The integration of cellular adhesion with transcriptional outputs of the Wnt signaling pathway has long been an important topic of study.1 Previous work has focused on the dual functionality of β-catenin, in adhesion and in transcriptional activation downstream of Wnt ligands, as the key nexus in mediating this integration. New work by Haÿ and colleagues2 presented in this issue of the Journal of Bone and Mineral Research (JBMR) has identified another point of interaction. The authors also provide evidence that manipulating this interaction can have functional consequences in regulating bone mass.
The Wnts are a family of 19 mammalian, cysteine-rich glycoproteins that play important roles in many developmental processes. Alterations in components of the Wnt pathway are among the most common changes associated with human disease.3 Low-density lipoprotein (LDL)–related receptors 5 and 6 (Lrp5 and Lrp6) are members of the LDL receptor family that specifically bind to Wnts in a heteromeric complex. This complex involves the simultaneous binding of Wnts to a member of the Frizzled family of seven-transmembrane receptors and to either Lrp5 or Lrp6.4 Formation of this receptor complex results in the phosphorylation of the carboxy terminus of Lrp5/6, creating a binding site for the Axin protein.5 The binding of Axin to phospho-Lrp5 or -Lrp6 leads to inhibition of the β-catenin destruction complex, which is composed of Axin, the adenomatous polyposis coli (Apc) protein, and glycogen synthase kinase-3 (GSK3), a serine/threonine protein kinase. GSK3 normally phosphorylates β-catenin, targeting it for ubiquitin-dependent proteolysis. When the destruction complex is inhibited, β-catenin increases in the cytoplasm and subsequently translocates to the nucleus, binding to members of the lymphoid enhancer factor (LEF)/T cell factor (TCF) family of DNA binding proteins and activating transcription from target gene promoters.
Several genetic alterations that affect the Wnt/β-catenin signaling pathway are causally linked to changes in bone mass. For example, patients homozygous for inactivating mutations in LRP5 develop osteoporosis pseudoglioma, a syndrome associated with extremely low bone mass at a young age.6 Patients carrying homozygous inactivating mutations in the sclerostin gene (SOST) develop high bone mass, and these inactivating mutations are linked to the inability to bind to and down-regulate signaling from LRP5.7–9 Finally, point mutations in LRP5 that decrease the ability of the protein to bind to SOST or to the Wnt pathway inhibitor Dickkopf-related protein 1 (DKK-1) also produce dramatically increased bone mass.10, 11 These observations have focused a great deal of effort on targeting this pathway for treatment of low bone mass, and many such approaches are in various stages of clinical development.12
Cadherins are a class of type-1 transmembrane glycoproteins that form Ca2+-dependent trans-interactions between adjacent cells.13 They were first identified almost 25 years ago as molecules involved in cellular adhesion or early developmental processes. In the late 1980s, it was demonstrated that three proteins with molecular weights of 102, 88, and 80 kDa (named α, β, and γ-catenin) could be coimmunoprecipitated with E-cadherin (also known as uvomorulin).14 Over the last two decades, our understanding of the functions of these proteins has grown dramatically.13 Extracellular association of cadherin molecules leads to the association of β-catenin with the cytoplasmic portion of the cadherin. This association, through interactions with α-catenin, then mediates reorganization of the actin cytoskeleton. The overall effect of these interactions is to link extracellular adhesion to the actin cytoskeleton. γ-Catenin (also known as plakoglobin15) can interact with cadherins and also plays a central role in mediating desmosomal adhesion by virtue of its interaction with desmoglein. Subsequent work has found that cadherins form a large family of proteins that can be classified into four distinct groups: classical, desmosomal, protocadherins, and unconventional.16
Several studies have identified N-cadherin as being of particular interest in the context of bone development. For example, N-cadherin is abundantly expressed in osteogenic cells and is involved in osteochondral differentiation.17–20 In addition, osteoblast-targeted overexpression of dominant-negative N-cadherin—a truncated N-cadherin lacking the extracellular domain but retaining its cytoplasmic β-catenin binding site—delays the acquisition of peak bone mass. This delay is possibly due to inhibition of Wnt signaling, because coexpression of constitutively active β-catenin reverses the delay in osteoblast differentiation.21 Furthermore, overexpression of N-cadherin inhibits osteoblast proliferation and survival, and targeted expression of N-cadherin in osteoblasts results in defective osteoblast function and osteopenia in vivo.22, 23
The demonstration by several laboratories in the early 1990s that β-catenin and plakoglobin were homologues of the Drosophila segment polarity gene armadillo further intensified interest in the interactions between cadherins and β-catenin.24 This was largely driven by the fact that β-catenin plays a key role in mediating transcriptional activation downstream of Wnt-induced cellular signaling.25 This dual functionality of β-catenin provides a potential mechanism for integrating cellular adhesion with growth and differentiation control. The importance and complexity of this regulation is highlighted by the fact that several receptor tyrosine kinases can phosphorylate residues in the cadherin/catenin complex to alter the stability of these interactions, potentially regulating the amount of β-catenin in the cytoplasmic pool that might be available for transcriptional control of Wnt signaling. Several models have emerged for how this integration is modulated at the molecular level.1 For example, by keeping β-catenin tethered to the cell membrane, cadherins could reduce the available cytosolic pool of β-catenin and inhibit Wnt/β-catenin signaling. It is also plausible that cadherins retain β-catenin at the cell membrane as a readily available pool for nuclear translocation and activation of transcriptional targets.
New work by Haÿ and colleagues2 presented in this issue of JBMR highlights another point of integration between cadherins and Wnt signaling and the role of this interaction in controlling bone homeostasis. Their work focused on the role of N-cadherin in regulating Lrp5. In a previous manuscript, the authors proposed that, in addition to sequestrating β-catenin at the cell membrane, N-cadherin is able to inhibit Wnt/β-catenin signaling through a direct interaction with Lrp5.22 This complex also included the Axin protein, and its functional relevance was suggested by the fact that transgenic mice overexpressing N-cadherin in osteoblasts have increased β-catenin degradation, reduced bone formation, and delayed bone mass acquisition. In addition, they found that the last 28 AA of Lrp5 were required to mediate the functional interaction with the cytoplasmic tail of E-cadherin.
In their new work, the authors further investigated the function relevance of the N-cadherin/Lrp5 interaction and identified specific domains that mediate the association. They found that an N-cadherin protein lacking the last 62 amino acids (AAs) of the cytoplasmic tail (Δ62 N-cad) failed to associate with Lrp5 or the related Lrp6. Δ62 N-cad is no longer able to impair Wnt/β-catenin signaling, suggesting that the inhibitory effect of N-cadherin on Wnt/β-catenin signaling and osteoblast differentiation depends on its interaction with Lrp5. To extend these observations, they synthesized a peptide corresponding to the last 28 AAs of Lrp5, which mediates the interaction with N-cadherin, and found that this peptide was sufficient to mediate a physical interaction with N-cadherin. Furthermore, they confirmed the functionality of this 28-AA peptide by showing that it abrogated the N-cadherin/Lrp5/6 interaction and could bind concurrently to N-cadherin and Axin. Consistent with this, exposure of MC3T3-E1 cells to the 28-AA fragment increased β-catenin protein levels and stimulated TCF-dependent transcriptional activation in the presence of Wnt3a. Importantly, in vitro treatment of MC3T3-E1 cells with this peptide increased markers of bone formation and matrix mineralization, and also enhanced cellular proliferation and survival. Finally, the authors found that subcutaneous injection of the 28-AA peptide (dissolved in SAINT-PhD, a cationic pyridinium amphiphile and helper lipid) resulted in increases in the matrix appositional rate, mineralizing surface, and bone formation rate in mouse calvarial bone. This anabolic effect was similar in magnitude to that induced by local injection of bone morphogenic protein 2 (BMP2).
The article in this issue by Haÿ and colleagues2 identifies a novel mechanism by which activation of the Wnt/β-catenin pathway can be influenced and further emphasizes the complex integration between adherens junctions and Wnt/β-catenin signaling. It will be of interest to evaluate whether the binding affinity between the last 62 AAs of N-cadherin and the 28-AA segment of Lrp5 is modified by posttranslational modifications. For example, Src family kinases can phosphorylate Tyr-860 within N-cadherin (which is within its last 62 AAs) and change functional outcomes in cell migration.26 Although the initial focus of this work has been on the effect of the N-cadherin/Lrp5 interaction on Wnt/β-catenin signaling, it should be of great future interest to examine how this interaction may affect cellular adhesion, which would further integrate Wnt signaling, adhesion, and the structure of the cytoskeleton.
Finally, given the widespread expression of N-cadherin, it will be of interest to examine the physiological relevance of this interaction with Lrp5 or Lrp6 in extraskeletal tissues, particularly its role in neural tissues.27 Such information could be important in determining whether therapies designed to disrupt this interaction and thus increase bone mass will have deleterious side effects.