenic capacity of mouse and human MSFs after implantation in vivo.

METHODS

After in vitro expansion, MSFs were loaded into a number of different vehicles and transplanted subcutaneously into immunodeficient mice.

RESULTS

Mouse MSFs transplanted within gelatin, polyvinyl sponges, and collagen matrices all formed a capsule of cortical-like bone surrounding a cavity with active hematopoiesis. In transplants of MSFs from transgenic mice harboring type I procollagen-chloramphenicol acetyltransferase constructs, chloramphenicol acetyltransferase activity was maintained for up to 14 weeks, indicating prolonged bone formation by transplanted MSFs. New bone formation by human MSFs was more dependent on both the in vitro expansion conditions and transplantation vehicles. Within gelatin, woven bone was observed sporadically and only after culture in the presence of dexamethasone and L-ascorbic acid phosphate magnesium salt n-hydrate. Consistent bone formation by human MSFs was achieved only within vehicles containing hydroxyapatite/tricalcium phosphate ceramics (HA/TCP) in the form of blocks, powder, and HA/TCP powder-type I bovine fibrillar collagen strips, and bone was maintained for at least 19 weeks. Cells of the new bone were positive for human osteonectin showing their donor origin. HA/TCP powder, the HA/TCP powder-type I bovine fibrillar collagen strips, and HA/TCP powder held together with fibrin were easier to load and supported more extensive osteogenesis than HA/TCP blocks and thus may be more applicable for therapeutic use.

CONCLUSIONS

In this article, we describe the differences in the requirements for mouse and human MSFs to form bone, and report the development of a methodology for the consistent in vivo generation of extensive bone from human MSFs.

Synthesis of membrane- and matrix-bound colony-stimulating factor-1 by cultured osteoblasts

Colony-stimulating factor-1 (CSF-1) is synthesized as a secreted or membrane-bound molecule. We investigated whether osteoblastic cells produce these forms of CSF-1. Glutaraldehyde-fixed cell layers supported proliferation of the macrophage cell line BAC1.2F5, suggesting the presence of membrane- or/and matrix-associated CSF-1. Furthermore, CSF-1 activity could be either extracted from the matrix or released from the cell membrane. A neutralizing antiserum against CSF-1 inhibited these activities. After labeling the cellular proteins with [35S] met/cys or [35S] SO4(2-), CSF-1 was immunoprecipitated and analyzed by SDS-PAGE. Under nonreducing conditions, bands with MW more than 200, 200, 100, and 50 kd were detected. These bands shifted to lower MW under reducing conditions. Treatment with chondroitin lyase ABC decreased the MW of the 200 kd monomer, proving the proteoglycan structure. Much smaller quantities of CSF-1 were found in the matrix extract than in the conditioned medium. Transforming growth factor beta (TGF-beta) increased both the synthesis of CSF-1 and its accumulation in the matrix. CSF-1 released with trypsin from the membrane fraction yielded on SDS-PAGE a band with MW of 60 and 30 kd under nonreducing and reducing conditions, respectively. Transcripts encoding both the secreted and the membrane-associated forms of the cytokine were detected in osteoblasts by reverse transcription polymerase chain reaction. These data indicate that osteoblastic cells produce the secreted forms, either remaining in the culture supernatant, or being associated to the matrix, and the membrane associated form of CSF-1.

Attachment of human bone cells to tissue culture polystyrene and to unmodified polystyrene: the effect of surface chemistry upon initial cell attachment

Cell culture studies have often been used in the determination of the suitability of biomaterials as surfaces for the attachment and growth of cells. For such studies of surfaces for potential use in bone implants, cells derived from bone may be maintained in culture on tissue culture polystyrene (TCPS). We have determined the contribution that serum fibronectin (FN) or vitronectin (VN) make to the attachment and spreading of cells cultured from explanted human bone (bone-derived cells) during the first 90 min following seeding on culture surfaces. The attachment of bone-derived cells to TCPS was simulated two-fold by the addition of 10% (v/v) fetal bovine serum (FBS) to the seeding culture medium. The roles of FN and VN were determined by selective removal of the FN or VN from the FBS prior to addition to the culture medium. FBS from which the VN had been removed did not have this stimulatory activity. In contrast, the attachment of bone-derived cells onto TCPS from medium containing FN-depleted serum (which contained VN) was the same as when intact FBS was used. There was incomplete attachment of bone-derived cells (27% of cells) when seeded in medium containing FBS depleted of both VN and FN. Our results show that for human bone-derived cells, the attachment onto TCPS of cells planted in medium containing FBS during the first 90 min of culture is principally as a result of adsorption onto the surface of serum VN. As unmodified polystyrene (PS) has also been used previously as a model biomaterial surface, PS was compared to TCPS for attachment of the bone-derived cells. Attachment of bone-derived cells to TCPS was twice that onto PS, both when the medium was serum-free and when it contained FBS. Bone-derived cells attached to TCPS or PS onto which purified VN or FN had been precoated, with VN adsorbed onto PS being as effective as was VN adsorbed onto TCPS. With FN, there was an effect of the polystyrene surface chemistry which was evident in that suboptimal concentrations of FN had a slightly higher potency when adsorbed onto TCPS than did the same concentrations of FN coated onto PS. When preadsorbed onto TCPS, the potency of FN for attachment of bone-derived cells was at least equal to that of VN.

In vivo osteogenic activity of isolated human bone cells

Human bone cells were obtained as the outgrowth from cancellous bone fragments pretreated with collagenase and DNase. The osteogenic potential of cells in primary culture was assessed upon intramuscular transplantation into young mice pretreated with cortisone. Transplants were recovered after 2 weeks and examined by light microscopy. Of 34 transplants, 6 showed evidence of osteogenesis and 12 the production of unmineralized matrix. Only cells were observed in the other transplants. In an attempt to find a biochemical marker for osteogenic cells we have assayed medium osteocalcin and alkaline phosphatase activity levels in cultures before transplantation. No correlation was found between the level of expression of the two osteoblast markers and the osteogenic potential of the cells.

Bone and cartilage formation in diffusion chambers by subcultured cells derived from the periosteum

Periosteal cells were enzymatically isolated from the tibiae of young chicks, introduced into cell culture, allowed to reach confluence, and subcultured. The freshly isolated or subcultured cells were loaded into diffusion chambers and implanted into the peritoneal cavity of athymic mice to test their osteo-chondrogenic potential in a contained in vivo location. Freshly isolated periosteal cells formed both bone and cartilage tissue in such test chambers, but with a relatively low incidence. In contrast, cultured periosteal cells consistently gave rise to bone and cartilage even after 10 population doublings. With further passages of cells, the osteo-chondrogenic potential diminished substantially, until complete loss of expressivity at 16 population doublings or longer. Cultured muscle fibroblasts, when loaded into diffusion chambers under identical conditions to those of cultured periosteal cells, formed neither bone nor cartilage. These observations suggest that periosteal cells of young chicks contain subsets of progenitor cells or mesenchymal stem cells which possess the potential to differentiate into osteoblasts or chondrocytes, and this potential is retained after enzymatic isolation and for several population doublings in culture.

The adenylate cyclase response to parathyroid hormone in cultured rabbit marrow fibroblastic cells

The ability of fibroblastic cells to respond to parathyroid hormone (PTH) by an increase in adenylate cyclase activity is accepted as a characteristic of the osteogenic phenotype. Whether marrow fibroblastic cells, which have osteogenic potential when assayed in vivo, demonstrate this hormonal response when cultured in vitro has been investigated. Our study has shown a level of stimulation of adenylate cyclase activity by PTH in cultured rabbit marrow fibroblasts comparable with other osteogenic cells in vitro. The effect is seen in fibroblasts grown either from multiple colonies or from single colonies. Only a proportion of colonies had osteogenic potential in vivo assay and our results show a similar finding for the PTH response in vitro. To what degree the two parameters are expressed by the same colony has not yet been established.

Expression of the chondrogenic phenotype by mineralizing cultures of embryonic chick calvarial bone cells

Cells released by sequential enzymatic digestion of 18-day chick calvariae were cultured over a 4-5 week period in Alpha modified Eagles medium. In some cultures the medium was supplemented with ascorbate and/or Na-beta-glycerophosphate. Microscopic examination of these cultures showed both polygonal and spindle-shaped cells. The biochemical nature of these cells was investigated by incubating the cultures with radiolabelled proline and subsequently analysing the medium and cell layer proteins by SDS/PAGE and fluorography. Osteoblast and chondrocyte-containing cultures were clearly distinguished in this way as the former cells secreted type I collagen while the latter secreted types II and X collagens as the major medium macromolecules. Type X collagen synthesis occurred after 14 days, but only in cultures supplemented with both ascorbate and Na-beta-glycerophosphate, and was maintained for the duration of the culture period. Unsupplemented cultures and those containing either ascorbate alone or Na-beta-glycerophosphate alone failed to synthesize type X collagen after 28 days. Isolated cells pulsed with radiolabelled proline at confluence and organ cultures of embryonic chick calvaria synthesized types I and V collagens only. These data demonstrate that the expression of phenotype by heterogeneous populations of bone cells could be modulated by a combination of culture conditions including the length of time in culture and conditions favourable for the formation of a mineralized matrix.

Comparison of bone formed intramuscularly after transplantation of scapular and calvarial osteoblasts

Previous work suggested that osteoblasts determine the size of the bone marrow area within the bone and that calvarial osteoblasts differ from those induced intramuscularly by cartilage formed by transplanted epiphyseal chondrocytes. This study reports morphological observations of bone formed by transplanted scapular and calvarial osteoblasts isolated from bones of young rats. In intact scapulas of 28-day-old rats the percentage area occupied by bone tissue in relation to bone marrow was 6 times larger than in parietal bones of comparable age. Isolated syngeneic scapular osteoblasts usually produced an ossicle with similar general structure and ratio bone tissue/bone marrow area as in intact scapulas. In transplants of calvarial osteoblasts numerous islands of bone tissue with a small amount of bone marrow appeared. Bone formed in allogenic transplants was rejected. These results suggest that osteoblasts from endochondral scapular bone may have different properties than those from intramembranous calvarial bones. Alternatively, the large amount of medullary space in bone produced by transplanted scapular osteoblasts could result from their contamination with bone marrow stromal cells.

Mineralization in osteoblast cultures: a light and electron microscopic study

Osteoblasts isolated mechanically from newborn mouse calvaria produced a calcified matrix when cultured in the presence of 10 mM beta-glycerophosphate or 3 mM inorganic phosphate. The uncalcified matrix revealed numerous matrix vesicles scattered among collagen fibrils. The calcified matrix showed mineralized collagen fibrils and calcified nodules whose underlying organic matrix was detected after decalcification. These structures resembled those described in fetal and woven bone. In partially decalcified areas, calcification was shown to spread out from these structures along collagen fibrils. Alkaline phosphatase activity was found associated with the plasma membrane and matrix vesicles. X-ray diffraction analysis demonstrated that the mineral phase deposited in culture was hydroxyapatite. These observations which demonstrate that the isolated cells elaborate in culture a mineralized matrix with chemical and ultrastructural properties of woven bone further support the osteoblastic nature of the cells.

Ultrastructural analysis of bone nodules formed in vitro by isolated fetal rat calvaria cells

When cells enzymatically digested from 21 d fetal rat calvaria are grown in ascorbic acid and Na beta-glycerophosphate, they form discrete three-dimensional nodular structures with the histological and immunohistochemical appearance of woven bone. The present investigation was undertaken to verify that bone-like features were identifiable at the ultrastructural level. The nodules formed on top of a fibroblast-like multilayer of cells. The upper surface of the nodules was lined by a continuous layer of cuboidal osteoblastic cells often seen to be joined by adherens junctions. Numerous microvilli, membrane protrusions, and coated pits could be seen on the upper surface of these cells, their cytoplasm contained prominent RER and Golgi membranes, and processes extended from their lower surfaces into a dense, highly organized collagenous matrix. Some osteocyte-like cells were completely embedded within this matrix; they also displayed RER and prominent processes which extended through the matrix and often made both adherens and gap junctional contacts with the processes of other cells. The fibroblastic cells not participating in nodule formation were surrounded by a less dense collagenous matrix and, in contrast to the matrix of the nodules, it did not mineralize. An unmineralized osteoid-like layer was seen directly below the cuboidal top layer of cells. A mineralization front was detectable below this in which small, discrete structures resembling matrix vesicles and feathery mineral crystals were evident and frequently associated with the collagen fibrils. More heavily mineralized areas were seen further into the nodule. Electron microprobe and electron and X-ray diffraction analysis confirmed the mineral to be hydroxyapatite.(ABSTRACT TRUNCATED AT 250 WORDS)

Structural differences between bone formed intramuscularly following the transplantation of isolated calvarial bone cells or chondrocytes

Bone formed in intramuscular transplants of isolated syngeneic calvarial bone cells in mice, was compared with endochondral bone induced by cartilage produced by analogous transplants of isolated epiphyseal chondrocytes, as well as with parietal bones forming the bulk of the calvaria. Transplanted calvarial cells produced islands of bone, some of which contained intraosseous cavities. Osteoclasts inside these cavities were observed only in 14-day-old transplants and bone marrow cells in 28-day and older transplants. On the contrary, bone marrow appeared soon after formation of bone trabeculae in endochondral bone. The percentage area occupied by bone marrow in these specimens was about twentyfold larger than in the bone formed by transplanted bone cells. On the other hand, the bone marrow area in the latter type of bone was somewhat smaller but of similar order as in parietal bones. Moreover, both in parietal bones and in bone formed by isolated bone cells, the bone marrow was devoid of fat cells which were numerous in bone arising by endochondral ossification. It appears, therefore, that the ratio of bone marrow to the bone tissue area in parietal bones depends more on the intrinsic properties of osteoblasts than on the local factors in the environment of the developing bone. In the case of bone induced by cartilage, the bone marrow/bone tissue area could be determined both by the extent of cartilage resorption by vascularized tissue and by the properties of osteoblasts.

Remodeling of bone and bones: effects of altered mechanical stress on the regeneration of transplanted bones

We divided 116 rats weighing 50 gm into four groups with tails either left in situ or transplanted as follows: straight in situ: untreated controls; bent in situ: five caudal vertebrae (CV) in the loop; straight transplants: three CV skinned and transplanted autologously; and bent transplants: five CV skinned, bent to form a loop, and transplanted autologously. Tails were radiographed weekly up to 6 weeks and at 12 weeks, and microradiographic and histological studies were undertaken on selected specimens. At 12 weeks the bones in the apex of the loop of tails left in situ appeared bent with a straight-to-convex shaft on the outer side and a thicker, more concave one on the inner side. In the transplanted bent segments the bone shaft died and initially the reverse occurred: the outer shaft thickened and the inner resorbed completely. A new concave inner diaphysis then formed so that the bones in both instances were essentially similar in final shape. In the bent transplants the surviving osteogenic tissues regenerated and, adapting to the altered forces, formed a new bone shaft. This involved a change in the direction, amount, and nature of endochondral, periosteal, and regenerative growth and subsequent remodeling of bone. The results support previous observations that, within limits, the strain in the osteogenic envelope is an important factor in adaptation of bones to changing stress and that, where the envelope is deficient, the surviving tissues have the capacity to regenerate and repair defects in the bone so that it best resists the changing stresses applied to it.

Mineralized bone nodules formed in vitro from enzymatically released rat calvaria cell populations

Single-cell suspensions obtained from sequential enzymatic digestions of fetal rat calvaria were grown in long-term culture in the presence of ascorbic acid, Na beta-glycerophosphate, and dexamethasone to determine the capacity of these populations to form mineralized bone. In cultures of osteoblastlike cells grown in the presence of ascorbic acid and beta-glycerophosphate or ascorbic acid alone, three-dimensional nodules (approximately 75 micron thick) covered by polygonal cells resembling osteoblasts could be detected 3 days after confluency. The nodules became macroscopic (up to 3 mm in diameter) after a further 3-4 days. Only in the presence of organic phosphate did they mineralize. Nodules did not develop without ascorbic acid in the medium. Dexamethasone caused a significant increase in the number of nodules. Histologically, nodules resembled woven bone and the cells covering the nodules stained strongly for alkaline phosphatase. Immunolabeling with specific antibodies demonstrated intense staining for type I collagen that was mineral-associated, a weaker staining for type III collagen and osteonectin, and undetectable staining for type II collagen. Nodules did not develop from population I and the number of nodules formed by populations II-V bore a linear relationship to the number of cells plated (r = .99). The results indicate that enzymatically released calvaria cells can form mineralized bone nodules in vitro in the presence of ascorbic acid and organic phosphate.