Ontogeny Informs Regeneration: Explant Models to Investigate the Role of the Extracellular Matrix in Cartilage Tissue Assembly and Development

Abstract

Osteoarthritis (OA) is typically managed in late stages by replacement of the articular cartilage surface with a prosthesis as an effective, though undesirable outcome. As an alternative, hydrogel implants or growth factor treatments are currently of great interest in the tissue engineering community, and scaffold materials are often designed to emulate the mechanical and chemical composition of mature extracellular matrix (ECM) tissue. However, scaffolds frequently fail to capture the structure and organization of cartilage. Additionally, many current scaffold designs do not mimic processes by which structurally sound cartilage is formed during musculoskeletal development. The objective of this review is to highlight methods that investigate cartilage ontogenesis with native and model systems in the context of regenerative medicine. Specific emphasis is placed on the use of cartilage explant cultures that provide a physiologically-relevant microenvironment to study tissue assembly and development. Ex vivo cartilage has proven to be a cost-effective and accessible model system which allows researchers to control the culture conditions and stimuli, and perform proteomics and imaging studies, that are not easily possible using in vivo experiments, while preserving native cell-matrix interactions. We anticipate our review will promote a developmental biology approach using explanted tissues to guide cartilage tissue engineering and inform new treatment methods for OA and joint damage.

Introduction

The extracellular matrix (ECM) directs the form and function of articular cartilage, both in healthy and diseased tissue. Alterations in the ECM incurred by traumatic injury or dysregulated biochemical signals initiate a degenerative cascade in the joint, leading to advanced osteoarthritis (OA). OA is the most common joint disease, and is mainly characterized by a failure to repair damaged cartilage, which contributes to a cascading biochemical and mechanical decline in joint tissues ¹. Cartilage degeneration can be treated in late stages by total joint replacement prostheses, with over 1 million surgeries performed every year in the United States, making it the most common elective surgical procedure ^2,3. Prosthetic joints have a high risk of post-surgery complications and a limited lifetime ⁴, and as lifespan increases, the likelihood of another surgery increases. Earlier interventions, such as hydrogel implants, are currently of great interest in the treatment of cartilage defects to prevent or delay the need for total joint replacement; however, success of such scaffolds as a treatment of OA is limited ^5–7.

Current tissue regeneration strategies focus on emulating the structure and chemical composition of mature cartilage ^8,9. The goal of this approach is to develop a biocompatible scaffold that mimics the mature native ECM, yet provides support for new tissue to develop and integrate with the host. A successful scaffold must provide direct biophysical and biochemical cues for cell growth, signaling, and migration in a relatively avascular and aneural tissue ^10,11. Chondrogenesis is most efficient when the scaffold degradation rate matches the matrix deposition rate ¹². Emerging scaffold designs that aim to match the stratified physical structure of the bone-cartilage interface use mechanical density gradients ¹³ or varied porosity within scaffold layers ¹⁴. Considering the tissue constructs that aim to mimic the multilayer structure of mature cartilage, researchers have yet to realize architectures that encapsulate both mechanical and complex chemical cues from the ECM that are required to maintain the cartilage cell (chondrocyte) phenotype ¹¹. Moreover, researchers have struggled to achieve integration of these repair constructs with host tissue ^15–17 since local chondrocytes maintain low proliferation rates and minimal regenerative abilities. Therefore, numerous challenges remain for the regeneration of articular cartilage before we can achieve successful restoration of key features, including biomechanical integrity.

Surprisingly, researchers rarely leverage knowledge of ECM remodeling that takes place during embryogenesis as design inspiration for engineered cartilage. The initial tissue assembly process plays a significant role in directing cellular behavior to form mature tissue that is functional and stable. Many embryonic and fetal tissues are capable of scar-free repair ^18–21. Thus, scaffolds using cues from developing environments could better promote repair without the adverse effects of scar tissue formation such as restricted movement and compromised mechanical integrity. One reason for limited success in scaffold integration may be that the stiffness and matrix composition of these engineered tissues mimic that of adult tissue. Cellular behavior is significantly affected not only by tissue stiffness, but also by the composition of ECM in that environment ²². In contrast with the homeostatic adult, assembling articular cartilage consists of different relative amounts of ECM, which turn over more quickly and have vastly different mechanical properties. Researchers have been unable to capitalize on these instructive cues for scaffold design due to the limited knowledge regarding the composition, turnover, organization, and mechanical properties of developing cartilage undergoing ontogenesis.

We propose that explanted cartilage tissue is an effective tool to track cartilage ontogeny and to understand the fundamentals of tissue regeneration. Explanted articular cartilage refers to an intact tissue extracted from a viable limb and investigated while sustained in media to maintain viability, tissue integrity, and cell-matrix interactions; cartilage explants therefore allow for the study of chondrocytes surrounded by the native ECM ²³. Cartilage explant models can be beneficial for research purposes because they accurately emulate in vivo conditions, especially in short-term (days to weeks of) culture, while avoiding the challenges of in vivo work. It is also difficult to study certain mechanisms of musculoskeletal development in vivo. For instance, studying the process of endochondral ossification is complicated by the interdependence of chondrocyte differentiation and need for vascularization ²⁴. Murine knockout models for many genes essential for musculoskeletal development are expensive to maintain or are embryonic lethal, limiting the possibilities to study musculoskeletal assembly in vivo ²⁵. Meanwhile, chemically-induced OA murine models have a different pathophysiology than in humans, so characterizing matrix degradation in vivo is not clinically informative ²⁶. However, the relatively avascular and aneural tissue composition of cartilage makes it suitable for an ex vivo culture system to observe and understand factors related to tissue assembly and healing. Freshly harvested cartilage maintained via tissue culture more closely represents native ECM architecture compared to frozen or preserved tissue sections ²⁷, further validating the use of viable explant models.

Our goal is to promote novel cartilage tissue engineering approaches informed by cartilage assembly and development during musculoskeletal ontogenesis (Figure 1). Additional information about musculoskeletal ontogenesis, and specifically articular cartilage development, is needed to identify design parameters that will effectively facilitate the formation of new, functional tissues. We focused our review on recent ex vivo systems, as well as the most effective metrics to evaluate matrix biochemistry, matrix mechanical properties, and resulting chondrogenic stage to argue that cartilage explant models are an employable system to study cartilage development.

Figure 1.

Informed cartilage tissue engineering via emulation of biomechanical and compositional parameters of the developing cartilage extracellular matrix. Conventional cartilage tissue scaffold parameters are guided by the composition and architecture of mature cartilage, which does not capture the rapid turnover of the ECM during development essential for stable tissue assembly. Mature, static cartilage has sparsely distributed chondrocytes with a localized pericellular matrix in a stiffer ECM containing type II collagen fibers and aggrecan. In contrast, during embryogenesis, chondrocyte progenitor cells and tissues begin to differentiate within a disorganized matrix of type VI collagen and widely distributed perlecan. During postnatal development, post-translational modifications (PTMs) stabilize the collagen structure while increased levels of EMILIN-1 ensure structural integrity of the deposited matrix fibrils. Typical nanoscale matrix stiffness values are indicated at each developmental stage. Hydrogel or decellularized scaffolds that imitate the turnover of chemical and mechanical parameters during development show great potential for healing.

Structure and composition of mature and developing cartilage

Mature cartilage functions to maintain a homeostatic environment to preserve robust structural integrity, and thus has a limited capacity for repair. This complex tissue is composed of chondrocytes embedded in ECM consisting of 70–80% water, 10–25% type II collagen, and 5–15% proteoglycans, noncollagenous proteins, and glycosaminoglycans (GAGs) ^28,29. Hyaluronic acid (HA), a GAG composed of Ν-acetyl-D-glucosamine and D-glucuronic acid, is prevalent throughout connective tissues, particularly in synovial fluids ³⁰ and interacts with chondrocytes via various surface receptors ³¹. The ECM is synthesized and maintained by chondrocytes that are sparsely distributed and occupy between 1–6% of the total tissue volume ²⁸. In mature cartilage, a pericellular matrix tightly surrounds each chondrocyte and type II collagen fibrils and GAGs fill the interstitial space between cells, providing resistance during tension and compression, respectively. The mechanical forces within each region are felt by chondrocytes and influence mechanotransduction pathways that regulate gene stability ²⁹. Mature articular cartilage is avascular, and migration and proliferation of chondrocytes is limited by the dense fibrous structure of the tissue ³². As a result, the tissue metabolic rate is slow, matrix synthesis is severely restricted, and the overall healing rate of mature cartilage is low ³³. Moreover, the conventional tissue engineering paradigm, which often relies on activation of resident cells and degradation and remodeling of the implanted scaffold ³⁴, may not prove an effective strategy for cartilage.

Developing cartilage, in contrast to adult cartilage, has a vastly different mechanical and biochemical composition with increased capacity for repair, which has the potential to facilitate scar-free, fetal tissue regeneration ²¹. For example, perlecan, tenascin-C and versican are highly abundant at the onset of cartilage condensation during embryogenesis, when chondrocytes are densely packed ³⁵. During postnatal maturation, perlecan is localized to the pericellular matrix ³⁶, tenascin-C becomes restricted to the perichondrium ³⁷ and versican is downregulated ³⁸. Throughout postnatal growth, articular cartilage continues to function as a surface growth plate ³⁹ and gradually acquires a zonal organization thought to be important for its long term biomechanical stability ⁴⁰. As cartilage matures, a well-defined surface zone forms that consists of elongated chondrocytes that produce lubricant proteins to protect the tissue from shear stress. A middle zone contains rounded chondrocytes and high GAG content. A deep zone has progressively larger chondrocytes organized in columns that are involved in matrix production and interact with the underlying subchondral bone, and the interior epiphysis is highly vascularized and thought to serve as a reservoir for cells that support chondrocyte enlargement and expansion of cartilage tissue ^41–43. Chondrocytes begin depositing extensive amounts of ECM more indicative of mature cartilage (e.g. type II collagen, aggrecan ^38,44), which reduces cell density. However, much remains to be discovered regarding the dynamics of ECM composition and cell organization during chondrogenesis, and the regulatory mechanisms by which articular cartilage stratifies during postnatal maturation remains an open question. Many developmental and maturation processes, including the upregulation of a transitional ECM ^45,46, are recapitulated in tissues capable of scar-free repair and regeneration ⁴⁷. Expected values for Young’s moduli of assembling, maturing, and adult homeostatic cartilage generally increases between each developmental stage; matrix stiffness measured with AFM are illustrated in Figure 1. Tissue engineers will benefit from more detailed knowledge of how the cells, surrounding ECM, and signaling factors are naturally employed and progressively change during initial tissue assembly to synthesize functional cartilage.

Architectural and chemical factors that dictate the initial assembly of cartilage provide a foundation for tissue formation distinct from the healing response of mature musculoskeletal tissues. In mature cartilage healing, fibrin strands span tissue defects to facilitate population by mesenchymal stem cells, which differentiate into chondrocytes ⁴⁸. Within the bone compartment of deeper defects, osteogenesis occurs alongside chondrogenesis and subchondral bone is deposited along the surface of the defect ⁴⁹. Scar-like tissue dominates over cartilage tissue and resident chondrocytes adopt a fibroblast cellular morphology ⁵⁰. Native articular cartilage adjacent to the defect site becomes necrotic, and no resorption or remodeling of cartilage tissue occurs ⁵¹. In stark contrast, ECM synthesis and remodeling during chondrogenesis is accompanied by a dynamically changing architecture ⁵². It is this complex and changing environment that results in the assembly of functional cartilage tissue during development.

Current investigation of chondrocyte and cartilage culture systems that inform scaffold design

Demands for developing in vitro engineered cartilage tissues have led researchers to consider multiple reductionist approaches. One such approach is the study of aggregate (pellet) cultures of MSCs and chondrocytes. These cells are harvested from tissues, expanded in culture, and seeded at a high cell density, resembling early cartilage condensation. After 3 to 4 weeks in culture, pellet cultures of MSCs show high expression of collagen II and aggrecan reminiscent of mature cartilage ⁵³; similarly, pellet culture of articular cartilage progenitor cells present chondrocyte-like phenotypes ⁵⁴. Pellet culture models are useful to elucidate differentiation of chondrocytes in defined microenvironments, as well as demonstrate the self-organization of cartilage tissue constructs, but information on the assembly of mature cartilage structure and function is lacking. Meanwhile, explant cultures may be used alongside pellet cultures to investigate and define specific developmental timepoints while maintaining structural complexity of neotissues.

Studies that focus on cartilage explant culture systems are affordable, readily available, and maintain complex cell-matrix interactions. Ex vivo studies sustain viable cartilage tissue for several weeks in culture without the cost or complexity of in vivo studies. Once culture protocols are established, explant studies are generally cheaper and faster than in vivo studies. A major advantage of ex vivo cartilage studies is the ability to control and quantitatively monitor specific treatments because biochemical conditions are held constant. Thus, comparison between samples is based primarily on varied parameters and less influenced by biological variations between animals. In vivo studies are best to provide a natural environment to study OA or the systemic effects of reactions to treatments, or procedures including inflammation or bone adaptation. Explant studies are thus best used to answer more fundamental research questions to understand the specific nature of cartilage, while in vivo studies may be best used to investigate a response under natural conditions over long periods of time. Because explanted tissue cultures are extracted from their dynamic in vivo environments and kept in a static culture environment, this model system becomes limited for studying regeneration. Each model system is valuable depending on the research question, and studying explant cartilage tissues is a valuable stepping stone to in vivo applications.

Cartilage explants have been used for decades to characterize cell and tissue mechanical function and response to exogenous (e.g. biomechanical, biochemical) stimuli. In well-known studies, explants were tested in confined and unconfined compression to establish the bulk multiphasic nature of the tissue electromechanical properties ^55–57. Related studies further described surface (tribological) properties ⁵⁸, and intra-tissue depth-dependent properties and mechanics ^59–62. More recently, cartilage explant studies investigated the responsiveness of chondrocytes to mechanical stimuli like shear loading ^63–66 and chemical factors including growth factors and cytokines ^67–69. Moreover, early studies using progenitor cells in explant culture revealed the ability to differentiate cells toward cartilage and bone phenotypes ⁷⁰. A recent study comparing cartilage-only to osteochondral tissues in explanted culture systems demonstrate that intact osteochondral plugs provide a more physiologically relevant system after 28 days in culture, as they preserve processes and interactions between tissue types ⁷¹. These systems may be used to study degenerative and regenerative properties of mature cartilage in vitro.

Commercially available scaffolds vary in their composition and fabrication; as a result, the success of these scaffolds in clinical application varies widely. Decellularized cartilage is considered a suitable basis for tissue engineering, which is a process that clears cellular material to reduce risk of host rejection while maintaining matrix composition and architecture. However, success of decellularized scaffolds leads to impairment of the mechanical properties of the tissue and shows poor cellular infiltration in larger defects ^72,73. Meanwhile, biopolymers such as hydrogels contain components of the ECM and can regulate migration, adhesion, and differentiation of embedded cells better than traditional polymers ⁷⁴. The use of hydrogels in tissue engineering is relatively new and growing in popularity because of their capacity to mimic the physiological chemistry of the ECM while having a high degree of biocompatibility ^75,76. Although these hydrogel features are advantageous, some suffer from poor biocompatibility leading to inflammation, proliferation of fibroblasts leading to scar tissue, and subsequent implant degradation ^47,77. Additionally, adhesion between hydrogel scaffolds and adjacent tissue is critical to promote integrative repair ⁷⁸. Researchers have reported a variety of methods to promote adhesion, such as gelatin-gelatin adhesive interactions at the tissue surface ⁷⁹; combining collagen with electrospun poly-L-lactic acid nanofibers ⁸⁰; constructing a bi-layer scaffold in which the top layer is a sponge containing proteins or a gene delivery system ^81,82; or simply engineering scaffolds that are inherently adhesive ⁸³. Researchers in the field are working to closely mimic the complex structure of mature cartilage with these hydrogel structures ^8,11 but the quiescent nature of embedded chondrocytes has not been widely shown to facilitate the generation of new tissue.

While current scaffold success is limited, we note that few are modeled from ontogeny-informed parameters that may guide cartilage tissue assembly and promote regeneration upon implantation. Hydrogels are a promising technology and have been further developed to time-release growth factors to facilitate healing ^84,85 or change stiffness, affecting cell fate ^86,87. These studies for treatment of articular cartilage defects commonly focus and assess their scaffold success based on induced differentiation of seeded cells into chondrocytes, increasing production of type II collagen present in mature host cartilage tissue, and limiting the host inflammatory response, all of which have shown limited success in vivo. Select hydrogel scaffolds have demonstrated success of cartilage-like assembly in vivo, namely a recently developed injectable hydrogel with time-release of kartogenin (KGN) ⁸⁸, a small molecule shown to coordinate synovial joint development ⁸⁹. There is great promise in the cartilage tissue engineering field to tune current chemically and mechanically dynamic hydrogels to mimic the ontogenic phenotype, and a great need to define instructions for developmentally-informed scaffolds.

Explant models correlate ECM composition and mechanics in development and maturation

The use of cartilage explant models may be applied to support the study of development to inform regeneration. In particular, they have the potential to enable the quantification of ECM turnover to identify stage-specific tissue remodeling that guides cartilage assembly. Below, we outline studies that describe how the composition and microscale mechanics of developing cartilage can be investigated using viable explant model systems.

There is a pressing need to identify ECM composition and structure during development. Historically, identification of key matrix components (e.g. collagen, GAGs) at distinct timepoints in fetal, calf, and adult articular cartilage tissues has been conducted using biochemical assays ⁹⁰. While this approach has provided foundational knowledge of structure-composition relationships, there still exists significant knowledge gaps regarding the precise composition of the ECM that are required to understand chondrogenesis. To this end, liquid chromatography-tandem mass spectrometry (LC-MS/MS) can be used to identify ECM composition by mapping measured peptide masses to a database of protein fragmentation patterns ⁹¹. Since only a small mass of tissue is needed (~ 1 µg), LC-MS/MS is amenable for investigating developing tissues. ⁹².

LC-MS/MS has been used to identify the ECM of extracted articular cartilage during postnatal murine development and reveals a breadth of protein changes during chondrocyte maturation ⁹³. By analyzing the matrix proteome of developing postnatal mice at discrete timepoints P3 and P21, the relative amounts of different ECM components were compared to show how composition changes during postnatal maturation. Postnatal cartilage is enriched in a number of cell adhesion proteins, notably EMILIN-1 (Elastin microfibril interface located protein 1) which binds fibrillin microfibrils ⁹⁴. Post-translational modifications of collagen α chains stabilize the intermolecular collagen structure and interactions. Throughout postnatal maturation, glycoproteins tenascin-X and tenascin-C are both upregulated ⁹³. These findings demonstrate the wide changes of ECM contents during just two stages of postnatal maturation and establish a starting point for using these instructional cues in cartilage implants that may improve tissue stability and integration.

The field will benefit from future studies focusing on the dynamics of ECM proteins at embryonic timepoints using emerging metabolic labeling methods such as non-canonical amino acid (ncAA) labeling. ncAAs, such as the methionine analog azidohomoalanine, enable the quick labeling and enrichment of newly synthesized proteins from complex mixtures for LC-MS/MS analysis ⁹⁵. This method contrasts traditional radiolabeling methods that require longer labeling periods and are unable to resolve kinetics of newly synthesized proteins that are in low abundance. It is easier to restrict methionine consumption within in vitro culture, which enables labeling of proteins with higher fidelity than in vivo. Notably, the labeling and enrichment of ncAA-labeled ECM has been recently demonstrated ^95–97. By combining these recently established ncAA-enrichment protocols with in vitro explant culture of cartilage from different developmental timepoints and LC-MS/MS, the precise composition and turnover of ECM will be identified. In addition, the use of tissues from small animal models will provide the ability to probe developmental time steps with fine resolution (e.g., E16.5 and E18.5 days) ³⁶. The proteomic analysis of different tissue sets will contribute to our understanding of both the molecular mechanisms and ECM composition underlying cartilage assembly.

In addition to the chemical composition of the developing ECM, it is crucial to measure the stiffness of cells and the microscale architecture within the tissues. It is well established that cell behavior and intracellular signaling pathways are significantly influenced by the mechanics of the external cellular matrix environment, and the stiffness of cells can be modulated through differential ECM-based adhesion molecules on a substrate of otherwise constant mechanical properties ^98,99. Cells that bind to an assortment of ECM through surface receptors differentially regulate cytoskeletal architecture and cellular stiffness. Atomic force microscopy (AFM) is the primary methodology to resolve differences in mechanical properties at the cellular level using tissue sections generated via a cryotome. However, a recent study demonstrated that freezing alters the bulk matrix mechanics and leads to cell death, preventing the direct measurement of mechanics of the native ECM and chondrocytes ²⁷. Therefore, to collect quantitative data relevant to native ECM architecture, ex vivo cartilage tissues must be mapped on a range of scales.

To address the need for mapping microscale mechanical properties, our lab established an AFM technique that addressed the need for microscale mechanical measurements in viable developing cartilage explants by generating slices using a vibratome ³⁶. We demonstrated that perturbations in ECM composition directly affect cell stiffness. Interestingly, altering the organization of the actin cytoskeleton via cytochalasin D treatment reduced chondrocyte stiffness without affecting the matrix. Meanwhile, a hyaluronidase treatment that disrupts HA had a significant effect on the matrix and chondrocyte stiffness. These results indicate that an explant approach could be applied to measure the cartilage response to chemical stimuli, such as enzymatic digestions and chemotherapy exposure, to extract information about the response of living cartilage explant tissues. Additionally, AFM techniques can be used to closely monitor mechanical properties in viable tissues during cartilage assembly (Figure 2) ³⁶. These methods will allow us to gain a better understanding of tissue-mediated cell behavior throughout development.

Figure 2.

Explant models enable the study of simultaneous changes in matrix structure and mechanical function through development. Perlecan knockdown significantly decreases the stiffness of the cells and ECM in developing articular cartilage. A) By utilizing atomic force microscopy, high-resolution stiffness maps of intact tissues show an increase in stiffness with age and perlecan incorporation. B) Both cell and ECM stiffness are influenced by perlecan content, where the compressive modulus significantly increases as a function of age and increased perlecan incorporation to the matrix. Figure reproduced, with permission, from Xu et al.³⁶

The local chondrocyte environment is also extremely important in mediating cell behavior and needs further classification. For instance, perlecan knockdown disrupts the pericellular matrix (PCM) and significantly affects cell and matrix stiffness during a time course of embryonic and postnatal development in ex vivo tissue samples ³⁶. Viable forelimb tissues from embryonic and postnatal tissues were sectioned using a vibratome for AFM testing on cells and their surrounding matrix in situ. Overall, ECM compressive modulus increases from E16.5 to P3 in wild type, which reflects the transition from a perlecan-rich matrix in embryonic joints to an aggrecan-rich ECM in rapidly maturing pups ¹⁰⁰. Disrupting the PCM lowers the overall ECM stiffness, likely due to disrupted growth factor signaling and alteration to the interstitial matrix structure. Knockdown of perlecan alters matrix organization and significantly decreases chondrocyte and matrix stiffness, revealing the role of the PCM in regulating intracellular mechanisms required for the functional development of cartilage. These results reveal the insights gained from measuring cellular-scale structures in developing cartilage systems to correlate matrix components with mechanical properties, all within viable tissue explants where native tissue characteristics are preserved.

Maintaining tissue viability through ex vivo systems makes the study of microenvironment/cell interactions possible. Likewise, studying sample sections in an aqueous environment permits the application of reagents to target components of the matrix and enable measurement of the mechanical influence felt by cells in the ECM. These methods provide the framework for future studies of cartilage degeneration through pathogenesis, but they may also be implemented during a time course of embryonic development to identify in situ tissue-cell interactions that lead to functional cartilage assembly.

Developing explant cultures demonstrate cartilage assembly

Whole-limb explant models, which may contain an undisrupted joint, support the study of cartilage development and may lead to insights to induce tissue regeneration scaffolds in culture. Ex vivo limb cultures have been used for decades as a model for limb development. Embryonic chick development has been the most common model due to the accessibility of the egg during development ¹⁰¹ but there is a significant increase in translatability to human tissue regeneration, particularly between differentially regulated gene expression profiles, if the model organism is murine ¹⁰². The components of ECM tissues in the musculoskeletal system are largely conserved across mammalian species ^103–105, permitting the use of scaffolds informed from xenogeneic cartilage in clinical applications. Synovial joint formation is far from understood and includes articular cartilage tissue assembly, so may be considered of great interest in the tissue engineering field. Previous studies have shown that interzone cells actively participate in joint formation and can give rise to multiple musculoskeletal tissue types ¹⁰⁶ but much remains to be deciphered about specific cues that result in tissue type differentiation. Viable limb explants containing a synovial joint provide a method to investigate microscale architecture among chondrocytes by enabling observation of the cell and tissue differentiation processes that comprise tissue assembly. However, there are a limited number of studies that specifically investigate the early limb via viable murine tissue explant.

Developing limb cultures of murine organs provide distinct advantages as a model system that includes maintained integrity of joint regions, simultaneous assessment of tissue interactions, and observation of cartilage assembly from distal to proximal regions. In the early limb, mesenchymal condensations prefigure the location of eventual joint elements ¹⁰⁷. After blood vessels are eliminated from the tissue site, cells undergo chondrogenic differentiation and organize into maturation zones, eventually ossifying and leaving a stable, mature joint consisting of a variety of chondrocyte and fibroblast phenotypes ¹⁰⁸. This distinctive and important process is only observable by studying tissue that is composed of developing tissue types without compromising integrity.

Murine whole-limb cultures have been shown to be more useful for investigating drugs that promote synovial joint formation than traditional 2D or micromass cultures ⁸⁹. For example, KGN is a small molecule that was identified to have chondrogenic potential using a high-throughput screen using in vitro mesenchymal culture ¹⁰⁹; however, its usefulness in the context of tissue development was unknown. More recently, a group maintained embryonic mouse limbs in culture and found that KGN promotes multiple limb developmental stages. Fluorescence imaging further demonstrated that entire synovial joints were formed in the ex vivo culture even without the cyclic loading normal in developing embryos. These studies indicate that KGN promotes mechanisms that orchestrate overall development of limb structure, including proper chondrocyte differentiation ⁸⁹. Explants of developing joints have been extracted from their in vivo milieu and cultured in a static environment, eliminating the imperative function of embryonic movement that coordinates musculoskeletal assembly. However, this reductionist approach reveals key insights to the impact of KGN treatment alone, which has already shown promise to promote regeneration of cartilage ^88,110–112.

The emphasis on limb development as a means to inform tissue engineering is relatively novel, and its application is an example of the potential of using limb explants to demonstrate the effects of drugs on tissue development. The use of explant limb culture for drug discovery is a way to bridge the gap between cell culture and in vivo studies, and to demonstrate the promising results of drugs without the cost or systematic effects that couple with in vivo studies. This approach demonstrates the usefulness of a structure more complex than a typical cartilage explant or monolayer of chondrocytes in vitro.

Advantages and limitations of cartilage and developing explants

In vivo model systems are frequently utilized, but as previously discussed, are expensive and may not allow for the investigation of the nature of cartilage directly. Cartilage tissue is ideal for ex vivo studies because it does not rely on vascularization or nerves to remain viable. Cartilage explants and whole-limb cultures typically survive for a few days and up to 2 weeks in culture ^113,114, and provide a way to investigate the tissue in nearly in vivo conditions. The traditional cartilage explant-based model is simple and easy to produce and has the major benefit of the ability to monitor cells in their natural ECM so tissue features such as matrix degradation can be observed ^115–117. Meanwhile, whole-limb cultures have the further benefit of leaving the cell microenvironment intact and maintaining interactions between tissue types. As described above, performing drug studies on intact limb explants can be more informative than chondrocyte cultures alone because the matrix is undisrupted and chondrocytes remain in their native 3D environment.

However, the use of tissue explants creates new challenges. For instance, results may be influenced by death at dissection edges of the tissue ¹¹⁸, few replicates are possible from the same source, and long-term alterations of tissue properties may occur in culture ¹¹⁹. In developing whole-limb cultures, vascularization and ossification of musculoskeletal tissue are major steps during development that are not maintained in vitro. As a result, the hypertrophic zone of cartilage is thinner than normal and filled with elongated chondrocytes ¹¹⁴. Explant and developing whole-limb cultures that are maintained in media culture lack mechanical stimulation that is necessary to support development and maintain homeostasis. Resident chondrocytes of both developing and mature cartilage experience mechanical forces with each muscle contraction, then transduce the loading into biological signals, ultimately changing gene expression and protein synthesis ¹²⁰.

To remedy these limitations, novel in vitro bioreactors have been developed to culture cartilage under controlled biological and mechanical conditions ¹²¹. A recently published benchtop bioreactor system functions as a cartilage-on-a-chip device capable of mechanical actuation ¹²². This device uses pressurization of an actuation compartment capable of achieving up to 30% compression and can recapitulate mechanical stimuli involved in OA pathogenesis. Additional bioreactors for tissue culture of articular cartilage were recently reviewed elsewhere. ¹²³. The approach using a bioreactor culture system is beneficial to investigate chondrocyte responses to individual biomechanical perturbations. Future studies with explant culture systems that employ these principles may also be applied to whole-limb culture by attachment to an actuator at joint boundaries, are needed to assess cartilage assembly and hydrogel integration. Studying whole-limb cultures are more appropriate for identifying developmental processes such as the assembly and differentiation of cells into joint tissues, while studying mature cartilage explants are most useful for matrix degradation and understanding the mature cartilage tissue response to treatment.

Conclusion and outlook: need for development-informed tissue engineering and cartilage repair

Our goal is to promote novel cartilage tissue engineering approaches informed by cartilage assembly and development during musculoskeletal ontogenesis, and especially the use of explants to explore ontogeny. Knowledge of the structure of immature cartilage may encourage the design of novel scaffolding materials and engineered explants that replicate the developing matrix, e.g. through the use of dynamic changes of perlecan localization and collagen content (Figure 1). Because embryonic tissues undergoing tissue assembly exhibit scar-free tissue repair ²¹, mimicking such processes would lead to more structurally competent tissue regeneration. Several in vitro models to investigate cartilage tissue assembly are of interest and will provide complementary information. While pellet culture or whole limbs may be ideal to study early condensation of cartilage (i.e. initial tissue assembly), investigation of matrix and tissue maturation (occurring both during embryogenesis as well as postnatally) will be better investigated using explant cultures of embryonic and postnatal cartilage. The use of explants harvested from developing animals allow the controlled study of specific factors important to promote tissue assembly, such as KGN and perlecan, using reductionist approaches ^36,89.

We expect this review will encourage the researchers to utilize structure-property turnover relationships of embryonic and postnatal cartilage to design three-dimensional scaffolds and bioreactor mechanical loading systems to better direct stem cells toward a chondrogenic cell fate. This can be achieved using ex vivo tissues that have an intact ECM, which are not only widely affordable and accessible for laboratories but also offers easy monitoring of desired tissue parameters. Understanding the structure-function relationship between chondrogenesis and the changing ECM will better inform our understanding of tissue regeneration. At a minimum, more successful regenerative scaffolds will use cues from the developing ECM and seed chondrocytes at a much higher density than that of adult cartilage with the goal of kickstarting the cells to recapitulate development and build their own homeostatic cartilage tissue.

The fields of tissue engineering and regenerative medicine aim to restore normal cellular and tissue function. Progress with new approaches to engineer or regenerate cartilage has been limited, partly due to the need for data describing the spatiotemporal dynamics of ECM and how that influences the mechanical properties, and chondrocyte gene expression. In this review, we identified that cartilage explant models aid in the study of ontogeny informed regeneration. The avascular, aneural nature of mature cartilage makes it unable to effectively repair itself and maintain an articular phenotype over a fibrocartilage phenotype. Alternatively, focusing research around developing cartilage will yield data on ECM turnover in chemistry and architecture, which ensures long-term tissue stability. By using limb ontogeny to inform tissue regeneration, current issues in chondrocyte dedifferentiation, scarring, and scaffold integration may be resolved. Future studies that reveal the instructional cues that guide natural cartilage tissue assembly are essential before the cartilage tissue engineering community may move forward with new approaches for engineered scaffold design.