Main

Following the formation of a synovial joint, cartilage growth and maturation allow for the enlargement of the joint and attainment of the adult properties allowing the cartilage to function for many decades as a low friction, wear resistant, bearing material. However, injury and degenerative conditions of the articular cartilage may compromise these functions and result in pain and disability. The treatment paradigm for articular cartilage damage frequently involves surgical intervention, since mature cartilage has a poor capacity for repair. Surgical methods range from penetrating the subchondral bone to elicit a repair response to implantation of cells and tissues to the replacement of partial or whole joints (1). Cartilage tissue engineering has been an active area of study focused on bringing biologically-active tissue substitutes into clinical use. Tissue engineering strategies offer the potential to provide benefits beyond those of the current repair procedures, including the formation of more consistent hyaline-like repair tissue, the use of nonimmunogenic cells and tissues, enhanced integration with surrounding tissues, and reduced rehabilitation times. The ability to realize these advances will likely depend equally on knowledge of cartilage biology and engineering innovations.

The diversity of form and functional properties in the body's many joints arises from the genetic, epigenetic, and environmental factors which govern cartilage morphogenesis, growth, and maturation. Disruptions of these regulating factors, occurring spontaneously or in a controlled manner in vivo or in vitro, can alter the normal course of cartilage development and growth. Each of these occurrences serves as an opportunity to learn about the mechanisms underlying these phenomena and apply that knowledge to the treatment of cartilage diseases and injuries.

The goal of this article is to review aspects of articular and epiphyseal cartilage growth, maturation, and morphogenesis with applications toward bioengineering chondral tissue substitutes for therapeutic purposes. A thorough review of the cartilage developmental biology is not intended, since a number of good reviews of the subject already exist (2–4). However, the lessons learned from cartilage developmental biology are important to understanding and improving the in vitro engineering of cartilaginous tissues. These translational aspects of cartilage biology as they relate to bioengineering analysis and cartilage tissue engineering will emerge as the focus of this article.

Cartilage growth, remodeling, maturation, and morphogenesis typically occur concurrently during the course of in vivo joint development, yet these processes might be manipulated independently in an in vitro system. Some definitions are provided here to clarify the meaning of these terms, as used herein. Although the term growth may be used loosely to encompass many of the changes in a tissue before reaching maturity, a more rigorous interpretation is useful for bioengineering analyses. Simply, growth is the increase in tissue volume resulting from the accumulation of material similar to the original. As is often the case, growth may be accompanied by remodeling of the existing tissue to incorporate the new material. However, remodeling, which refers to the modification of tissue properties via biochemical and biophysical changes in structure, may also occur in the absence of growth (5,6). Maturation is the progressive change of tissue composition and structure to achieve the functional properties of the adult state. Last, morphogenesis is the development of form resulting from specific patterns of cellular organization and tissue growth and remodeling during embryogenesis (7). With these underlying definitions, it is possible to exam cartilage growth, maturation, and morphogenesis using both physical and mathematical approaches.

GROWTH AND MATURATION

Mechanisms and modes of growth.

The growth of cartilage may arise nonuniformly in distinct regions of the tissue through specific biologic activities of the indwelling cells termed chondrocytes. These cells are embedded in a highly hydrated extracellular matrix consisting predominantly of collagen type II and aggrecan, a large, negatively charged proteoglycan (8). Growth may be mediated through increases in either the cell volume fraction by proliferation or hypertrophy or in the matrix volume fraction by a net accretion of constituent molecules. Early in cartilage development, the tissue is highly cellular, so significant changes in cell number or volume would likely have proportional effects on the tissue volume. As sources of matrix secretion, chondrocytes at high density would also be well poised to contribute to growth through their anabolic production of matrix molecules. However at later stages of growth, chondrocyte density is reduced (9), and the capacity for further growth is diminished.

The growth of a tissue may also occur in two distinct modes: growth throughout the volume of the substance, termed interstitial growth, and growth at a surface or interface, termed appositional growth (10). Interstitial growth is simply illustrated as a spheroidal material occupying region R1 at time t = t1 which grows throughout the volume to occupy region R2 at t = t2 (Fig. 1A). Assuming that the set of material points in R1 is dense in the set of points in R2, points in R2 (e.g., P2 and Q2) can be mapped back to points in R1 (e.g., P1 and Q1) (11). Similarly, appositional growth is illustrated as an ellipsoidal material with a two-sided growth surface, S0, separating material occupying regions R1 and R2 at t = t1 (Fig. 1B). New material is deposited on both sides of the growth surface forming regions R3 and R4 by t = t2, whereas the original regions R1 and R2 might deform only to accommodate these new regions.

Figure 1
figure 1

(A) Interstitial growth is illustrated as a material occupying region R1 at t = t1 and R2 at t = t2. (B) Two-sided appositional growth at surface S0 results in no growth within regions R1 and R2, but formation of new regions R3 and R4. (C) Cartilage growth potentially includes both interstitial growth by matrix accretion and chondrocyte hypertrophy and appositional growth by proliferation of cells in the superficial-most layer demarcated by S0. Illustrations in (A) and (B) are adapted, from Skalak R et al., J Theor Biol 94:555–577, Copyright © 1982 Elsevier, Ltd., with permission.

Full size image

Examining which biologic mechanisms of growth are active in the different regions of cartilage is critical to determining its modes of growth. Using tritiated thymidine, two bands of proliferating chondrocytes were identified in immature rabbit knees, one near the articular surface and a second, deeper and close to the ossification front (12). By detecting proliferating cell nuclear antigen or the incorporation of bromodeoxyuridine, the presence of a proliferating layer of cells at the articular surface in immature rats, bovines, and South American possums was confirmed, though these studies did not find a deeper band of dividing cells (13,14). This thin layer of proliferating cells near the surface may serve as a means of appositional growth (15), which is consistent with the reported existence of a progenitor cell population at the articular surface (16–18). However, the lack of mitotic activity in the bulk of the growing cartilage suggests that cell division plays little or no role in interstitial growth of articular cartilage.

Interstitial growth of cartilage is likely to arise from the enlargement of chondrocytes and their robust production of matrix molecules. In the growth plates of rats, these two processes result in much of the total tissue elongation, with chondrocyte hypertrophy accounting for 44–59% of the total growth and matrix accretion accounting for 32–49% (19). Similar quantification of the contribution of these processes to overall tissue growth in epiphyseal cartilage is lacking, though there is evidence that both occur. Increasing chondrocyte size with depth in mature articular cartilage lends support to the contribution of hypertrophy (20). Chondrocyte density also decreases significantly with growth and maturation, which in the absence of appreciable cell death is indicative of abundant matrix accretion (9). Thus, cartilage growth may proceed via a combination of appositional and interstitial growth (Fig. 1C). Understanding these modes of growth and the underlying biologic processes may facilitate tissue engineering efforts to grow cartilage of a required size.

Changes accompanying maturation.

The maturation of cartilage in vivo is coincident with the growth of the tissue and can be characterized by a number of changes in the tissue's structure, composition, and functional properties. The bovine stifle joint provides a well-studied example of articular cartilage properties at various stages of maturity (third trimester fetal, 1–3 wk-old calf, and 1–2-y-old young adult). As mentioned previously, cellularity decreases with maturation from fetus to adult by roughly half or more (9,21,22). The solid content (dry weight/tissue volume) increases from ∼12% in the fetus to ∼18% in the calf and ∼20% in the adult (21). This change in solid content is largely attributable to an increasing collagen concentration which may nearly triple from fetus to adult (21,22). During this period the collagen network is also stabilized through the formation of pyridinoline crosslinks, which may increase in concentration from the fetal stage by 2- to 4-fold in the calf and greater than 7-fold in the adult (22,23). On the other hand, the glycosaminoglycan (GAG) concentration is relatively maintained (21,22), though the size of individual chondroitin sulfate chains and their ratio to keratan sulfate typically decrease with maturation (24). These compositional changes from fetus to adult are manifested in marked alterations of the cartilage mechanical properties including a ∼180% increase in compressive aggregate modulus and a ∼450% increase in equilibrium tensile modulus (21,22). A large portion of these biochemical and functional changes takes place in the few short weeks between the late fetal and neonatal periods. This demonstration of rapid maturation during normal growth supports the concept that under certain conditions the functional properties of engineered cartilage may be significantly augmented during brief in vitro culture or during a short, in vivo rehabilitation time.

Cartilage maturation is also reflected by the architecture of the tissue. The thicker immature cartilage is marked by features which identify it as both articular cartilage and epiphyseal growth plate. The bulk of the immature cartilage is often described as having a fairly isotropic structure with chondrocytes arranged at random (25,26). However, cell density may decrease with depth, particularly near the articular surface (25), whereas GAG and collagen concentrations and tensile and compressive moduli may increase (27,28). On the other hand, mature articular cartilage possesses distinct superficial, middle, and deep zones. In mature tissue, this transition, from the superficial most layer to the deepest, is characterized by cellular changes including a decrease in density, a rounding of morphology, and an organization into clusters in the middle zone and columns in the deep zone (25,26). Likewise, aggrecan content increases with depth (29), and the predominant collagen orientation shifts from parallel to perpendicular to the articular surface (30). These compositional and structural differences are also reflected in the depth-varying mechanical properties in compression and tension (31,32). Mature articular cartilage also bears the distinguishing feature of a calcified cartilage interface with the subchondral bone. This feature provides a strong interdigitated bond between uncalcified cartilage and bone through which joint forces are transmitted. The formation of similar architectural features in engineered tissues may provide mechanical and tribological properties that are important for successful joint repair.

Mathematical descriptions of growth.

Complementing the physiologic aspects of growth, mathematical descriptions of growing bodies can provide a framework in which growth phenomena can be analyzed. General approaches for analytically describing both volumetrically distributed (interstitial) growth and surface (appositional) growth have been summarized previously (10). Volumetrically distributed growth may be modeled using a growth extension tensor analogous to a material strain tensor in solid mechanics, while surface growth may be represented through a velocity field at the growth surface (10). These fundamental growth models are widely adaptable to better understand any number of growing tissues including cartilage and bone.

Expanding upon the volumetrically distributed growth model, cartilage growth mixture models have been developed using a continuum mechanics approach (5,6). In these models, cartilage is represented as a mixture of components: typically two growing elastic constituents, namely proteoglycans and collagen, and an inviscid fluid. Other cartilage constituents, such as collagen crosslinks or cells, may be included, since the theory allows for an arbitrary number of growing components. The process of growth is decomposed into two separate tensors which describe growth deformations from mass deposition and mechanical deformations which provide compatibility of the constituents in an equilibrium stress state. Changes in the size, biochemical composition, and tensile properties of growing cartilage explants in vitro have been successfully described using this growth mixture model approach (6). With similar assumptions, a cartilage growth finite element model has been proposed recently and may facilitate the description of specific experimental protocols such as compression or perfusion (33). Such models of cartilage growth may ultimately be useful for predicting the geometry, composition, and biomechanical properties of a tissue.

Compartmental and continuum metabolic models of cartilage growth and remodeling provide an alternative means of describing changing tissue composition. Metabolic models are based on the principle of mass balance for one or more components such as proteoglycan or collagen. Single compartment models have been frequently used in the analysis of radiolabel experiments to determine rates of molecular synthesis and degradation. These models have been used to quantify the effects of chemical and mechanical stimuli on the metabolism of cartilage proteoglycans and collagens (34–36). The kinetics of collagen crosslinking in cartilage explants has also been described using similar compartmental models (37). Although a compartmental model can only account for the spatially averaged concentration of a component over time, a continuum model may describe the spatial distribution of such a component. One such continuum model was developed with consideration of the synthesis, transport, binding, and degradation of matrix components and was applied to understanding proteoglycan distributions in articular cartilage (38). Recently, this model was expanded to analyze the effects of fluid flow and membrane permeability on the development of depth varying composition of engineered cartilage in a perfusion bioreactor (39). Together, these types of models of cartilage growth and metabolism help provide a quantitative description of the dynamics of growth, remodeling, and maturation which may be useful in creating or evaluating engineered tissues.

In vivo regulation of growth and maturation.

Chondrocyte cell fates and metabolic activities provide the means of cartilage growth and maturation and are regulated by chemical and mechanical stimuli (Fig. 2). Applied stimuli are transmitted through the tissue according to the physical and chemical properties and become a chemical or mechanical change in the chondrocyte microenvironment. These microenvironmental signals potentially act upon the chondrocytes to alter cell fate and matrix metabolism. In turn, these cellular activities may change the composition, structure, and functional properties of the tissue, thereby modifying the way subsequent stimuli are perceived. Determining what stimuli regulate cartilage growth and maturation and by what mechanisms they act is critical to understanding cartilage development and may be applicable to the engineering of biologic tissue substitutes.

Figure 2
figure 2

Dynamic processes govern the growth and maturation of articular cartilage as shown for the fetal to adult transition of the bovine knee. Regulatory stimuli are transformed according to the physical and chemical properties of the tissue into signals in the chondrocyte microenvironment to alter matrix metabolism and cell fates. These cellular processes, in turn, are responsible for the enlargement of the joint and the evolving tissue composition, structure, and functional properties. Some of these changes, particularly the decrease in cell density and attainment of a mature, stratified architecture, are demonstrated by cartilage digital volumetric images. Reprinted from Jadin KD et al., J Histochem Cytochem 53:1109–1119, Copyright © 2005 The Histochemical Society, Inc., with permission.

Full size image

A large effort has been put forth to identify the molecular mediators such as growth factors and transcription factors which govern joint formation, growth, and maturation. Regulatory molecules from various growth factor families such as the insulin-like growth factors, fibroblast growth factors, and transforming growth factors-β (TGF-β), which includes the bone morphogenetic proteins (BMP) and growth differentiation factors, have been implicated in these processes. Many human skeletal defects are associated with mutations in genes encoding these growth factors as well as their receptors and antagonists, a selection of which is nicely summarized by Shum et al. (2). The development of transgenic mice has provided great insight into how these chemical regulators affect cartilage development. However, many of the chemical signaling molecules which influence cartilage growth and maturation have functions earlier in development, for example in cellular differentiation or joint site patterning. The disruption of these earlier functions can mask the effects of these molecules in later stages of growth. In such cases, the conditional inactivation of the gene in cartilage at a specified time may help clarify the roles of these signaling molecules.

Mechanical stimuli also serve a regulatory function for cartilage growth and maturation in vivo. Joint morphology, which is the physical manifestation of spatially nonuniform cartilage growth, is sensitive to the mechanical environment of the joint. Although further detail will be given in a later section, it is important to note that abnormal cartilage growth and joint morphology may result from altered physical forces on the tissue during prenatal development and beyond. In addition, a number of in vivo studies with immature dogs have provided evidence that different regimes of joint loading, via exercise or limb immobilization, may lead to significant differences in articular cartilage structure (e.g., thickness), composition (e.g., proteoglycan content), and functional properties (e.g., indentation stiffness) (40).

In vitro manipulation of growth and maturation.

The culture of isolated cells or cartilage explants in a controlled in vitro environment has yielded additional insight into the biochemical regulation of cartilage growth and maturation. Although the processes of growth and maturation proceed concurrently in vivo, an in vitro system may allow for independent manipulation of these processes (Fig. 3). It has been hypothesized that changes in cartilage size and in maturation (as quantified by the tensile properties) could result by altering the balance of proteoglycan and collagen metabolism (27,41,42). Culturing immature cartilage explants with fetal bovine serum (FBS), insulin-like growth factors-1, or bone morphogenetic protein-7 (BMP-7) resulted in an expansive growth phenotype characterized by large increases in size, stimulation of proteoglycan in excess of collagen, and decreases in tensile stiffness and strength (42). Alternatively, culture with TGF-β1 promoted cartilage homeostasis with little change in size, composition, or mechanical properties (42). The identification of biochemical stimuli which promote maturation in immature cartilage explants has been elusive, though depletion of proteoglycans by chondroitinase-ABC treatment before culture resulted in increased tensile stiffness (27). These distinct treatments to promote expansion, homeostasis, or maturation may find use in tissue engineering cartilage of specific size or maturity.

Figure 3
figure 3

Expansion (or size), maturation (e.g., tensile or compressive stiffness) and composition are properties of cartilage controlled by the application of regulatory stimuli. In vivo, cartilage grows and matures concurrently; however, expansion and maturation may be manipulated independently within a controlled in vitro environment. Explants of immature cartilage cultured with fetal bovine serum (FBS), insulin-like growth factors, or bone morphogenetic protein-7 (BMP-7) exhibit expansive growth, while those pretreated with chondroitinase ABC (C-ABC) followed by culture in FBS undergo maturational growth. Culture in transforming growth factor-β1 (TGF-β1) yields a near homeostatic state.

Full size image

In vitro experiments have also been useful for examining the effects of mechanical stimuli, such as static and dynamic compression, hydrostatic pressure, and fluid shear, on the metabolic activities, particularly those of aggrecan, in cartilage explants and cells (43). The application of such stimuli in a controlled manner in vitro provides an accurate knowledge of the loading history, though these simplified loading conditions do not fully reflect the complexities of joint loading and articulation. The effects of loading on metabolism are dependent on the magnitude and frequency of loading, with the general paradigm that static stresses inhibit aggrecan synthesis while moderate dynamic stresses promote synthesis.

Enhancement of engineered cartilage composition and mechanical properties may also be prescribed through specific regimes of biochemical and mechanical stimuli. The effects produced by biochemical stimuli are dependent on magnitude (or dose) and frequency (duration and timing) just like their mechanical counterparts. Further complexity arises from the possibility of additive and synergistic effects of combinations of different stimuli. As an example, one recent study reports that sequential stimulation of agarose-chondrocyte constructs with TGF-β3 followed by dynamic compressive loading resulted in significantly higher compressive modulus and GAG content compared with constructs which received the same growth factor stimulus but were mechanically stimulated or left free-swelling throughout the culture (44). That study and others have been able to produce engineered cartilage containing proteoglycans at or near native levels following weeks or months of culture (44,45). However, engineering cartilage with collagen contents approaching native levels has proved to be more challenging. This problem is made more difficult by the rapidly changing tissue dynamics illustrated in Figure 2. As is often the case, engineered cartilage consists of cells elaborating a matrix into a nonnative environment, such as a scaffold, hydrogel, or high-density culture. The different physicochemical properties, initial lack of cell-matrix interactions, and altered cell morphology may cause cells in engineered tissues to respond to stimuli in a manner unlike that of chondrocytes in their native extracellular matrix (46–48).

Engineering the architecture of mature cartilage.

The zonal stratification of articular cartilage is a hallmark of maturation and the recapitulation of this architecture has been a focus of recent tissue engineering efforts. Several studies have taken advantage of the phenotypic differences of chondrocytes from different zones of native cartilage. For instance, chondrocytes isolated from the middle and deep zones readily produce an aggrecan-rich matrix to a greater extent than those near the articular surface (49). However, cells in the superficial-most layer synthesize and secrete proteoglycan 4 (PRG4), a contributor to the boundary lubrication function of cartilage (50,51). Stratified tissues have been engineered using chondrocyte subpopulations isolated from different zones and layered through sequential seeding or hydrogel encapsulation (52–54). On the other hand, constructs created from mixed subpopulations of chondrocytes at high density have also exhibited a propensity to develop a zonal architecture resembling native tissue through a process independent of cell sorting (55). It remains unclear how such organization arises but may result from differences in physical constraints or the availability of biochemical factors and nutrients between the surface layer and the deeper tissue.

The formation of a calcified cartilage interface has been another target of recent tissue engineering efforts. The presence of this feature, which anchors the cartilage to the subchondral bone and assists in load distribution, may be pivotal in strengthening the integration of engineered cartilage with bone. Conceptually, the calcification of the interface could be engineered in osteochondral constructs in vitro or induced in vivo following implantation of an engineered chondral or osteochondral graft. The feasibility of the first approach has been shown by researchers using deep zone chondrocytes and culturing them on top of a calcium polyphosphate scaffold in the presence of beta-glycerophosphate to promote calcification (56). This technique also makes use of phenotypic differences in chondrocyte subpopulations, since deep zone chondrocytes are known to produce markers of hypertrophy and calcification such as collagen type X and alkaline phosphatase (57). The in vivo formation of a calcified cartilage interface, as indicated by the presence of a tidemark, has also been observed following the implantation of tissue engineered constructs in rabbit models of osteochondral repair (58,59). However, the appearance of this feature was inconsistent, and the ability to manipulate the in vivo calcification of the cartilage-bone interface has not yet been demonstrated.

THE DEVELOPMENT OF FORM

Regulation of joint morphology.

The process of joint morphogenesis begins around the time of cavitation and rapidly produces congruent articular surfaces with specialized forms. The diversity of these forms is evident from the many types of joints and their associated ranges of motion, from the ball and socket configuration of the hip to the hinge-like knee with dual condyles. The development of the hip in humans is a particularly well documented process. Cavitation occurs around the 7th week of gestation, and in the subsequent 4 weeks the predominant forms of the spherical femoral head, short femoral neck, and primitive greater trochanter become apparent (60). It is evident that specific patterns of differentiation and growth provide the means of such morphogenetic changes. The processes of growth (i.e., matrix deposition and cell division and hypertrophy) can serve as morphogenetic factors only if they occur differentially throughout a tissue (7). However, the mechanisms which guide these patterns of growth to produce such different forms warrant further study. Recent evidence suggests that specific patterns of growth or differentiation factor 5 and cell division direct morphogenetic changes in the embryonic chick interphalangeal joint (3). Additionally, the role of the local biomechanical environment in the morphogenesis of joints has long been of interest, and mechanical stimuli have been deemed crucial in the proper development of joint form. Paralysis of chick embryos may result in the failure of cavitation at joints sites or the fusion of cartilage anlagen already separated (61,62).

Although morphogenesis results in the development of basic joint forms during embryogenesis, the actual morphology of the joint may continue to change beyond the embryonic stage as a result of growth or remodeling. Fetal joints are not proportionately scaled-down versions of their mature counterparts (see images in Fig. 2). Rather, the proportions of joint features may change relative to one another in a manner dependent on the rates and directions of growth. Models of cartilage growth and ossification as mechanically-dependent processes have illustrated how distinct profiles of spatially varying growth rates may alter subsequent joint morphology (63,64). Experimental evidence confirms that postnatal mechanical loading of joints may influence their morphology. Exercise may result in cartilage thickening and bone remodeling to improve congruency in the knees of immature dogs (40). Likewise, exercise can produce cartilage thickening and longer, flatter subchondral surfaces in the femoral heads of immature mice (65).

Biomechanics also plays a pivotal role in the development and treatment of common disorders of joint morphology. Developmental dysplasia of the hip (DDH) is characterized by abnormal growth and morphology of the femoral head and acetabulum. Risk factors for DDH include those linked to genetics (e.g., gender, ethnicity, and family history) and an altered mechanical environment (e.g., breech presentation and oligohydramnios) (66). Dislocation or lateral subluxation of a morphologically normal hip and the resulting alteration in contact stresses may induce dysplasia in the femoral head and acetabulum (67). Nonsurgical treatment for DDH may rely on harnesses or casts to provide reduction of the femoral head within the acetabulum to achieve a mechanical environment which facilitates molding of the articulating surfaces (68,69). Similarly, mechanical manipulation of joint morphology through casting has been used in the treatment of club foot and shown to alter growth rates of cartilage anlagen during this corrective process (70). These findings indicate that biomechanical regulation of cartilage growth continues to be an important factor in determining joint morphology beyond the embryonic stage.

The role of proper form in joint repair.

The congruity of apposing joint surfaces is critical to the cartilage's ability to distribute load and provide smooth articulation. Consequently, the restoration of physiologic joint geometry should be an important consideration in therapeutic strategies employing engineered cartilage. The effects of using improperly sized and shaped engineered tissues may resemble those produced by poorly inserted osteochondral grafts. Malpositioned osteochondral grafts may result in abnormal biomechanics including increased contact pressures (71) and cartilage strains (72). To minimize incongruity produced by osteochondral allograft transplantation, orthopedic surgeons frequently use orthotopic grafts from donor tissue matched to the recipient based on size (73,74). Similar care in designing the sizes and shapes of engineered chondral and osteochondral grafts may well be a factor in successful cartilage repair.

Engineering cartilage forms.

Chondral tissues of specific shapes and sizes have been generated using various scaffold-based and scaffold-free techniques. Scaffolds for cartilage tissue engineering may be given an initial shape via casting, free-form fabrication, and similar methods before seeding with cells. The application of these techniques in creating osteochondral constructs resembling a phalanx and mandibular condyle have been demonstrated (75,76). Hydrogels with encapsulated cells are particularly well suited for the creation of constructs with complex shapes, such as those of the patella and mandibular condyle, since they can be polymerized within a negative mold (77,78). In a scaffold-free approach, forms resembling the femoral head, acetabular, or mensical cartilages have been generated by high density seeding of primary (79) or alginate-recovered cells (manuscript in preparation) in negative molds. One concern regarding these techniques is that they only set the initial form of a construct which may change through subsequent growth and remodeling. It may be necessary to predict how such growth would alter the shape of these tissues, perhaps through improvements of the growth models previously described, or to implement measures to control and direct the shape of these tissues during culture.

Just as the morphology of a growing joint is influenced by physical forces in vivo, the shapes of chondral tissues may also be altered via mechanical stimuli in vitro. Although the use of mechanical stimuli to promote functional maturation of engineered cartilage has been an active area of study over the last decade (80), the application of such stimuli to selectively alter the form of cartilaginous tissues is relatively underexplored. Recent work has demonstrated that the application of static bending to immature cartilage explants in a deformation-controlled manner can alter the shape of these tissues during culture (81). The process by which this reshaping occurred was largely independent of chondrocyte biosynthesis, suggesting the involvement of a biophysical or biochemical remodeling of the extracellular matrix rather than growth. Combining this mechanical reshaping technique with one of the methods which provide initial form may emerge as a complementary approach to engineering shaped chondral tissues.

CONCLUSIONS

Growth and maturation of cartilage of articular joints are dynamic processes mediated by cell fates and extracellular matrix metabolism. These processes are regulated by complex biochemical and mechanical stimuli which are being elucidated through both in vivo and in vitro approaches. Mathematical models also provide a means of quantitatively describing the changing composition, functional properties, size and shape of tissues, and may be a useful tool for designing experimental conditions to improve engineered tissues. Many lingering questions about the mechanisms and regulation of cartilage growth remain to be answered, and many challenges still exist for the bioengineer trying to create engineered cartilage of a specified size, shape, and maturity. Integrating various fabrication and culture techniques, such as those mentioned here, will become important to meet the clinical requirements for engineered cartilage tissues.