Endochondral bone formation involves a highly coordinated program of chondrocyte differentiation, proliferation, maturation, and hypertrophy. The Bone Morphogenetic Protein (BMP) and Transforming Growth Factor beta (TGFβ) and Activin signaling pathways are regulators of these processes. The importance of the BMP type I receptors had been well documented; however, the physiological role of the BMP type I receptor Activin Like Kinase 2 (ALK2/ACVR1) during the process of mammalian endochondral bone formation was unknown. The ALK2 receptor gained its infamy when activating mutations in it were discovered to cause the devastating disease fibrodysplasia ossificans progressiva (FOP), an autosomal dominant condition that causes paralysis and early death due to ossification of soft connective tissue. Early in vivo and in vitro chick studies revealed that the ALK2 receptor plays an important role in the formation of cartilage and bone. However, the physiological relevance of ALK2 in mammalian skeletal tissues remained unknown. Furthermore, whether ALK2 receptor impacts BMP signaling in the development of skeletal tissues as strongly as and/or has overlapping functions with those of the other BMP Type I receptors ALK3/BMPR1A and ALK6/BMPR1B remained unclear. My studies addressed these unknowns.

Previous studies in the laboratory of Professor Karen Lyons had shown that the main transducers of the BMP signaling pathways in cartilage are the receptor regulated Smads (R-Smads) 1, 5, and 8. This was demonstrated by the observation that cartilage-specific knockouts of all three Smads recapitulated the phenotype seen in mice lacking the BMP receptors ALK3 and ALK6. Smad4, a co-Smad that complexes with R-Smads, was thought to be required for all transcriptional R-Smad activity. However, cartilage specific knockout of Smad4 reported by another group surprisingly rendered a viable mouse. The phenotypic analysis of pre- and postnatal Smad4 mutants was characterized, however, the mechanism by which BMP signaling persisted in the absence of Smad4 remained unclear. The studies in this thesis address the mechanisms underlying this apparent R-Smad-dependent, Smad4-independent signaling. Cartilage specific knockout of Smad4 was found to result in an accumulation of phosphorylated Smads 1/5/8 in the cell that could not be targeted for degradation. Overall these studies showed that the requirement of Smad4 for Smad1/5/8-mediated transcription is, in part, obsolete in systems with persistent BMP signaling such as the developing cartilage growth plate. However, Smad4 appears to be required in events that require temporal, rapid, and potent signaling, like that of limb bud development. Therefore, Smad4 plays a role in regulating the duration and strength of BMP signaling.

This thesis includes one published review and two original articles (one of which has been published). The known mechanisms by which BMP signaling regulates chondrogenesis and cartilage maintenance will be highlighted in Chapter One, “BMP signaling in Skeletogenesis” and Chapter Two “TGFβ signaling in Cartilage Development and Maintenance”. The roles of the BMP type I receptors ALK2/ALK3/ALK6, with a focus on ALK2, will be highlighted in Chapter Three. This work has been published. The novel Smad4-independent, Smad1/5/8-dependent signaling pathway will be the subject of Chapter Four “ Challenging the Dogma of Canonical BMP signaling in the absence of Smad4 during endochondral bone formation”. Next, a short methods paper highlighting the art of skeletal tissue staining will be outlined in Chapter Five “Whole-Mount Skeletal Staining”. This work has been published. Finally, Chapter Six will discuss the overall conclusions and future directions of my thesis project.

The CCN (Cyr61/Ctgf/Nov) family of matricellular proteins integrates the extracellular environment with the intracellular environment, with involvement in many cellular processes such as adhesion, proliferation, mechanical stretch, fibrosis, and more. A large question lurks regarding how these CCNs interact with each other to regulate many normal physiological events, as well as disease pathogenesis. While CCN1 and CCN2 are two of the best-studied CCN family members, it remains elusive whether they have similar roles in endochondral bone formation despite similar patterns of mRNA and protein expression within developing bones. Furthermore, in vivo data have already shown the significance of CCN2 for endochondral bone formation, but a role for CCN1 in chondrogenesis has only been shown in vitro. In order to study whether CCN1 and CCN2 have overlapping functions in endochondral bone formation, we utilized a collagen II/cartilage-selective deletion of Ccn1 and/or Ccn2 to analyze the phenotypes of Ccn1 and Ccn2 single and double mutant mice. A generally more severe skeletal phenotype was evident in the double mutant axial skeletons, implying the possibility that CCN1 and CCN2 may have overlapping functions. Surprisingly, in the appendicular skeletons there was not a more severe double mutant phenotype but rather a phenotype closely resembling that of Ccn2 single mutants, indicating that CCN2 plays a more dominant role within the appendicular skeleton. This supports the new idea that CCN2 is epistatic to CCN1. In addition, CCN1 and CCN2 may work antagonistically to regulate chondrocyte proliferation in the resting zone of tibial growth plates. Also, a novel phenotype was discovered as the double mutant exhibited increased apoptosis in cells at the perichondrial border. These studies will facilitate the understanding of embryonic bone development and how the proteins CCN1 and CCN2 are involved in this process.

Members of the C(Cyr61) C(CTGF) N(Nov) family of matricellular proteins are involved in multiple aspects of embryonic and postnatal development and disease pathologies. CCN2, or connective tissue growth factor, is the second member of the family and has gained importance due to its prominent role in fibrotic disease. CCN2 has been extensively studied both in vitro and in vivo, but its specific roles remain unclear due to discrepancies in cell specific signaling mechanisms and experimental design. The current use of CCN2 blocking antibodies in clinical trials to treat symptoms associated with fibrosis and kidney function during diabetes warrants a clearer understanding of the function of CCN2 during normal physiology in order anticipate side effects. In order to investigate the functions of CCN2 during development, we have further analyzed the Ccn2 global knockout mouse and found that CCN2 plays essential roles in angiogenesis during vascular remodeling and cellular stress during endochondral ossification. CCN2 plays a dual role in vascular remodeling, first through mediating platelet derived growth factor (PDGF) signaling between endothelial cells and pericytes, and second through inducing the secretion of provisional and permanent vascular basement membrane components. Further analysis of the ECM defect observed in the growth plates of Ccn2 mutant mice revealed that the loss of CCN2 results in increased endoplasmic reticulum (ER) stress. Ccn2 mutants also exhibited decreased Nuclear Factor κB (NFκB) and autophagy-mediated cellular survival. Conversely, the overexpression of CCN2 results in attenuated ER stress and increased cellular survival during chemically induced ER stress. These results highlight a novel protective role for CCN2 during chondrocyte differentiation. Taken together, these results demonstrate that CCN2 plays important physiological roles in vivo and these roles should be considered during therapeutic interventions.

Endochondral ossification involves a highly coordinated program of chondrocyte differentiation, proliferation, maturation, and hypertrophy. The bone morphogenetic protein (BMP) and transforming growth factor beta (TGFβ)/activin pathways are important regulators of these processes. The importance of extracellular antagonists as regulators of the duration, intensity and extent of BMP and TGFβ/activin signaling has been defined. For example, mice lacking the BMP receptor antagonist Noggin exhibit cartilage overgrowth concurrent with excessBMP activity. However, very little is known about the roles of intracellular inhibitors of BMP and TGFβ/activin pathways, such as the inhibitory Smad (I-Smad) proteins, Smad6 and Smad7. In vitro studies reveal that I-Smads 6 and 7 can regulate BMP- and TGFβ-mediated effects on chondrocytes. Although in vivo studies in which I-Smads were overexpressed in cartilage have shown that I-Smads have the potential to limit BMP signaling in vivo, the physiological relevance of I-Smad activity in skeletal tissues remains unknown. Furthermore, whether I-Smads impact TGFβ signaling in cartilage during development is unclear.

This thesis includes two reviews (both have been published) and two original articles (one of which has been published). First, the known mechanisms by which BMP signaling regulate chondrogenesis, osteogenesis, and adipogenesis will be highlighted Chapter One, "BMP signaling in Skeletogenesis." Then, the roles of Smad proteins, which are the mediators of the canonical BMP/TGFβ pathways, in regulating skeletal development will be highlighted in Chapter Two, "Smad signaling in skeletal development and regeneration." Finally, the roles of Smad6 and Smad7 in endochondral bone formation will be described in Chapters Three ("Smad6 is essential to limit BMP signaling in cartilage development") and Four ("Smad7 regulates terminal maturation of chondrocytes during cartilage development in mice"). The last chapter will discuss the overall conclusions and future directions of my thesis project.