Concise Review: Unraveling Stem Cell Cocultures in Regenerative Medicine: Which Cell Interactions Steer Cartilage Regeneration and How?

This review provides a systematic overview of the literature on stem cell cocultures with cartilage cells and tries to elucidate the mechanisms that lead to chondrogenesis. It could serve as a basis for research groups and clinicians aiming to design and implement combined cellular technologies for single-stage cartilage repair and treatment or prevention of osteoarthritis.

Abstract

Cartilage damage and osteoarthritis (OA) impose an important burden on society, leaving both young, active patients and older patients disabled and affecting quality of life. In particular, cartilage injury not only imparts acute loss of function but also predisposes to OA. The increase in knowledge of the consequences of these diseases and the exponential growth in research of regenerative medicine have given rise to different treatment types. Of these, cell-based treatments are increasingly applied because they have the potential to regenerate cartilage, treat symptoms, and ultimately prevent or delay OA. Although these approaches give promising results, they require a costly in vitro cell culture procedure. The answer may lie in single-stage procedures that, by using cell combinations, render in vitro expansion redundant. In the last two decades, cocultures of cartilage cells and a variety of (mesenchymal) stem cells have shown promising results as different studies report cartilage regeneration in vitro and in vivo. However, there is considerable debate regarding the mechanisms and cellular interactions that lead to chondrogenesis in these models. This review, which included 52 papers, provides a systematic overview of the data presented in the literature and tries to elucidate the mechanisms that lead to chondrogenesis in stem cell cocultures with cartilage cells. It could serve as a basis for research groups and clinicians aiming at designing and implementing combined cellular technologies for single-stage cartilage repair and treatment or prevention of OA.

Introduction

Diseases of the musculoskeletal system, such as articular cartilage injuries and osteoarthritis (OA), are major contributors to disability. Active young and middle-aged patients suffering from these degenerative diseases can be faced with pain and limited mobility, resulting in reduced quality of life [1–4]. In fact, focal cartilage defects in the knee can impair quality of life as much as severe OA [1]. Eventually, damage to the articular cartilage leads to OA [5–8].

Over the past two decades, exponential growth in research of regenerative medicine has led to a variety of cartilage repair strategies [9]. For articular defects, microfracture and autologous chondrocyte implantation (ACI) are rapidly becoming the standard of care in specialized knee clinics [10]. In microfracture, which is generally used for smaller defects, the subchondral bone plate is penetrated to allow for bone marrow infiltration, blood clot formation, and subsequent fibrocartilaginous tissue generation [11, 12]. In the ACI technique, which is used for larger defects, chondrocytes are isolated from a biopsy, expanded in culture, and reimplanted in a second surgical procedure [13]. Positive clinical results of up to 20 years have recently been reported for ACI [14]. In addition, early results suggest that ACI can be used for treatment of (early) OA [15]. Although autologous cell-based treatments such as ACI are increasingly being used clinically, the practical limitations of harvesting chondrocytes from uninvolved areas of the joint and the associated high costs of prolonged culture have spawned research aimed at alternative cell sources and single-stage procedures.

Coculture-Induced Tissue Regeneration

Multipotent stem cells are considered alternative sources to autologous cartilage cells and are believed to be capable of differentiating into different cell types including osteoblasts, chondroblasts, and adipocytes [16]. Mesenchymal stromal or stem cells (MSCs) on their own, either in vitro or in vivo, have limited capacity to generate new cartilage matrix de novo unless they are stimulated by growth factors. The coculture approach for cartilage tissue engineering was initially introduced to design complex bone-cartilage scaffolds [17, 18]. For cartilage tissue engineering, the coculture technique was developed to enhance the cartilage forming potential of MSCs, to eliminate the need for chondrocyte expansion, and to understand the mechanisms of MSC differentiation to the chondrogenic lineage. Subsequently, from several studies on in vitro combination of stem cells and chondrocytes, it was concluded that chondrocytes induced the chondrogenic differentiation of MSCs (Fig. 1A) [19–44]. Since then, many different combinations have been introduced targeting articular cartilage regeneration and have shown promising results [45, 46]. In the debate on the mechanisms behind the coculture-induced regeneration, there seems to have been a paradigm shift recently, based on evidence suggesting that MSCs, rather than differentiating, have trophic effects on chondrocytes (Fig. 1B) [46–55]. Nonetheless, regeneration may also result from both chondroinduction and MSC differentiation (Fig. 1C) [56, 57]. These theories are based on a vast and heterogeneous literature that is difficult to grasp. Numerous questions remain as to what happens in these coculture systems: do the MSCs differentiate, disappear, and signal through trophic factors or a combination of the above? In addition, the mechanisms underlying the stimulatory effects of the cell combination technologies remain elusive.

Figure 1.

The three theories of the coculture approach. (A): After mixing chondrocytes and MSCs, chondrocytes stimulate MSCs to differentiate to the chondrogenic lineage. (B): Coculture of chondrocytes and MSCs results in chondroinduction; that is, MSCs excrete trophic factors such as FGF-1 and TGF-β that stimulate chondrocyte proliferation. (C): MSC differentiation plus chondroinduction: both chondroinduction and MSC differentiation contribute to chondrogenesis. Abbreviations: FGF-1, fibroblast growth factor 1; MSC, mesenchymal stem cell; TGF-β, transforming growth factor β.

In this paper, we systematically review the literature on combined cellular technologies that aim at single-stage procedures for cartilage repair and prevention of osteoarthritis. This allows the reader to find answers as to which coculture approaches enhance cartilage regeneration and what mechanisms lay at their foundations.

Justification of Search Strategy and Results

We performed a systematic review of the literature aimed at cocultures and coimplantation including stem cells for cartilage tissue engineering. A search was conducted in the electronic databases of Medline, Embase, and the Cochrane Collaboration, using the following search strategy: (Coculture) OR (Co culture) OR (coculture) OR (cocultures) OR (cocultures) OR (cocultured) OR (cocultured) OR (coculturing) OR (coculturing) OR (coculture-driven) OR (side-by-side) OR (side by side) OR (cell-to-cell) OR (cell to cell) OR (coimplantation) OR (coimplantation) OR (mixed) AND (cartilage) OR (cartilaginous) OR (chondrocytes) OR (chondrons) OR (chondrocytic) OR (chondrogenesis) OR (chondrogenic) OR (chondrogenically) OR (chondrogenically committed). The articles were screened by title and abstract by two separate observers (T.S.d.W. and L.A.V.) using the following inclusion criteria: papers describing an in vitro and/or in vivo coculture system for cartilage tissue engineering. Papers describing cocultures aimed at intervertebral disc regeneration were excluded. Reviews, case reports, missing full texts, and papers aimed at bone formation were also excluded (Fig. 2). When there was discrepancy in paper selection between observers, a consensus was reached. Using the same syntax, the search was updated regularly until October 14, 2013. The literature search performed in the Medline, Embase, and Cochrane Collaboration databases yielded 1,296 articles, 1,487 articles, and 0 articles, respectively. After removing 1,269 duplicates and screening titles and abstracts, 40 articles could be included. Of the included papers, one related article was identified from the references (Fig. 2) [40].

Figure 2.

Literature search flowchart. Abbreviation: IVD, intervertebral disc.

Characteristics of the Coculture Systems

Cell Types and Outcomes of Cocultures for Cartilage Regeneration

In general, there are three types of coculture approaches aimed at cartilage tissue engineering. The majority of studies have used a direct coculture with cell-cell contact of articular chondrocytes (ACs) and bone-marrow-derived MSCs (BM-MSCs). Other coculture systems were based on indirect coculture using either Transwell systems (Corning Enterprises, Corning, NY, http://www.corning.com) or conditioned medium (Table 1). The latter two approaches both aim at the eventual therapeutic use of a single cell source. Cells of equine, rat, porcine, bovine, rabbit, and human origins have been used for the coculture approach. In most studies, increased expression of cartilage markers such as type II collagen were found compared with MSC monocultures and similar or superior results were found compared with chondrocyte monocultures. Different groups found cocultures to reduce cell hypertrophy based on the downregulation of type X collagen (Table 1; studies 2, 4, 13, 25, 26, 31, 32, and 34). SRY (sex determining region Y)-box 9 (SOX9), which has been characterized as a transcription factor for chondrogenic differentiation, was found to be upregulated in cocultures using articular chondrocytes and MSCs in vitro (Table 1; studies 2, 7, 18, 22, 25, 29, and 40). SOX9 and its downstream genes have been shown to activate type II collagen expression and to be essential for normal skeletogenesis [58–61]. In only one study a significantly reduced deposition of cartilaginous extracellular matrix was demonstrated in a permeable insert coculture of rabbit chondrocytes and MSCs encapsulated in alginate beads (Table 1; study 39). In addition to BM-MSCs, other stem cells have been applied in cocultures such as adipose-derived stem cells (ASCs) (Table 1; studies 6, 8, 18, 28, 31, and 36) and synovium-derived MSCs (Table 1; studies 15 and 23). Three studies found limited effects on chondrogenic markers such as type II collagen when combining chondrocytes with ASCs (Table 1; studies 6, 18, and 36). This may suggest that, compared with BM-MSCs, ASCs are less efficient in coculture. The variable results found for ASCs in cocultures should stimulate future research to optimize their role in cartilage regeneration, especially because they are a highly accessible cell source. In contrast, coculture of chondrocytes with synovium-derived MSCs, embryonic MSCs, and induced pluripotent stem cells all induced consistent upregulation of markers of matrix formation compared with MSC monoculture (Table 1; studies 15, 19, 22–24, 29, 30, and 35). No significant difference in cartilaginous tissue formation (histological and quantitative glycosaminoglycan [GAG] analysis) was shown in cocultures between chondrocytes and human dermal fibroblasts compared with chondrocytes and MSCs (Table 1; study 20). However, when chondrocytes were replaced by murine embryonic fibroblasts in chondrocyte and ESC cocultures, the chondrogenesis in terms of GAG production and type II collagen expression dramatically decreased (Table 1; study 29). These findings suggest that chondrocytes are crucial for coculture-induced chondrogenesis. Alternative cell sources for cocultures such as fibroblasts are promising because they have the potential to replace the more expansive and scarce MSCs. Recent studies found promising regenerative results in cocultures with other cell types such as bone marrow mononuclear cells and expanded xenogeneic chondrocytes [62, 63]. Consequently, because coculture-induced chondrogenesis may not be stem cell specific, a thorough comparison of different cell types in cocultures may be of great value. It should be acknowledged that there is methodological heterogeneity among the studies included; for example, there is variety in controls, growth factors, and culture systems (Table 1). Indeed, not all studies included chondrocyte monocultures as a control (Table 1; studies 24, 29, 30, and 36). This is an important finding because different studies found that chondrocyte monocultures perform equally or superior to cocultures (Table 1; studies 9, 10, and 40). Furthermore, some studies chose to use the same amount of chondrocytes and/or stem cells in both monocultures and cocultures, resulting in different overall cell counts between the two conditions (Table 1; studies 29 and 33).

Table 1.

Cocultures and outcomes

The Cell Ratio in Cocultures

It has been hypothesized that the ratio of primary chondrocytes to MSCs used in coculture is important for chondrogenesis. In several studies, a 50:50 ratio of human chondrocytes and MSCs was found to provide optimal results in terms of chondrogenic markers such as type II collagen expression and GAG production (Table 2; studies 4, 9, and 38). Others found a clear advantage using a lower number of chondrocytes (up to 10%) (Table 2; studies 1, 5, 13, and 32). In hyaluronic acid hydrogels, a higher chondrocyte concentration (50:50 compared with ratios of 5:95 and 20:80 ACs to MSCs) was even shown to produce significantly lower mechanical stiffness, GAG, and collagen content (Table 2; study 32). Taken together, for chondrocytes cocultured with MSCs, higher MSC concentrations were found to produce the desired chondrogenic effects.

Table 2.

Coculture method specifications

Coimplantation for Cartilage Regeneration In Vivo

From this literature review, 12 studies were identified that applied the cell combination approach in vivo to determine its applicability in cartilage tissue engineering (Table 2; studies 1, 5, 13, 17, 24, 29, 34, 35, and 37). In all of these studies, a subcutaneous mouse model was used. Subcutaneously, threefold higher amounts of GAG per wet weight were found for cotransplantation of cells (ratio of ACs to MSCs: 30:70) compared with an MSC-only group using the same total amount of cells in a copolymer gel (Pluronic-F127) (Table 2; study 37). In addition, the mixed group showed similar GAG content with a chondrocyte-only group and a higher GAG content than a low-chondrocyte group. MSCs labeled with green fluorescent protein (GFP) were also found to be incorporated in the cartilage-specific matrix after 8 weeks (Table 2; study 37). Similarly, type I collagen meshes with human chondrocytes and MSCs in different ratios and either after immediate seeding or preculture resulted in a 3.8-fold higher GAG content and an increase in type II collagen staining compared with MSC- or chondrocyte-only implants (Table 2; study 13). In addition, higher aggrecan and cartilage oligomeric matrix protein (COMP) expression was found after 4-week implantation of chondrocytes and MSCs (50:50) in a hydrogel compared with MSCs alone (Table 2; study 17). Positive type II collagen staining was also shown after coimplantation of ESCs combined with human chondrocytes on a poly-D, L-lactide scaffold in a subcutaneous mouse model (Table 2; study 29). Implants with ESCs only appeared disorganized without type II collagen staining. Histological examination showed extensive vascular infiltration without evidence for inflammation. Although Fisher et al. (Table 2; study 34) found a significantly reduced mineralization staining score for mouse explants containing AC/MSC pellets compared with MSC-only pellets, lower semiquantitative histological scores were shown for the mixed group. Finally, in one study, different ratios of chondrocytes with their pericellular matrix (chondrons) and BM-MSCs were compared in a subcutaneous mouse model, and coimplantation consistently outperformed both chondron and MSC monoimplantation in terms of GAG production (p < .015) (Table 2; study 1). In their in vitro cocultures, the effect on GAG production was more than twofold higher for chondrons compared with chondrocytes (10%–50% chondrons or chondrocytes in coculture with MSCs). Next, a large-animal study was performed in which focal articular cartilage defects were created in both knees of eight goats and treated with a mixture of 10:90 chondrocytes and MSCs in fibrin glue or with microfracture. After 6 months, the AC/MSC group showed superior (p = .01) histological O’Driscoll [64] scores for the regenerated tissue and for biochemical regeneration (0.083 ± 0.037 mg of GAG per gram of tissue vs. 0.041 ± 0.013 mg of GAG per gram of tissue) over microfracture (Table 2; study 1). To date, the coimplantation approach has shown reproducible results for small-animal models and promising results for a large-animal model.

Mechanisms Behind the Coculture System

Cell-Cell Contact Versus Paracrine Signaling

To investigate the mechanisms behind the coculture model, in several studies, direct and indirect (permeable insert cultures; e.g., Transwell or Millicell [Millipore, Billerica, MA, http://www.millipore.com]) culture systems were compared (Table 1; studies 20, 27–29, 31, 32, 36–40). Only a few of these used the same culture conditions to investigate cells in direct and indirect contact (Table 1; studies 20 and 32). A greater than twofold Young’s and dynamic modulus as well as GAG and collagen content for mixed AC/MSC populations was found in a hydrogel compared with MSC- and chondrocyte-only groups with the same total cell amount (Table 1; study 32). When MSCs and chondrocytes were cultured in two distinct gels but in the same well (indirect contact through trophic factors), the outcomes were comparable to MSC monocultures. This implies that close cell proximity is essential. Meanwhile, another study also found that cell-cell contact was necessary to achieve cartilage matrix deposition for cocultures between primary chondrocytes and expanded chondrocytes in pellets and suggested that cell-cell contact by way of gap junctions could explain these cellular interactions because chondrocytes in culture express connexin 43 (Table 1; study 20). Chondrogenic matrix content in a pellet coculture of nasal cartilage or chondrocytes with BM-MSCs was 1.6 times higher compared with monocultures, but this was not observed in an AC/BM-MSC permeable insert system (Table 1; study 31). In polylactic acid/polyglycolic acid constructs, BM-MSCs produced cartilage-like tissue when cultured in a permeable insert system with chondrocytes (Table 1; study 37). The authors concluded that paracrine signaling in the form of soluble factors, not cell-cell contact, provided the chondrogenic signals. For both two-dimensional and three-dimensional permeable insert systems, however, BM-MSCs or amniotic MSCs were found to induce the morphological transformation of chondrocytes from a round phenotype to a spindle-like shape and to inhibit the generation of cellular aggregates and deposition of extracellular matrix (Table 1; study 37). Similarly, chondrocytes cocultured with ASCs using a permeable insert system showed a 3.5- to 4-fold reduction in collagen, COMP, and RUNX2 expression (Table 1; study 36). However, neither study used a direct coculture as a control. The majority of studies reported cell-cell contact to be crucial for cartilage regeneration. Because permeable insert cocultures do not seem to provide consistent chondroinduction, trophic factors alone may not be able to achieve optimal chondrogenic differentiation.

Differentiation Versus Chondroinduction

A limited number of studies investigated the fate of the different cells while in coculture (Table 1; studies 15, 24, 31, 33, and 38). In situ hybridization for Y chromosome, HLA characterization, detection of genomic DNA, and short tandem repeat loci in human/ xenogenic AC/MSC culture have been used to show a progressive loss in MSC numbers with an enhancement of chondrogenic proliferation (chondroinduction) (Table 1; studies 9, 31, and 38). The change in cell ratio during the first 3 days, in addition to apoptosis of the iMSCs, was attributed to bovine chondrocyte proliferation, as shown by Ki-67 labeling and terminal deoxynucleotidyl transferase dUTP nick end labeling staining (Table 1; study 31). In contrast, major histocompatibility complex I staining after a 4-week subcutaneous implantation of a scaffold seeded with mouse ESCs and porcine chondrocytes in nude mice suggested that 20% of the mouse ESCs survived (Table 1; study 24). In a coculture of a human MSC line (Kp-MSC) and GFP-labeled chondrocytes, the number of Kp-MSCs increased substantially, and they were found to express type II collagen (Table 1; study 33). It seems that there is evidence that MSCs both stimulate chondrocyte proliferation in vitro and are capable of differentiating into the chondrogenic lineage. However, this mechanism has yet to be thoroughly investigated, especially in vivo.

The Role of Soluble Factors in Cocultures

Exogenous Soluble Factors

Several studies have sought to determine the role of soluble or trophic factors in stimulating cartilage regeneration in cocultures (Table 1; studies 2, 7, 18, 25, 26, 30, 31, and 34–37). Greater chondrogenic gene expression (type II collagen: 948-fold; aggrecan: 353-fold; SOX9: 3.81-fold) was observed for cocultures of chondrocytes and MSCs compared with MSCs solely treated with transforming growth factor β (TGF-β) (Table 1; study 2). Similarly, limited effects of TGF-β1 on chondrogenic differentiation of ASCs were found, whereas these effects became synergistic when combined with conditioned medium obtained from chondrocytes (Table 1; study 28). In two separate studies, Wu et al. (Table 1; studies 15 and 38) found no differences in chondrogenesis of a coculture system when using chondrocyte proliferation medium or chondrocyte differentiation medium containing TGF-β3, dexamethasone, and bone morphogenetic protein 6. In the absence of these growth factors, however, BM-MSCs almost disappeared in the coculture, whereas 40% remained using differentiation medium. This suggests that MSCs show better survival in a chondrogenic environment. Others found no additional effects of TGF-β1 on chondrogenic gene expression in cocultures (Table 1; studies 7 and 25). However, transfection of both induced pluripotent stem cells with TGF-β1 and chondrocytes with TGF-β3 in cocultures have been shown to have a positive effect on the expression of cartilage-related genes such as type II collagen, aggrecan, and COMP (Table 1; studies 23 and 30). Finally, a 10-fold lower TGF-β3 concentration was sufficient to achieve equivalent chondrogenesis compared with AC monocultures, suggesting cocultures have a higher sensitivity to TGF-β3 (Table 1; study 3).

Endogenous Soluble Factors

Using a three-dimensional hydrogel permeable insert system with human osteoarthritic chondrocytes and human MSCs, an increase in type II collagen and aggrecan expression was demonstrated compared with MSCs treated solely with conditioned medium (Table 1; study 26). Mass spectrometry indicated there were up to 53 proteins that were unique in the coculture system compared with monoculture, in which 44 were extracellular histones. The remaining candidates were extracellular matrix proteins and growth factors. Adding chondrocytes to BM-MSCs in pellet cocultures was found to reduce alkaline phosphatase activity in vitro and (histological) calcification in vivo (Table 1; study 34). Prolonged parathyroid hormone-related protein (PTHrP) was found to be chondrocyte specific. In fact, by treating MSC monocultures with unconditioned medium containing PTHrP, both alkaline phosphatase activity and type X collagen expression showed a more than twofold decrease. These findings suggest chondrocytes inhibit calcification and hypertrophy when in coculture with MSCs (Table 1; study 34). Both separate and combined blockage of TGF-β1, insulin-like growth factor 1, and bone morphogenetic protein 2 (BMP-2) in conditioned medium was also shown to suppress the proposed chondrogenic differentiation (type II collagen expression) of MSCs (Table 1; study 37). For ASCs in coculture with chondrocytes in both monolayers and microbeads, adding a high concentration of a neutralizing antibody against fibroblast growth factor (FGF) type 2 eliminated the apparent apoptotic effect of vascular endothelial growth factor A on chondrocytes (Table 1; study 36). A recent study also showed the potential stimulatory trophic role of FGF-1 in coculture (Table 1; study 16). In this study, FGF-1 expression was found to be upregulated with a fold change of 2.25 with quantitative polymerase chain reactions after a 48-hour coculture of human chondrocytes and MSCs. Positive FGF-1 staining was found in 72% of labeled MSCs that were in close proximity to or directly in contact with chondrocytes. Blockage of FGF signaling by specific FGF inhibitors or FGF-1-neutralizing antibodies blocked chondrocyte proliferation, as shown by 5-ethynyl-2′-deoxyuridine staining. In summary, several studies could not reproduce the effects of cocultures by using soluble factors alone, whereas others have shown the added value of adding trophic factors to the culture medium or transfecting stem cells with TGF-β. Finally, FGF-1 could be an important trophic factor, secreted by MSCs, that stimulates short-term chondrogenesis in coculture.

Conclusion

Research in regenerative medicine has rapidly evolved in recent decades, giving rise to a variety of tissue-engineering approaches for the treatment of musculoskeletal injuries. For cartilage repair, the focus has shifted from in vitro expansion and subsequent reimplantation, to single-stage procedures. This eliminates the need for expanded cells that seem to have less regenerative capacity and that entail a large financial burden.

This is the first review that provides an extensive and systematic overview of the literature that describes coculture models aimed at articular cartilage tissue engineering. It gives a first global insight in the interactions between the cells as described for the different coculture approaches. Currently, the available literature suggests that cocultures using chondrocytes and a variety of stem cells are able to produce cartilaginous tissue both in vitro and in vivo. The majority of these articles used AC/BM-MSC cocultures with 10%–50% chondrocytes and found the technology to stimulate chondrogenesis both in vitro and in vivo.

Different studies included in this review tried to identify key mechanisms behind the chondrogenesis found in coculture systems. The majority of this literature concluded that the increased chondrogenesis is attributable to differentiation of MSCs. However, few studies have investigated the actual fate of the individual cells. The increase in chondrogenesis that was found to be accompanied by MSC apoptosis strongly suggests that, instead of MSC differentiation, MSCs produce trophic factors that stimulate regeneration (Table 1; studies 15 and 38). The facts that different markers were found to be unique for the coculture system and that chondrogenesis in cocultures was related to TGF-β, insulin-like growth factor 1, BMP-2, PTHrP, and FGF-1 expression also suggest there is a role for trophic factors [65–69] (Table 1; studies 26, 34, and 37). Although the first steps have been taken, understanding the exact cellular mechanisms that lay at the foundation of coculture-induced chondrogenesis may give rise to more simplified and easily translatable techniques. If a cocktail of trophic factors could be identified, for example, and/or a material constructed that could replace either one of the cells used in the current coculture systems, an off-the-shelf product providing the cues for chondrogenesis is conceivable.

In this review, we have identified several studies that have compared the direct and indirect coculture systems and found cell-cell contact to be essential for chondrogenesis (Table 1; studies 31 and 32). The study by Xu et al. (Table 1; study 39) underscores the importance of cell-cell contact because the authors have demonstrated that a permeable insert culture of chondrocytes and MSCs in alginate beads gave a morphological change of chondrocytes from a round shape to a spindle-like shape with less deposition of chondrogenic extracellular matrix [75]. This leads to the authors’ interpretation that both cell-cell contact and trophic factors are required for cartilage regeneration in cocultures (Fig. 1C) [56, 57]. Further understanding of these complex interactions and signals may allow for the development of cell constructs in which these signals are optimized either in cocultures or monocultures.

Although a number of studies used small-animal models to test the coculture constructs, only one study applied a large-animal model and implanted autologous chondrons and allogeneic MSCs in a fibrin glue carrier in fresh cartilage defects (Table 1; study 1). The promising (histological) results, the lack of adverse events in this study, and the extensive preclinical findings of several other studies have warranted the recent initiation of two human trials. One of these trials uses primary autologous chondrocytes in combination with autologous bone marrow cells (INSTRUCT trial, NCT01041885), and the other uses autologous chondrons and allogeneic MSCs (IMPACT Trial, 2012-001570-29). The initial safety of such combined cell therapies could support further development of these tissue-engineering techniques. At least as important, it may allow for better understanding of the behavior of coimplanted cells in vivo.

The majority of studies that were included in this review have shown that cocultures outperform MSC monocultures and provide equal or superior results to AC monocultures. For now, the use of MSCs alone for cartilage repair may not be optimal because in vitro-differentiated MSCs have been shown to express the hypertrophic marker type X collagen and may not be capable of forming ectopic stable cartilage in vivo [70–72]. This review supports future translational research that aims to elucidate the complex cellular interactions in coculture-based treatments and to develop single-stage cell therapies targeting cartilage tissue engineering and (prevention of) osteoarthritis.

Author Contributions

T.S.d.W., J.A.A.H., X.Z., and L.A.V.: conception and design, data analysis and interpretation, manuscript writing, final approval of manuscript; L.B.C. and W.J.A.D.: manuscript writing, final approval of manuscript; M.A.R. and D.B.F.S.: conception and design, manuscript writing, final approval of manuscript.

Disclosure of Potential Conflicts of Interest

M.A.R. has compensated research funding.