Connexins as Active Regulators of Signal Transduction in Bone

Abstract

Intercellular communication by gap junctions plays an important role in achieving peak bone mass and determining bone quality. Deletion of the gap junction protein connexin43 (Cx43) in mice impairs bone quality (osteopenia), and results in a cortical structure resembling aging and disuse. Although it is clear that Cx43 plays a role in osteoblast function and bone mineralization, the molecular mechanism underlying the specific roles of Cx43 are not well defined. Contrastingly, osteoblasts also express connexin45 (Cx45) and its role in osteoblasts has not been extensively studied and nothing is known about its role in skeletal tissue in vivo. Cx43 and Cx45 form gap junctions with different molecular permeability, or pore sizes (<1kD <0.3kD) and seem to have opposing functions in bone biology. While gap junctions are generally thought of as passive conduits for small molecules to be shared between cells, growing evidence indicates that connexins actively contribute to downstream signaling. Modulation of Cx43 expression impacts PKCδ, ERK1/2 and β-catenin signaling pathways in vitro, which influence osteoblast gene expression and function. The regulation of signal transduction downstream of intercellular communication may involve assembly of a specific subset of effectors to the gap junction. The work presented in this thesis examines the influence of the C-terminus (CT) of connexins on osteoblast signaling and function through physical interactions with effectors like β-catenin, ERK(1/2), CREB, and PKCδ, both in vivo and in vitro. The central hypothesis of this work is that in addition to having distinct molecular permeability, each connexin can assemble a unique "interactome" of locally recruited signaling machinery that can affect downstream signaling from the gap junction and ultimately, bone cell function. We characterized the skeletal phenotype of a Cx43 truncation mouse model. The absence of the Cx43 CT in mice resulted in an osteopenic skeletal phenotype analogous to osteoblast conditional deletion of the entire Cx43 gene including cortical thinning and increased cross-sectional area, defective signaling, reduced gene expression, and altered collagen processing. These data imply that Cx43-containing gap junctions not only exchange signals, but also recruit signaling machinery to the Cx43 CT domain to optimally affect cell signaling, cell function, and bone acquisition. We also determined the requirement for both pore permeability (pore function) and signal effector protein recruitment (tail function) to Cx43 and Cx45 in osteoblast signaling and function using chimeric constructs composed of portions of Cx43 and Cx45. By luciferase reporter assays, western blot analysis, and qPCR, we found that, in general, both the Cx43 pore and Cx43 tail are required for optimal osteoblast signaling and gene expression, as Cx45 overexpression was ineffective or inhibitory in all aspects, while Cx43 overexpression stimulated signaling and gene expression. The chimeras largely mimicked Cx45 or had an intermediate effect. We also further defined the interactomes of Cx43 and Cx45 to determine their requirement for affecting osteoblast function. Through co-immunoprecipitation, our data shows binding of ERK(1/2), CAMKII, and β-catenin to Cx43 but not to Cx45, which is consistent with the idea of differential interactomes. In total, these data imply that each connexin can differentially regulate downstream signaling and gene expression from the gap junction by local recruitment of different signaling effector molecules to each connexin's CT in order to affect bone cell function and bone modeling and remodeling.