David M. Rosen

Studies in Cranial Suture Biology: Part I. Increased Immunoreactivity for TGF-β Isoforms (β1, β2, and β3) During Rat Cranial Suture Fusion†

The mechanisms involved in normal cranial suture development and fusion as well as the pathophysiology of craniosynostosis, a premature fusion of the cranial sutures, are not well understood. Transforming growth factor‐β isoforms (TGF‐β1, β2, and β3) are abundant in bone and stimulate calvarial bone formation when injected locally in vivo. To gain insight into the role of these factors in normal growth and development of cranial sutures and the possible etiology of premature cranial suture fusion, we examined the temporal and spatial expression of TGF‐β isoforms during normal cranial suture development in the rat. In the Sprague‐Dawley rat, only the posterior frontal cranial suture undergoes fusion between 12 and 22 days of age, while all other cranial sutures remain patent. Therefore, immunohistochemical analysis of the fusing posterior frontal suture was compared with the patent sagittal suture at multiple time points from the fetus through adult. Whereas the intensity of immunostaining was the same in the posterior frontal and sagittal sutures in the fetal rat, there was increased immunoreactivity for TGF‐β isoforms in the actively fusing posterior frontal suture compared with the patent sagittal suture starting 2 days after birth and continuing until approximately 20 days. There were intensely immunoreactive osteoblasts present during fusion of the posterior frontal suture. In contrast, the patent sagittal suture was only slightly immunoreactive. A differential immunostaining pattern was observed among the TGF‐β isoforms; TGF‐β2 was the most immunoreactive isoform and was also most strongly associated with osteoblasts adjacent to the dura and the margin of the fusing suture. Since the increased expression of TGF‐β2 during suture fusion suggested a possible regulatory role, recombinant TGF‐β2 was added directly to the posterior frontal and sagittal sutures in vivo to determine if suture fusion could be initiated. Exogenously added TGF‐β2 stimulated fusion of the ectocranial surface of the posterior frontal suture. These data provide evidence for a regulatory role for these growth factors in cranial suture development and fusion. Additionally, the intense immunostaining for TGF‐β2 in the dura mater underlying the fusing suture supports a role for the dura mater in suture fusion. It is possible that premature or excessive expression of these factors may be involved in the etiopathogenesis of craniosynostosis and that modulation of the growth factor profile at the suture site may have potential therapeutic value.

Transforming Growth Factor-β2 Enhances the Osteoinductive Activity of a Bovine Bone-Derived Fraction Containing Bone Morphogenetic Protein-2 and 3

Abstract We previously identified a novel glycoprotein in an osteoinductive fraction from bovine bone (Bentz et al.: Amino acid sequence of bovine osteoinductive factor. J. Biol. Chem. 265: 5024-5029, 1990). We now find that this fraction also contained small amounts of bone morphogenetic protein-2 and 3 (BMP-2+3) previously identified by others (see Wozney, J. M.: Bone morphogenetic proteins. Prog. Growth Factor Res. 1: 267-280, 1990). Separation of BMP-2+3 from the glycoprotein was achieved with a modified reversed phase-high pressure liquid chromatographic procedure. When assayed in the rat subcutis using a collagen-ceramic carrier, the osteoinductive activity was found in the subtraction containing BMP-2+3. This activity was potentiated and the ratio of cartilage to bone was increased by transforming growth factor-β2. The glycoprotein, originally called osteoinductive factor, has been renamed osteoglycin. In its precursor form, osteoglycin is a member of the leucine-rich family of proteins showing the characteristic 24-residue internal homology. Its biological function is unknown.

Systemic administration of recombinant transforming growth factor beta 2 (rTGF-β2) stimulates parameters of cancellous bone formation in juvenile and adult rats

Transforming growth factor beta is a multifunctional protein with known actions on bone and bone cells. The ability of TGF-beta to stimulate osteogenic parameters in vitro and osteogenesis adjacent to injection sites in vivo is well established. The purpose of this study was to determine if systemic administration of recombinant TGF-beta 2 (rTGF-beta 2) could stimulate bone formation in rats of different ages. Juvenile (25-day-old) and adult (160-day-old) rats were treated daily for 5 days and 14 days, respectively, with rTGF-beta 2 given by subcutaneous injection. Bone formation was measured in cancellous bone of the lumbar vertebrae in juvenile rats and the femoral epiphysis in adult rats. Endochondral bone growth rates were measured in the distal femurs from both juvenile and adult rats using histomorphometric methods. Systemic administration of rTGF-beta 2 resulted in substantial increases in bone formation rates (both surface and volume referent) in both juvenile and adult rats. In the juvenile rats, rTGF-beta 2 increased the percent double labeled surface and the mineral appositional rate. In the adult rats, TGF-beta 2 treatment increased the double labeled surface and also endochondral (longitudinal) growth parameters without changing the number of osteoclasts or the number of osteoclast nuclei per cell. These results demonstrate that short-term systemic administration of rTGF-beta 2 substantially increases cancellous bone formation rate in rats.

Treatment of ovariectomized rats with the complex of rhIGF-I/IGFBP-3 Increases cortical and cancellous bone mass and improves structure in the femoral neck

Sixteen-week-old Sprague-Dawley rats were ovariectomized (Ovx) or sham-operated and housed for 8 weeks to develop osteopenia prior to systemic administration of rhIGF-I (0.9 and 2.6 mg/kg) alone or the rhIGF-I/IGFBP-3 (0.9, 2.6 and 7.5 mg/kg) complex. After 8 weeks of treatment, proximal femurs were fixed, embedded, and cut through the midneck region. Structural and dynamic histomorphometric analyses were performed using standard techniques. Ovx increased endocortical resorption and modeling-dependent periosteal formation which resulted in decreased cortical bone area. Despite increased bone formation, trabecular number, thickness, and area were all reduced due to increased resorption. Structural changes following Ovx included fewer struts and nodes, a higher percentage of the simpler strut forms, and reduced endocorticotrabecular cnnnectivity. Eight weeks of treatment with rhIGF-I or rhIGF-I/IGFBP-3 promoted periosteal and endocortical bone formation and reduced the endocortical resorption induced by Ovx. Both rhIGF-I formulations stimulated bone formation on existing trabecular surfaces which increased trabecular thickness and area but not trabecular number. These treatments prevented further deterioration of the trabecular network caused by Ovx and preserved endocortico-trabecular connectivity. In summary, changes in the femoral neck following Ovx appear to be similar in rats and humans. The highest dose of rhIGF-I/IGFBP-3 used in this study showed the best results in promoting cortical and cancellous bone formation, and appears to be promising therapy for human osteopenias.

Systemic administration of rhIGF-I or rhIGF-I/IGFBP-3 increases cortical bone and lean body mass in ovariectomized rats

The purpose of this study was to compare dose-related effects on cortical bone and lean body mass following subcutaneous administration of rhIGF-I alone, or bound to an equimolar amount of rhIGFBP-3 to adult Ovx rats. At the age of 16 weeks, rats were ovariectomized or sham-operated and were allowed 8 weeks to develop osteopenia. After being divided into control (saline treated) or treatment groups, rats were injected daily during an 8-week period with 0.9 and 2.6 mg/kg of rhIGF-I, or with 0.9, 2.6, and 7.5 mg/kg of rhIGF-I bound to rhIGFBP-3. Fluorescent bone markers were given 9 and 2 days prior to necropsy. Body weights and lean body mass were monitored throughout the experiment. Cortical bone histomorphometry was performed on tibial cross-sections at the tibiofibular junction, and endochondral bone growth was measured at the distal femoral metaphysis. All rats treated with rhIGF-I or the rhIGF-I/IGFBP-3 complex had increased body weights, corresponding to a dose-dependent increase in lean body mass. Endochondral growth was slightly increased in all experimental groups, but was not dose-dependent. A dramatic increase in periosteal, modeling-dependent formation, coupled with decreased or unchanged resorption on the endocortical envelope resulted in a dose-dependent increase in cortical thickness and cross-sectional area in groups treated with the complex of rhIGF-I/IGFBP-3. This complex appeared to be more effective in promoting positive musculoskeletal changes than rhIGF-I alone.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of systemically administered rhIGF-I/IGFBP-3 complex on cortical bone strength and structure in ovariectomized rats.

The action of systematically administered recombinant human insulinlike growth factor-I (rhIGF-I) complexed to its natural binding protein-3 (rhIGFBP-3) on cortical bone dynamic, structural, and mechanical properties was tested in previously ovariectomized (Ovx) rats. Bilateral ovariectomy or sham surgery was performed on 16-week-old female Sprague-Dawley rats. Eight weeks after surgery basal Sham and Ovx rats were killed to establish baseline cortical bone values before the initiation of treatment with rhIGF-I/IGFBP-3 complex. At that time, Ovx rats had increased body weight and body fat mass with reduced femoral BMC and BMD relative to basal Shams. Bone formation rates in Ovx rats were increased on both cortical envelopes relative to time-matched controls. The thickness of the inner lamellar bone layer and average cortical width were reduced due to increased endocortical erosion. A similar ratio between Sham and Ovx rats in body mass and composition and femoral BMC and BMD continued throughout the experiment. Sixteen weeks after surgery bone formation rates at both cortical envelopes in Ovx rats were reduced relative to Shams, but endocortical erosion remained high causing a further decrease in thickness of the inner lamellar layer. As a result of periosteal bone modeling. Ovx rats exhibited a larger femoral cross-sectional area and periosteal perimeter, as well as a thicker outer lamellar layer. Newly deposited periosteal bone increased ultimate torque values in the Ovx rats relative to Shams at 16 weeks. Treatment of Ovx rats with the rhIGF-I/IGFBP-3 complex increased body weight, lean body mass, and femoral BMC and BMD.(ABSTRACT TRUNCATED AT 250 WORDS)

Differentiation of rat mesenchymal cells by cartilage-inducing factor: Enhanced phenotypic expression by dihydrocytochalasin B

The role of cell shape in chondrogenesis was studied by using rat mesenchymal cells cultured with cartilage-inducing factor (CIF). Here we report that enhanced expression of chondroblastic markers by induced cells was attained by culturing cells in monolayer in the presence of dihydrocytochalasin B (DHCB). This effect was optimal at 3 microM DHCB and was apparent after 3 days in culture. Mesenchymal cells cultured with DHCB alone exhibited no detectable increase in cartilage proteoglycan synthesis, whereas cells cultured with 3 microM DHCB and 0.1 nM CIF showed a 4-5 fold increase in proteoglycan synthesis. When cells were cultured with CIF alone on plastic, only small increases in proteoglycan synthesis were observed. Cells cultured with CIF in monolayer and then transferred to a permissive environment (either agarose or cultured with DHCB) showed enhanced synthesis of chondroblastic proteins. These results suggest that expression, but not induction, of a chondroblastic phenotype by CIF is inhibited by growth in monolayer. The altering of cell shape with DHCB releases that inhibition.

Bone morphogenetic proteins (BMP‐2 and BMP‐3) induce the late phase expression of the proto‐oncogene c‐fos in murine osteoblastic MC3T3‐E1 cells

Here we report that bone morphogenetic proteins 2 and 3 (BMP‐2 and BMP‐3) induced marked expression of c‐fos mRNA in a biphasic manner, i.e. the late phase (48 to 60 h) as well as the immediate‐early phase (0.5 h), in murine osteoblastic MC3T3‐El cells in vitro. The BMP‐induced late phase c‐fos gene expression was temporally associated with the onset of marked expression of the genes for osteocalcin and alkaline phosphatase, differentiation markers of mature osteoblasts. In contrast, none of TGF‐β1, 10% FBS, IGF‐I and IGF‐II, which induced only the immediate‐early c‐fos mRNA expression, stimulated the expression of osteocalcin and alkaline phosphatase genes. These data suggest that in osteoblasts BMP‐2 and BMP‐3 induce the late phase expression of c‐fos, which may play a role in transcriptional activation of the genes involved in differentiation of osteoblasts.

A fragment of the hypophosphatemic factor, MEPE, requires inducible cyclooxygenase‐2 to exert potent anabolic effects on normal human marrow osteoblast precursors

MEPE, 56.6 kDa protein isolated from tumors associated with hypophosphatemic osteomalacia, increases renal phosphate excretion and is expressed in normal human bone cells. AC‐100, a central 23‐amino acid fragment of MEPE, contains motifs that are important in regulating cellular activities in the bone microenvironment. Thus, we assessed in vitro effects of AC‐100 on multipotential normal human marrow stromal (hMS) cells that have the capacity to differentiate into mature osteoblasts. Proliferation was quantified by [H3]thymidine uptake and cell counting and differentiation by the levels of mRNA for the α2‐chain of type I procollagen (COL1A2), alkaline phosphatase (AP), and osteocalcin (OC) measured using real time reverse transcriptase PCR (RT‐PCR) and by the formation of mineralized nodules. AC‐100 increased proliferation by 257u2009±u200989% (Pu2009<u20090.005), increased gene expression of COL1A2 by 339u2009±u200985% (Pu2009<u20090.005), AP by 1,437u2009±u200940% (Pu2009<u20090.001), and OC by 1,962u2009±u2009337% (Pu2009<u20090.001). In addition, it increased mineralized nodule formation by 81u2009±u200914% (Pu2009<u20090.001) in a dose‐ and time‐dependent fashion. In equimolar dosages, the parent compound, MEPE, had the full activity of the AC‐100 fragment. AC‐100 elicited a comparable response to both IGF‐I and BMP‐2 with respect to proliferation and differentiation of hMS cells. Using gene expression microarray analysis, we demonstrated that AC‐100 increased (by ∼3‐fold) the mRNA for cyclooxgenase‐2 (COX‐2), an inducible enzyme required for prostaglandin synthesis. Moreover, NS‐398, a specific inhibitor of COX‐2 action completely blocked AC‐100‐induced increases in proliferation and differentiation. Thus, AC‐100 has potent anabolic activity on osteoblast precursor cells in vitro and these effects require the induction of COX‐2.

Bone Induction and Transforming Growth Factor-β

In the mid 1960s. Marshall Urist first reported that implantation of demineralized bone particles can induce new bone formation at an ectopic site. It was later shown that this bone-inducing activity can be extracted from bone with dissociative agents such as guanidine-HCl: suggesting that this activity is due to an osteoinductive factor or factors. This biological response has been characterized by Reddi and others?.3 Bone formation follows an endochondral pathway, which includes chemotaxis of PMNs and macrophages, followed by fibroblastic mesenchymal cell recruitment, a differentiation of these cells into chondroblasts, and eventual bone formation complete with all associated marrow elements. Our interest in TX3F-p began with our efforts to isolate osteoinductive factors from bone. Initial studies using an in vitro chondrogenic assay led to the purification of cartilage-inducing factors A and B? These factors are now also known as TGF-fll and TGF-p2.5,6 In our in vitro assay, these factors induce embryonic rat mesenchymal cells to express a chondroblastic phenotype when cultured in agarose gels? At subcutaneous sites in vivo, however, they elicit the well known fibrotic response without any evidence of cartilage or bone formation. We have consistently observed that the cellular responses to demineralized bone particles and crude bone extracts, during the initial stages of bone induction, very closely resemble those of E F p l and XiF-p2 three to four days postimplantation. These observations as well as the in vitro chondrogenic activity of the TGF-(3s prompted us to investigate their potential role in osteoinduction.