[Frontiers in Bioscience 3, d781 -794, August 1, 1998] |
FGF SIGNALING IN SKELETAL DEVELOPMENT
Michael C. Naski and David M. Ornitz
Department of Molecular Biology and Pharmacology, Washington University School of Medicine, Campus Box 8103, 660 S. Euclid Ave., St. Louis, MO 63110
Received12/22/98 Accepted 2/17/98
5. INTRAMEMBRANOUS BONE FORMATION
5.1. FGFs and intramembranous bone formation
Calvarial bone forms directly from mesenchymal cells derived from neural crest (120). These cells migrate to predetermined sites of the embryo where they condense into a multilayered membrane (121). Under the appropriate morphogenetic signals, osteogenesis begins within the core of this membrane. The instructed cells differentiate into osteoblasts which in turn secrete and initiate the mineralization of a matrix rich in type I collagen. Ossification begins at predefined sites of the membrane and radiates outward. In the calvarium, separate osteogenic fronts meet at the cranial sutures. During early post-natal life these sutures remain patent, allowing the cranial vault to grow and expand to accommodate the enlarging brain. Growth of the calvarial bones occurs through the proliferation and differentiation of osteoblasts and the deposition of bone matrix at the margins of the suture. Altered signaling within the regulatory pathways controlling osteoblast growth and differentiation in the calvarium results in dysmorphic facial features. For example, premature closure of the sutures (craniosynostosis) results in characteristic craniofacial features, such as those of the hereditary craniosynostosis.
FGFs have potent effects on the survival, proliferation and differentiation of osteoblasts and their precursors. FGFs have been shown to stimulate the proliferation of cells derived from fetal rat calvarium (122, 123). Cultured rat osteosarcoma cells and osteoblastic MC3T3 cells also proliferate in response to FGFs (124, 125). Additionally, apoptosis of osteoblasts is inhibited by FGF (126). Several studies have shown that FGF inhibits the differentiation of cultured osteoblasts, as evidenced by the inhibition of matrix mineralization, alkaline phosphatase activity and osteocalcin gene expression (124, 127). Concomitantly, an increase in the expression of interstitial collagenase and tissue inhibitor of metalloprotease was observed (128). Interestingly, others have observed that the expression of osteocalcin can be activated by FGF (129-132). These differences in the profiles of gene activation may reflect the duration of the treatment with FGF, the responsiveness of the cells to FGF, or the specific developmental stage of the cells. Prolonged in vitro stimulation with FGF may result in the synthesis of collagenases that degrade the surrounding matrix, a feature essential to the fate of many cell lineages. FGF receptors are subject to degradation by matrix proteases (133). Thus, proteolysis could alter osteoblast differentiation by directly affecting FGF receptor signaling. Additionally, the response of cells to FGF may be dependent on the stage of differentiation of the cells; stimulation of mesenchymal precursors may recruit additional cells to the osteoblast lineage, while prolonged or late treatment may cause reversion to an undifferentiated phenotype. Intravenous administration of FGF 1 or 2 in rats results in enhanced endosteal bone formation (12, 13). Thus, prolonged in vivo treatment with FGF does not irreversibly inhibit programs for osteoblast differentiation. Local factors such as the availability of glycosaminoglycans that regulate the activity of FGF (44) or the presence of other regulators of bone growth, such as TGF-beta (13) or BMPs may also modulate the in vivo activities of FGF.
5.2. FGF receptors and intramembranous bone formation
The important role for FGF signaling in osteoblast differentiation, and in particular the control of the development of cranial sutures, is highlighted by the human craniosynostoses that result from mutations in FGF receptors. Several mutations in FGF receptor 2 that cause Crouzon syndrome are known to be gain of function mutations (97, 134), and it is likely that other mutations causing craniosynostoses are activating as well. This observation suggests that FGF receptor activity within the cranial suture directs an anabolic signal for osteoblast differentiation. In fact, recent experiments studying the relationship of FGF receptor 2 expression and osteoblast differentiation suggest that the two events are linked (135). These studies showed that within the cranial sutures FGF receptor 2 expression is localized to pre-osteoblastic mesenchyme. FGF 2, a potent ligand for FGF receptor 2, is found at highest levels at sites of osteoblast differentiation. The authors speculate that the high levels of FGF 2 strongly activate FGF receptor 2 resulting in accelerated osteoblast differentiation and the down regulation of receptor expression. Consistent with this, studies using calvarial tissues derived from fetuses with Apert syndrome demonstrated enhanced mesenchyme condensation and bone formation (136). It is interesting to note once again that as with the distal extremities, the phenotypic effects of the FGF receptor mutations are localized. During the complex developmental program that creates the calvarial bones, the effects of FGF receptor mutations are localized to the suture. Its is likely that a similar developmental program to that which initiates the differentiation and ossification of the calvarial bones is also responsible for the growth and differentiation of these bones at the suture margin. Several of the FGF receptor mutations responsible for premature ossification of the sutures have been demonstrated to activate the FGF receptor. The similar phenotype seen with other mutations suggests that they may be gain of function mutations as well. This suggests that the early, very rapid events that lay-down the calvarium are insensitive to an activated FGF receptor. Perhaps because these early developmental steps, which occur very rapidly, operate under conditions of excess ligand. Therefore, an activating mutation in the receptor would have no consequence. However, later developmental steps, such as growth at the suture margin, occur more slowly and may function under conditions of limiting ligand, matching the amount of ligand to the desired degree of growth. In this case, an activated receptor would have profound consequences.
Mutations in FGF receptors 1, 2, and 3 cause numerous human craniosynostoses (figure 3). This observation suggests some overlap in the expression and function of the FGF receptors. Indeed, the same clinical syndrome (Pfeiffer syndrome) can result from mutations in FGF receptor 1 or FGF receptor 2. The receptors are clearly not completely redundant, however, because the same proline to arginine substitution occurring in an absolutely conserved region of FGF receptors 1, 2 and 3 results in distinct clinical syndromes, Pfieffer syndrome, Apert syndrome and a nonsyndromic craniosynostosis, respectively. It is particularly interesting that a few of the mutations in FGF receptor 3 have been found to cause craniosynostoses. While all the mutations described in FGF receptor 1 and 2 affect craniofacial development, most of the mutations in FGF receptor 3 cause dwarfing conditions, implying a primary role for FGF receptor 3 in endochondral rather that intramembranous bone growth. The two mutations in FGF receptor 3 that affect craniofacial development result from amino acid substitutions in the transmembrane domain or linker region between Ig-like folds II and III. Both of these domains are thought to be important for receptor dimerization (74, 137), raising the interesting possibility that these mutant receptors may exert their effects by forming a heterodimer between wild type FGF receptor 2 and the mutant FGF receptor 3, thereby activating FGF receptor 2 signaling pathways. Precedence exists for the formation of receptor heterodimers in that a truncated, dominant negative FGF receptor 1 can inhibit signaling through multiple FGF receptors (137).