Thomas A. Linkhart

Growth factors for bone growth and repair: IGF, TGFβ and BMP

Abstract Current research is reviewed regarding the actions of three growth factor systems on bone formation: insulin-like growth factors (IGFs), transforming growth factor-βs (TGFβs), and bone morphogenetic proteins (BMPs). Each growth factor family consists of multiple related growth factor genes. TGFβs and BMPs 2–7 are subfamilies of a larger TGFβ superfamily. IGFs, TGFβs and BMPs are produced by osteoblasts and other bone cells and affect osteoblast proliferation and differentiation. They are also incorporated into mineralized bone matrix and retain activity when extracted from bone. Various hormones, growth factors, and mechanical stress influence bone cell production of IGFs, TGF βs, and BMPs. Thus these growth factors may function in local regulation of bone formation. Currently there is much interest in the function of IGF binding proteins, which are also produced by bone cells, in regulating IGF activities in bone. Recently, mechanisms for activation of the TGFβ serine/threonine kinase receptors have been investigated, and receptors for BMPs have been identified which are structurally related to TGFβ receptors. In vivo studies are discussed which demonstrate the applicability of IGFs, TGFβs and BMPs to increasing bone formation systemically, promoting fracture healing and inducing bone growth around implants.

Primary structure of human skeletal growth factor: homology with human insulin-like growth factor-II.

Human skeletal growth factor (human SGF) extracted from human bone has been purified to homogeneity by hydroxyapatite chromatography and gel filtration under dissociative conditions followed by FPLC heparin-Sepharose affinity chromatography and reverse phase HPLC. Human SGF was homogeneous except that in each preparation about 30% of SGF molecules lacked the N-terminal alanine. 75% of the human SGF sequence has been determined. The amino acid sequences of the N-terminal 20 amino acids and of several tryptic fragments were identical to the corresponding sequences of human insulin-like growth factor-II (IGF-II) purified from serum. However, since the C-peptide (variable region) of human SGF has not yet been sequenced, we cannot conclude that SGF is identical to IGF-II. Comparison of the amino acid sequence of human SGF with that of IGF-II variants that have been described in the literature revealed that human SGF is not one of the known IGF-II variants. IGF-I was also found in human bone extract but was several-fold less abundant than SGF/IGF-II. The relative abundance of SGF/IGF-II and IGF-I in bone corresponded to the relative rates of production of these two mitogens by human bone cells in vitro. Regarding the physiological significance of IGF-II in bone, previous studies on the biological actions of SGF in vitro suggest that this growth factor can have both paracrine and autocrine functions on cells of the osteoblast line. In addition, we have proposed the concept that SGF is a mediator of the coupling of bone formation to bone resorption, an important bone volume regulatory mechanism. In as much as SGF is very similar (if not identical) to IGF-II, it seems likely that these proposed regulatory functions of SGF in bone are attributable to IGF-II.

Evidence that IGF-binding protein-5 functions as a growth factor

Recent studies support the concept that IGF-binding protein-5 (IGFBP-5) stimulates bone formation, at least in part, via IGF-independent mechanisms. To evaluate this hypothesis further, we evaluated in vitro and in vivo effects of IGFBP-5 on bone formation parameters using the IGF-I knockout (KO) mouse. Treatment of serum-free cultures of osteoblast clones derived from IGF-I KO mice with recombinant human IGFBP-5 increased both proliferation and alkaline phosphatase (ALP) activity in a dose-dependent manner, an effect comparable to that seen with IGF-I. IGF-II levels from media conditioned by osteoblasts derived from IGF-I KO mouse were below those detectable by RIA. To eliminate possible actions of IGF-II, if any was produced by osteoblasts derived from IGF-I knockout mice, the IGFBP-5 effect was studied in the presence of exogenously added IGFBP-4, a potent inhibitor of IGF-II actions in bone cells. Addition of IGFBP-4 blocked IGF-I- but not IGFBP-5-induced cell proliferation in osteoblasts derived from IGF-I knockout mice. Consistent with in vitro results, a single local injection of IGFBP-5 to the outer periosteum of the parietal bone of IGF-I KO mice increased ALP activity and osteocalcin levels of calvarial bone extracts. The magnitudes of IGFBP-5-induced increases in ALP and osteocalcin in parietal bone extracts of IGF-I KO mice were comparable to those seen in C3H mice. In contrast to IGFBP-5, local administration of IGFBP-4 had no significant effect on bone formation in C3H and IGF-I KO mice. These results provide the first direct evidence to our knowledge that IGFBP-5 functions as a growth factor that stimulates its actions in part via an IGF-independent mechanism.

Circulating and skeletal insulin-like growth factor-I (IGF-I) concentrations in two inbred strains of mice with different bone mineral densities

Recent work has demonstrated differences in femoral bone mineral density between two common inbred strains of mice, C3H/HeJ (C3H) and C57BL/6J (B6), across a wide age range. To investigate one possible mechanism that could affect acquisition and maintenance of bone mass in mice, we studied circulatory and skeletal insulin-like growth factor-I (IGF-I) and femoral bone mineral density (F-BMD) by pQCT in C3H and B6 progenitor strains, as well as serum IGF-I obtained from matings between these two strains and mice bred from subsequent F1 intercrosses (F2). Serum IGF-I measured by radioimmunoassay was more than 35% higher in virgin progenitor C3H than virgin B6 at 1, 4, 8, and 10 months of age, and in 8-month-old C3H compared with B6 retired breeders (p < 0.001). In the progenitors, there was also a strong correlation between serum IGF-I and serum alkaline phosphatase (r = 0.51, p = 0.001). In the 4 month F1 females IGF-I levels and F-BMD were intermediate between C3H and B6 progenitors. In contrast, groups of F2 mice with the highest or lowest BMD also had the highest or lowest serum IGF-I (p = 0.0001). IGF-I accounted for > 35% of the variance in F-BMD among the F2 mice. Conditioned media from newborn C3H calvarial cultures had higher concentrations of IGF-I than media from B6 cultures, and cell layer extracts from C3H calvariae exhibited greater alkaline phosphatase activity than cultures from B6 calvarial cells (p < 0.0001). The skeletal content of IGF-I in C3H tibiae, femorae, and calvariae (6-14 weeks of age) was also significantly higher than IGF-I content in the same bones of the B6 mice (p < 0.05). These data suggest that a possible mechanism for the difference in acquisition and maintenance of bone mass between these two inbred strains is related to systemic and skeletal IGF-I synthesis.

Estradiol stimulates invitro the secretion of insulin-like growth factors by the clonal osteoblastic cell line, UMR106

UMR106 cells, a rat osteosarcoma derived clonal line, secreted insulin-like growth factors (IGF) in vitro. The IGF-II levels corrected for the cell numbers were 7-8 times higher than the IGF-I levels in the medium. Both growth factors were higher by 4-5 fold in medium conditioned by rapidly growing cells than in medium conditioned by confluent cells. The addition of 17-beta-estradiol (E) to the culture medium was associated with a statistically significant increase in the IGF concentrations. This increment was metabolite specific, not occurring with 17-alpha-E, the inactive epimer of E. 1,25(OH)2D3 also increased the IGF-I concentration but prior treatment with E blocked the response to 1,25(OH)2D3, demonstrating antagonistic actions of these two hormones on IGF secretion by osteoblast-like cells.

Insulin-like growth factor-binding protein 5 (IGFBP-5) interacts with a four and a half LIM protein 2 (FHL2)

Recent studies using insulin-like growth factor I (IGF-I) knockout mice demonstrate that IGF-binding protein (IGFBP)-5, an important bone formation regulator, itself is a growth factor with cellular effects not dependent on IGFs. Because IGFBP-5 contains a nuclear localization sequence that mediates transport of IGFBP-5 into the nucleus, we propose that IGFBP-5 interacts with nuclear proteins to affect transcription of genes involved in bone formation. We therefore undertook studies to identify proteins that bind to IGFBP-5 using IGFBP-5 as bait in a yeast two-hybrid screen of a U2 human osteosarcoma cDNA library. Five related clones that interacted strongly with the bait corresponded to the FHL2 gene, which contains four and a half LIM domains. Co-immunoprecipitation studies with lysates from U2 cells overexpressing FHL2 and IGFBP-5 confirmed that interaction between IGFBP-5 and FHL2 occurs in whole cells.In vitro interaction studies revealed that purified FHL2 interacted with IGFBP-5 but not with IGFBP-3, -4, or -6. Northern blot analysis showed that FHL2 was strongly expressed in human osteoblasts. Nuclear localization of both FHL2 and IGFBP-5 was evident from Western immunoblot analysis and immunofluorescence. The role of FHL2 as an intracellular mediator of the effects of IGFBP-5 and other osteoregulatory agents in osteoblasts will need to be verified in future studies.

Alkaline phosphatase levels and osteoprogenitor cell numbers suggest bone formation may contribute to peak bone density differences between two inbred strains of mice

Previous studies have shown that C3H/HeJ (C3H) mice have higher peak bone density than C57BL/6J (B6) mice, at least in part because of differences in rates of bone resorption. The current studies were intended to examine the alternative, additional hypothesis that the greater bone density in C3H mice might also be a consequence of increased bone formation. To that end, we measured two presumptive, indirect indices of bone formation and osteoblast number in these inbred strains of mice: alkaline phosphatase (ALP) activity in serum, bones, and bone cells; and the number of ALP-positive colony-forming units (CFU) in bone marrow stromal cell cultures. We found that C3H mice had higher serum levels of ALP activity than B6 mice at 6 (118 vs. 100 U/L, p < 0.03) and 32 weeks of age (22.2 vs. 17.2 U/L, p < 0.001). Tibiae from C3H mice also contained higher levels of ALP activity than tibiae from B6 mice at 6 (417 vs. 254 mU/mg protein, p < 0.02) and 14 weeks of age (132 vs. 79 mU/mg protein, p < 0.001), as did monolayer cultures of bone-derived cells from explants of 7.5-week-old C3H calvariae and femora (8.2 times more, p < 0.02, and 4.6 times more, p < 0.001, respectively). Monolayer cell cultures prepared by collagenase digestion of calvariae from newborn and 6-week-old mice also showed similar strain-dependent differences in ALP-specific activity (p < 0.001 for each). Our studies also showed more ALP-positive CFU in bone marrow stromal cell cultures from 8-week-old C3H mice, compared with B6 mice (72.3 vs. 26.1 ALP-positive CFU/culture dish, p < 0.001). A similar result was seen for ALP-positive CFU production at 6 and 14 weeks of age, and the difference was greatest for the CFU that contained the greatest numbers of ALP-positive cells. Because skeletal ALP activity is a product of osteoblasts and has been shown to correlate with rates of bone formation, and because the number of ALP-positive CFU is believed to reflect the number of osteoprogenitor cells, the current data are consistent with the general hypothesis that bone formation may be greater in C3H than B6 mice because of a difference in osteoblast number. Our data further suggest that peak bone density may be greater in C3H mice than B6 mice due to a combination of decreased bone resorption and increased bone formation.

Fibroblast Growth Factor Inducible 14 (Fn14) Is Required for the Expression of Myogenic Regulatory Factors and Differentiation of Myoblasts into Myotubes EVIDENCE FOR TWEAK-INDEPENDENT FUNCTIONS OF Fn14 DURING MYOGENESIS

Fibroblast growth factor-inducible 14 (Fn14), distantly related to tumor necrosis factor receptor superfamily and a receptor for TWEAK cytokine, has been implicated in several biological responses. In this study, we have investigated the role of Fn14 in skeletal muscle formation in vitro. Flow cytometric and Western blot analysis revealed that Fn14 is highly expressed on myoblastic cell line C2C12 and mouse primary myoblasts. The expression of Fn14 was decreased upon differentiation of myoblasts into myotubes. Suppression of Fn14 expression using RNA interference inhibited the myotube formation in both C2C12 and primary myoblast cultures. Fn14 was required for the transactivation of skeletal α-actin promoter and the expression of specific muscle proteins such as myosin heavy chain fast type and creatine kinase. RNA interference-mediated knockdown of Fn14 receptor in C2C12 myoblasts decreased the levels of myogenic regulatory factors MyoD and myogenin upon induction of differentiation. Conversely, overexpression of MyoD increased differentiation in Fn14-knockdown C2C12 cultures. Suppression of Fn14 expression in C2C12 myoblasts also inhibited the differentiation-associated increase in the activity of serum response factor and RhoA GTPase. In addition, our data suggest that the role of Fn14 during myogenic differentiation could be independent of TWEAK cytokine. Collectively, our study suggests that the Fn14 receptor is required for the expression of myogenic regulatory factors and differentiation of myoblasts into myotubes.

Osteoclast formation in bone marrow cultures from two inbred strains of mice with different bone densities.

For the purpose of identifying genes that affect bone volume, we previously identified two inbred mouse strains (C57BL/6J and C3H/HeJ) with large differences in femoral bone density and medullary cavity volume. The lower density and larger medullary cavity volume in C57BL/6J mice could result from either decreased formation or increased resorption or both. We recently reported evidence suggesting that bone formation was increased in vivo and that osteoblast progenitor cells are more numerous in the bone marrow of C3H/HeJ compared with C57BL/6J mice. In the present study, we determined whether osteoclast numbers in vivo and osteoclast formation from bone marrow cells in vitro might also differ between the two mouse strains. We have found that the number of osteoclasts on bone surfaces of distal humerus secondary spongiosa was 2‐fold higher in 5.5‐week‐old C57BL/6J mice than in C3H/HeJ mice of the same age (p < 0.001). Bone marrow cells of C57BL/6J mice cocultured with Swiss/Webster mouse osteoblasts consistently produced more osteoclasts than did C3H/HeJ bone marrow cells at all ages tested from 3.5–14 weeks of age (p < 0.001). Osteoclast formation was also greater from spleen cells of 3.5‐week‐old C57BL/6J mice than C3H/HeJ mice. The distribution of nuclei per osteoclast and the 1,25‐dihydroxyvitamin D3 dose dependence of osteoclast production from bone marrow cells were similar. Osteoclasts that developed from both C57BL/6J and C3H/HeJ marrow cells formed pits in dentin slices. Cultures from C57BL/6J marrow cells formed 2.5‐fold more pits than cultures from C3H/HeJ marrow cells (p < 0.02). We compared the abilities of C57BL/6J and C3H/HeJ osteoblasts to support osteoclast formation. When bone marrow cells from either C57BL/6J or C3H/HeJ mice were cocultured with osteoblasts from either C57BL/6J or C3H/HeJ newborn calvaria, the strain from which osteoblasts were derived did not affect the number of osteoclasts formed from marrow cells of either strain. Together, these observations suggest that genes affecting the bone marrow osteoclast precursor population may contribute to the relative differences in bone density that occur between C3H/HeJ and C57BL/6J mouse strains.

Local ex vivo gene therapy with bone marrow stromal cells expressing human BMP4 promotes endosteal bone formation in mice.

Bone loss in osteoporosis is caused by an imbalance between resorption and formation on endosteal surfaces of trabecular and cortical bone. We investigated the feasibility of increasing endosteal bone formation in mice by ex vivo gene therapy with bone marrow stromal cells (MSCs) transduced with a MLV‐based retroviral vector to express human bone morphogenetic protein 4 (BMP4).