SEARCH

SEARCH BY CITATION

INTRODUCTION

  1. Top of page
  2. INTRODUCTION
  3. BIOLOGIC FUNCTIONS OF CADHERINS
  4. CADHERINS IN THE SKELETAL SYSTEM
  5. Acknowledgements
  6. REFERENCES

The organization of cells in tissues and organs is controlled by molecular programs that afford cells the ability to recognize their neighbors and the extracellular matrix. Cell—cell and cell—matrix contact and adhesion are mediated by four major groups of molecules: cadherins, immunoglobulin-like molecules, integrins, and selectins.(1 Adhesive interactions are essential in embryonic development, when cell specification and tissue differentiation are dependent on the ability of undifferentiated cells to aggregate and sort out in specific areas of the embryo.(2 Cell adhesion is also involved in a variety of other biologic processes, such as cell polarity, the immune response, the inflammatory process, division and death, tumor progression, and metastases.(3–5 The biologic importance of cell—cell adhesion in certain tissues and organs is underscored by the findings of mutations of cell adhesion molecules in different pathologic conditions, such as pemphigus vulgaris, leukocyte adhesion deficiency,(6 inflammatory bowel disease,(7 and congenital forms of hearing loss and retinal degeneration.(8 Increasing evidence shops that direct cell—cell interactions are also intimately involved in a variety of processes critical for bone development, starting from embryonic cartilage formation to adult bone remodeling. Previous reviews have focused on cell—cell interactions mediated by cell—cell adhesion and gap junctions in these processes.(9–11 This article focuses on the role of cadherins and their interplay with β-catenin signaling for the maintenance and differentiation of skeletal stem cells in the adult bone microenvironment.

BIOLOGIC FUNCTIONS OF CADHERINS

  1. Top of page
  2. INTRODUCTION
  3. BIOLOGIC FUNCTIONS OF CADHERINS
  4. CADHERINS IN THE SKELETAL SYSTEM
  5. Acknowledgements
  6. REFERENCES

Cell—cell adhesion: cadherins and the adherens junction

Cadherins are single-chain integral membrane glycoproteins, which derive their name from their function in mediating calcium-dependent cell—cell adhesion. To date, −80 members of this large superfamily have been cloned, and it is now clear that many tissues express more than one cadherin.(1,12 Cadherins have a molecular mass of −120 kDa and are composed of a long extracellular domain, a single transmembrane domain, and a relatively small cytoplasmic C-terminal tail. Calcium-binding sites are located in the extracellular domain, which is composed of five repeats and confers the ability to bind to the same cadherin on neighboring cells. Based on structural differences, cadherins are classified as type I, which includes, among others, N-, E-, M- and R-cadherin; and type II, comprising cadherin-5 through −12. Another group of cadherins includes those lacking the intracellular tail (i.e., T-cadherin and cadherin-13), whose function is still obscure. This now large family of molecules is sometimes referred to as “classical” cadherins to distinguish them from protocadherins, cadherin-related neuronal receptor, and other structurally related molecules present primarily in the central nervous system.(12

Assembly of the adhesion structure begins with cadherin homodimerization and stabilization of the five extracellular domains into a rigid X-shaped structure, a Ca2+-dependent process (Fig. 1). Such cis-dimers can dock to opposing cadherin dimers through trans-interactions through the first extracellular domain, leading to the formation of a “zipper”-like structure.(13,14 The cytoplasmic tail, highly conserved among cadherins, is important for stabilization of the adhesion structure, which occurs by cadherin lateral clustering and linkage to the cytoskeleton. Such organization is orchestrated by cadherin binding to β-catenin and plakoglobin, which link cadherin molecules to the actin cytoskeleton through α-catenin and other intermediate proteins, including actinin, vinculin, and ZO-1.(15,16 Very recent data indicate that this multiplex structure may not be as rigid as one would think, because α-catenin binding to either the cadherin/β-catenin complex or to actin is not simultaneous,(17 implying that the connection between the adhesion structure and the cytoskeleton is dynamic rather than static.(18 Regulation of adhesion is also controlled by cadherin binding to other proteins, most importantly p120ctn, which is structurally similar to β-catenin and is thought to stabilize cadherin-β-catenin binding.(19,20 The assembly of cadherins and their associated cytoskeletal elements form the junctional structures known as adherens junctions,(2,21 which provide anchorage between two adjoining cells.

thumbnail image

Figure Figure 1. Schematic diagram of the adherens junction. Two symmetrical structures are shown on two opposing cells. Cadherin cis-homodimers form by lateral clustering on one cell, and the first extracellular domains bind to similar homodimers on an adjacent cell (trans-dimers) in a calcium-dependent fashion to form a zipper-like adherence junctional structure. The cadherin complex is stabilized intracellularly by binding to β-catenin and plakoglobin. These in turn bind to α-catenin, and through interaction with other proteins, including α-actinin, ZO-1, vinculin, and others, tether the complex to the actin cytoskeleton. Cadherin binding to p120ctn also contributes to stabilizing the adhesion structure.

Download figure to PowerPoint

Signaling: convergence of Wnt, β-catenin, and cadherin pathways

Among the proteins that stabilize the adherens junction, β-catenin has a particularly important role because of its involvement in the Wnt signaling system and its central role in osteogenic differentiation.(22–24 Canonical Wnt signaling results in β-catenin binding to members of the T cell factor/lymphoid enhancer factor (Tcf/Lef) family to regulate gene transcription. In the absence of Wnt signaling, cytoplasmic β-catenin is targeted for degradation through the ubiquitin/proteasome pathway by both casein kinase I and glycogen synthase kinase-3β (GSK-3β)—mediated serine/threonine phosphorylation.(25,26 Because β-catenin participates in both cell—cell adhesion and Wnt signaling, it represents a mechanism by which cadherins may modulate signaling.(16,27 By keeping β-catenin bound on adhesion structures on the cell surface, cadherins sequester β-catenin from its transcriptionally active pools,(28,29 thus potentially attenuating Wnt signaling.(30 In contrast, reduced abundance of cadherins on the cell surface can in theory make more β-catenin available for transcriptional regulation. At the same time, lower cadherin abundance destabilizes the adhesion complex, with loss of cell—cell adhesion and increased cell movement. The opposite occurs when cadherin—β-catenin binding is stabilized (Fig. 2). Thus, Wnt signaling induces conformational changes of β-catenin that favor assembly into transcriptional complexes, whereas β-catenin association with α-catenin favors cadherin binding and adhesion.(31,32

thumbnail image

Figure Figure 2. Convergence of Wnt and cadherins on β-catenin. Wnt signaling stabilizes β-catenin by inhibition of GSK-3 phosphorylation and release from the axin—APC—GSK3 complex. β-catenin translocates to the nucleus for gene transcription activation. On the other hand, β-catenin binding to α-catenin and to cadherins on the cell surface leads to assembly and stabilization of the adherens junction. Thus, by modulating the abundance and stability of β-catenin nuclear and membrane pools, both Wnt signals and cadherins can affect gene transcription and cell—cell adhesion. Whether the cadherin bound pool represents a reservoir of β-catenin for recruitment into canonical Wnt signaling or just a sequestration of β-catenin away from gene transcription remains unknown.

Download figure to PowerPoint

This model of cadherin/Wnt signaling interactions is widely accepted, but functional alternatives may exist. For example, β-catenin on the cell surface might serve as a reservoir for rapid recruitment of this signaling molecule into transcriptionally active pools in response to continuous Wnt signaling activation. In this scenario, a lower abundance of membrane-bound β-catenin might actually attenuate signaling because less β-catenin is available for mobilization into active pools on Wnt activation. In any case, intersection of cadherin-mediated adhesion and Wnt signaling through β-catenin offers a complex but flexible mechanism of cell regulation,(16,27,33 a mechanism of particular importance in the skeleton.

CADHERINS IN THE SKELETAL SYSTEM

  1. Top of page
  2. INTRODUCTION
  3. BIOLOGIC FUNCTIONS OF CADHERINS
  4. CADHERINS IN THE SKELETAL SYSTEM
  5. Acknowledgements
  6. REFERENCES

Cadherins in bone mass acquisition

A number of in vitro studies have shown that cells of the osteoblastic lineage express two major cadherins, N-cadherin (Cdh2) and cadherin-11 (Cdh11) and that disrupting cadherin function using adhesion inhibiting peptides,(34,35 function blocking antibodies,(36 antisense RNA,(37 or overexpression of dominant negative mutants,(35,38 consistently impairs osteoblast differentiation. More recently, new in vivo genetic models have provided more direct evidence of the biological role of cadherins in skeletal development and bone remodeling. Unfortunately, homozygous Cdh2 null mutation is embryonically lethal(39; however, mice genetically deficient of Cdh11 display minor calcification defects of the cranial sutures and become osteopenic by 3 months of age, an abnormality seen primarily in the trabecular bone and linked to a cell autonomous functional defect of bone-forming cells.(40 The relatively mild phenotype of these animals may be explained by partial redundancy between Cdh2 and Cdh11. Whereas tissue-specific Cdh2 ablation models are being developed, a recent study elegantly showed that ex vivo limb bud cultures from Cdh2 null embryos partially rescued by transgenic expression of E-cadherin were able to undergo cartilage condensation and develop into structured limbs in the absence of Cdh2.(41 Thus, either N-cadherin is dispensable for cartilage development or caderin-11 compensates for lack of N-cadherin. Preliminary results from our group suggest that the latter may be the case. Compound Cdh2 heterozygous/Cdh11 null mice develop a smaller skeleton with thinner cortices of long bones and much more severe osteopenia than single Cdh11 null mice.(42 Intriguingly, these animals exhibit both an osteogenic differentiation defect'consistent with loss of Cdh11(40'and a decrease in the number of colony forming units-fibroblast (CFU-F), suggesting that N-cadherin may be involved in maintaining bone marrow skeletal stem cell (SSC) niches.(42

In an attempt to overcome the partial redundancy of osteoblast cadherins, we developed a transgenic mouse model expressing a dominant negative N-cadherin mutant (NcadΔC) in osteoblasts in vivo using the mouse osteocalcin (Og2) promoter.(43 Because this mutant lacks almost entirely the extracellular domain, it is unable to dimerize and consequently, it disrupts cell—cell adhesion.(44,45 However, because NcadΔC retains the intracellular domain, it may have a sequestering effect on β-catenin when expressed at sufficiently high levels.(11 We found that Og2-NcadΔC mice exhibit a marked delay in the acquisition of peak bone mass, the result of impaired osteogenic differentiation.(43 Interestingly, there was a concomitant increase in percent body fat in these transgenic mice relative to their wildtype littermates, suggesting a shift from osteogenesis to adipogenesis. In fact, the number of adipogenic precursors isolated from bone marrow cells of Og2-NcadΔC mice was higher and the number of osteogenic precursors was lower relative to wildtype cells. Importantly, whereas cell—cell adhesion among osteoblasts was modestly decreased by expression of the dominant negative N-cadherin mutant, β-catenin nuclear localization was notably decreased, with a corresponding increase in membrane localization.(43 Thus, not only does NcadΔC disrupt cell—cell adhesion but it also sequesters β-catenin on the cell surface, and this correlates with a significant decrease of Tcf/Lef-dependent transcriptional activity.(46 That the osteogenic to adipogenic shift in Og2-NcadΔC transgenic mice was at least in part consequent to reduced β-catenin was shown by the rescue, in vitro, of this abnormality by expression of an active β-catenin mutant.(43 Obviously, these results do not exclude that interference with cell—cell adhesion might also be responsible for the osteogenic defect in these mice. Nonetheless, the phenotype of these transgenic mice provides the first in vivo evidence that osteoblast cadherins are involved in bone marrow stromal cell lineage allocation and supports the notion that cadherins influence Wnt/β-catenin signaling in regulating cell differentiation and function.

These new in vivo data can be used to update the “cadherin switch” model of SSC lineage allocation that we(9,47 and others(48 had proposed earlier based on in vitro observations. As shown in Fig. 3, bone marrow stromal cells express a relatively wide repertoire of cadherins, including Cdh2, Cdh11, and R-cadherin, although at low abundance.(47,48 N-cadherin may serve a major role in maintaining SSC in their undifferentiated state, perhaps controlling their self-renewal (i.e., the capacity of dividing giving rise to two daughter cells with the same multipotential capacity as the mother cells [symmetric division] or to a daughter cell that remains a stem cell and another that is committed [asymmetric division]).(49 Evidence for this role of N-cadherin in SSC maintenance is still circumstantial or indirect but very suggestive. Not only is N-cadherin relatively abundant compared with other cadherins in undifferentiated cells,(47,48 but overexpression of Cdh2 prevents full chondrocyte differentiation, even though it favors cell aggregation in micromass cultures.(50,51 Although embryonic chondrogenesis is obviously a different process than osteoblast differentiation in adult bone marrow, these observations suggest that N-cadherin may have a “braking” effect on mesenchymal cell differentiation. Such a conclusion is consistent with our unpublished data indicating that mice heterozygous for Cdh2 null mutation have reduced osteoprogenitors in the bone marrow, despite faster mineralization of calvaria cells. As discussed more in depth later, data from the hematopoietic field also support a role of N-cadherin in regulating stem cell self-renewal. Commitment to osteogenesis is associated with upregulation of Cdh11,(47,52,53 whereas downregulation of Cdh11 (and Cdh2) favors adipogenesis (Fig. 3). Thus, the presence of both N-cadherin and cadherin-11 defines osteogenic commitment. As osteoblast differentiation progresses, Cdh2 is progressively downregulated, and Cdh11 remains as the major osteoblast cadherin. The molecular mechanisms by which cadherins modulate this process are the object of current research. Certainly, cadherins do provide the adhesion structures for cell—cell aggregation and perhaps for guiding migration of committed cells to the areas of active bone remodeling. However, maintaining “appropriate” levels of intracellular β-catenin might well be another mechanism by which cadherins modulate SSC lineage allocation.

thumbnail image

Figure Figure 3. “Cadherin switch” model of SSC lineage allocation. In their undifferentiated stage, SSC express a relatively wide repertoire of cadherins, including N-cadherin (Cdh2), cadherin-11 (Cdh11), and R-cadherin (Cdh4), although at low abundance. SSC are maintained by self-renewal (i.e., production of daughter cells with conserved multipotential capacity). N-cadherin may serve to support SSC self-renewal, thus keeping SSC in an undifferentiated state. Commitment to osteogenesis is associated with upregulation of Cdh11, whereas downregulation of Cdh11 (and Cdh2) favors adipogenesis.

Download figure to PowerPoint

Heterotypic and “orthotypic” cell—cell interactions in bone

The notion that direct cell—cell contact between osteoclast precursors and cells of the stromal/osteoblastic lineage are necessary for induction of osteoclast differentiation has been known to bone biologists for over a decade.(54 We now know that this heterotypic cell—cell interaction is functionally mediated by engagement of RANK with RANKL.(55,56 Importantly, both receptor, RANK, and ligand, RANKL, are membrane bound molecules present in osteoclast precursors and osteoblastic cells, respectively. Thus, close proximity between the two interacting cells is required for initiation and maintenance of the osteoclastogenic signal, but it is unlikely that this interaction alone provides a strong cell—cell anchorage that would allow stable RANK—RANKL engagement and initiation of osteoclastogenic signaling. In fact, earlier work convincingly showed (even before the discovery of RANKL) that osteoclastogenesis could be prevented by inhibiting cadherin-dependent heterotypic (osteoblast/stromal cells—osteoclast precursors) interactions.(57,58 Cadherin-6 was found to be involved in this mechanism, and expression of a dominant negative, adhesion defective cadherin-6 mutant prevented stromal cell support of osteoclastogenesis.(58 More recent data suggest that cadherins expressed by osteogenic cells can also control the expression of RANKL through a β-catenin—dependent mechanism.(46 Considering that E-cadherin may be involved in osteoclast precursor fusion,(57 interference with cadherin-mediated cell—cell interactions becomes a reasonable target for inhibition of osteoclast differentiation.

Two breakthrough studies from two independent investigative teams have generalized the relevance of heterotypic osteoblast/stromal-osteoclast/hematopoietic interactions through cadherins, elegantly showing that osteoblasts are critical in maintaining the hematopoietic stem cell (HSC) niche(59 and that stimulation of osteogenic activity by either expression of a constitutively active PTH/PTH-related peptide (PTHrP) receptor mutant in osteoblasts or by exogenous administration of PTH increases the number of HSCs.(60 In this case, modulation of HSCs was attributed to modulation of Notch signaling, again, a system based on membrane-bound ligand and receptors, thus requiring cell—cell contact for optimal ligand presentation to the receptor. This heterotypic contact is maintained, at least in part, by N-cadherin,(61 because HSCs are located in close proximity to cells lying on the bone surface and both cell types express Cdh2 and β-catenin in vivo.(59 Furthermore, Cdh2 is necessary to maintain stem cells in Drosophila.(62 The role of N-cadherin in HSC niche maintenance has been further supported by the recent demonstration that the balance between HSC self-renewal and differentiation is controlled by c-Myc in a N-cadherin—dependent fashion.(63 Increased c-Myc expression decreases N-cadherin abundance, thus allowing HSCs to escape from the niche and undergo differentiation, whereas loss of c-Myc leads to increased N-cadherin and HSC accumulation into the niche.(63 These results reinforce the idea that N-cadherin contributes to maintain cells in an undifferentiated state but its persistence prevents differentiation. As noted earlier, such notion is consistent with the demonstration that overexpression of Cdh2 prevents progression from precartilage condensation to chondrocyte development, probably by stabilizing cell—cell adhesion or increasing the threshold for Wnt signaling.(51,64 In this new scenario, the concept of osteoblast/stromal cell support of osteoclastogenesis becomes a particular case of a more general mechanism of osteoblast—hematopoietic cell interactions, in which cadherins play a prominent role.

These studies and the results from our dominant negative cadherin transgenic mice suggest that osteoblast control of stem cell niches may also extend to the stromal compartment. As just mentioned, mice overexpressing a constitutively active PTH/PTHrP receptor have increased HSC number.(60 However, in the first few months of life, they also exhibit a major delay in development of a bone marrow cavity because of exuberant bone formation. Later, with continuous bone remodeling, hematopoietic supportive stroma develops, although with a decreased number of SSC.(65 This finding was attributed to a depletion of SSC as a consequence of multiple rounds of bone turnover. Because the promoter used in that model, a 2.3-kb fragment of αI(I) collagen, is expressed strongly only in cells that are already committed to osteogenesis, another possible interpretation is that osteogenic cells modulate the survival and fate of their own precursors, as they do with hematopoietic precursors. Such a mechanism would be consistent with our own data showing an osteogenesis to adipogenesis shift in bone marrow precursors of Og2-NcadΔC transgenic mice.(43 Because in these animals the transgene is only expressed in differentiated osteoblasts, it is reasonable to believe that the lineage allocation defect is mediated by cadherin modulation of cell—cell interactions between differentiated osteoblasts and undifferentiated stromal cells through regulation of cell—cell adhesion and/or β-catenin signaling. We propose this type of cell—cell interaction be called “orthotypic,” to distinguish between osteoblast—osteoblast (homotypic) and osteoblast—hematopoietic cells (heterotypic), because the interaction is between cells of the same lineage but of different phenotypic features.

Model of cadherin modulation of SSC commitment and bone cell differentiation

The data reviewed herein can be interpreted and summarized in a model that defines cadherins' role in bone cell differentiation (Fig. 4). As noted, Cdh2 is expressed at higher levels than any other cadherins in bone marrow stromal cells(47,48 and serves to keep these cells in an undifferentiated state, allowing aggregation in a niche, an organization similar to HSC niches.(61,66 In fact, SSC and HSC may even co-exist in the same microanatomical niches.(67 Extending the proposed mechanism by which N-cadherin serves as mediator of heterotypic interactions between osteoblasts and HSC,(61,63 differentiated osteoblasts also support the SSC niche through N-cadherin (and cadherin-11?)— mediated cell—cell adhesion (orthotypic interactions; Fig. 4, left). Close proximity between osteoblastic cells and SSC may allow engagement of membrane bound ligand-receptor systems (e.g., Notch signaling), which modulate SSC self-renewal and fate. Alternatively, modulation of SSC may occur through indirect, perhaps β-catenin—dependent paracrine mechanisms, and production of extracellular factors that feed back onto SSC. SSC lineage allocation (escape from the niche) requires loss of cell—cell adhesion attended by downregulation of Cdh2 and increased β-catenin signaling. As osteoblast differentiation progresses, cadherin-11 provides the main mechanism of homotypic (among osteoblasts) cell—cell interactions. In an analogous fashion, direct contact between osteoblastic cells and osteoclast precursors (probably mediated by cadherin-6) facilitates RANK—RANKL engagement and osteoclastogenesis (Fig. 4, right).

thumbnail image

Figure Figure 4. Proposed functions of cadherins in cell—cell interactions in the bone microenvironment. Both skeletal and hematopoietic stem cells (SSC, HSC) are organized in niches that are maintained and regulated by direct contact (orthotypic and heterotypic interactions, respectively) with cells on the bone surface (osteoblasts and bone lining cells), through cadherins. Osteoblast differentiation (left) is characterized by the presence of both N-cadherin and cadherin-11. Whereas the latter is progressively upregulated with commitment and differentiation, N-cadherin is downregulated, perhaps to allow escape from the SSC niche. Contact between osteoblastic cells and hematopoietic osteoclast precursors (right), mediated by cadherin-6 or N-cadherin allows osteoclast differentiation through stabilization of RANK—RANKL engagement.

Download figure to PowerPoint

By providing means of direct interactions among cells of different lineages and by modulating β-catenin signaling, cadherins represent fundamental players for stem cell specification and osteoblast differentiation and function. Future work will focus on the mechanisms by which cadherin mediated adhesion and modulation of β-catenin signaling drive bone modeling and remodeling. As we unlock these mechanisms, the potential for development of new therapeutic targets aimed at increasing one's potential for osteogenesis may become apparent.

Acknowledgements

  1. Top of page
  2. INTRODUCTION
  3. BIOLOGIC FUNCTIONS OF CADHERINS
  4. CADHERINS IN THE SKELETAL SYSTEM
  5. Acknowledgements
  6. REFERENCES

This study was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR43470 (to RC).

REFERENCES

  1. Top of page
  2. INTRODUCTION
  3. BIOLOGIC FUNCTIONS OF CADHERINS
  4. CADHERINS IN THE SKELETAL SYSTEM
  5. Acknowledgements
  6. REFERENCES