Cartilage regeneration using principles of tissue engineering.

It is well known that articular cartilage in adults has a limited ability for self-repair. Numerous methods have been devised to augment its natural healing response, but these methods generally lead to filling of the defect with fibrous tissue or fibrocartilage, which lacks the mechanical characteristics of articular cartilage and fails with time. Recently, tissue engineering has emerged as a new discipline that amalgamates aspects from biology, engineering, materials science, and surgery and that has as a goal the fabrication of functional new tissues to replace damaged tissues. The emergence of tissue engineering has facilitated the generation of new concepts and the revival of old ideas all of which has allowed a fresh approach to the repair or regeneration of tissues such as cartilage. The collaborations between scientists with different backgrounds and expertise has allowed the identification of some key principles that serve as the basis for the development of therapeutic approaches that now are less empiric and more hypothesis-driven than ever before. The current authors review some of the considerations regarding the various models used to test and validate the above repair methods and to address different aspects of the cartilage repair paradigm. Also, some key principles identified from past and current research, the need for the development of new biomaterials, and considerations in scale-up of cell-biomaterial constructs are summarized.

Tissue-engineered fabrication of an osteochondral composite graft using rat bone marrow-derived mesenchymal stem cells.

This study tested the tissue engineering hypothesis that construction of an osteochondral composite graft could be accomplished using multipotent progenitor cells and phenotype-specific biomaterials. Rat bone marrow-derived mesenchymal stem cells (MSCs) were culture-expanded and separately stimulated with transforming growth factor-beta1 (TGF-beta1) for chondrogenic differentiation or with an osteogenic supplement (OS). MSCs exposed to TGF-beta1 were loaded into a sponge composed of a hyaluronan derivative (HYAF-11) for the construction of the cartilage component of the composite graft, and MSCs exposed to OS were loaded into a porous calcium phosphate ceramic component for bone formation. Cell-loaded HYAFF-11 sponge and ceramic were joined together with fibrin sealant, Tisseel, to form a composite osteochondral graft, which was then implanted into a subcutaneous pocket in syngeneic rats. Specimens were harvested at 3 and 6 weeks after implantation, examined with histology for morphologic features, and stained immunohistochemically for type I, II, and X collagen. The two-component composite graft remained as an integrated unit after in vivo implantation and histologic processing. Fibrocartilage was observed in the sponge, and bone was detected in the ceramic component. Observations with polarized light indicated continuity of collagen fibers between the ceramic and HYAFF-11 components in the 6-week specimens. Type I collagen was identified in the neo-tissue in both sponge and ceramic, and type II collagen in the fibrocartilage, especially the pericellular matrix of cells in the sponge. These data suggest that the construction of a tissue-engineered composite osteochondral graft is possible with MSCs and different biomaterials and bioactive factors that support either chondrogenic or osteogenic differentiation.

BMP-2 induction and TGF-beta 1 modulation of rat periosteal cell chondrogenesis.

Periosteum contains osteochondral progenitor cells that can differentiate into osteoblasts and chondrocytes during normal bone growth and fracture healing. TGF-beta 1 and BMP-2 have been implicated in the regulation of the chondrogenic differentiation of these cells, but their roles are not fully defined. This study was undertaken to investigate the chondrogenic effects of TGF-beta 1 and BMP-2 on rat periosteum-derived cells during in vitro chondrogenesis in a three-dimensional aggregate culture. RT-PCR analyses for gene expression of cartilage-specific matrix proteins revealed that treatment with BMP-2 alone and combined treatment with TGF-beta 1 and BMP-2 induced time-dependent mRNA expression of aggrecan core protein and type II collagen. At later times in culture, the aggregates treated with BMP-2 exhibited expression of type X collagen and osteocalcin mRNA, which are markers of chondrocyte hypertrophy. Aggregates incubated with both TGF-beta 1 and BMP-2 showed no such expression. Treatment with TGF-beta 1 alone did not lead to the expression of type II or X collagen mRNA, indicating that this factor itself did not independently induce chondrogenesis in rat periosteal cells. These data were consistent with histological and immunohistochemical results. After 14 days in culture, BMP-2-treated aggregates consisted of many hypertrophic chondrocytes within a metachromatic matrix, which was immunoreactive with anti-type II and type X collagen antibodies. In contrast, at 14 days, TGF-beta 1 + BMP-2-treated aggregates did not contain any morphologically identifiable hypertrophic chondrocytes and their abundant extracellular matrix was not immunoreactive to the anti-type X collagen antibody. Expression of BMPR-IA, TGF-beta RI, and TGF-beta RII receptors was detected at all times in each culture condition, indicating that the distinct responses of aggregates to BMP-2, TGF-beta 1 and TGF-beta 1 + BMP-2 were not due to overt differences in receptor expression. Collectively, our results suggest that BMP-2 induces neochondrogenesis of rat periosteum-derived cells, and that TGF-beta 1 modulates the terminal differentiation in BMP-2 induced chondrogenesis.

Hyaluronan-based polymers in the treatment of osteochondral defects.

Articular cartilage in adults has limited ability for self-repair. Some methods devised to augment the natural healing response stimulate some regeneration, but the repair is often incomplete and lacks durability. Hyaluronan-based polymers were tested for their ability to enhance the natural healing response. It is hypothesized that hyaluronan-based polymers recreate an embryonic-like milieu where host progenitor cells can regenerate the damaged articular surface and underlying bone. Osteochondral defects were made on the femoral condyles of 4-month-old rabbits and were left empty or filled with hyaluronan-based polymers. The polymers tested were ACP sponge, made of crosslinked hyaluronan, and HYAFF-11 sponge, made of benzylated hyaluronan. The rabbits were killed 4 and 12 weeks after surgery, and the condyles were processed for histology. All 12-week defects were scored with a 29-point scale, and the scores were compared with a Kruskall-Wallis analysis of variance on ranks. Untreated defects filled with bone tissue up to or beyond the tidemark, and the noncalcified surface layer varied from fibrous to hyaline-like tissue. Four weeks after surgery, defects treated with ACP exhibited bone filling to the level of the tidemark and the surface layer was composed of hyaline-like cartilage well integrated with the adjacent cartilage. At 12 weeks, the specimens had bone beyond the tidemark that was covered with a thin layer of hyaline cartilage. Four weeks after surgery, defects treated with HYAFF-11 contained a rim of chondrogenic cells at the interface of the implant and the host tissue. In general, the 12-week defects exhibited good bone fill and the surface was mainly hyaline cartilage. Treated defects received significantly higher scores than untreated defects (p < 0.05), and ACP-treated defects scored significantly higher than HYAFF-11-treated defects (p < 0.05). The introduction of these hyaluronan-based polymers into defects provides an appropriate scaffolding and favorable microenvironment for the reparative process. Further work is required to fully assess the long-term outcome of defects treated with these polymers.

High variability in rabbit bone marrow-derived mesenchymal cell preparations.

The rabbit has been extensively used for preclinical models, especially in orthopedic applications. One of the more troubling features of this model is the high interindividual variability that is encountered and that requires a careful experimental design with sufficient sample size to make judgments valid. We have processed 241 individual preparations of rabbit bone marrow-derived mesenchymal progenitor cells (MPCs) over the last 3 years and have kept detailed records of the performance of these cells in various assays. This communication details the lack of correlation between the analyzed parameters. Bone marrow was harvested from 4-month-old rabbits; the cells were centrifuged, resuspended, and cultured. When cells reached 80% of confluence, they were removed from the plates with trypsin and assayed for their osteo- and chondrogenic potential. The average yield of the 241 individual MPC preparations exhibited a coefficient of variation of 77. An in vivo implantation assay with porous calcium phosphate ceramic cubes exhibited scores with a coefficient of variation of 65. Lastly, an in vitro assay of alkaline phosphatase enzyme activity exhibited the most variability with a coefficient of variation of 132. All of the cell preparations tested in an in vitro aggregate culture assay underwent chondrogenic differentiation. No relationships between any of these parameters were found. The variability of the results within the different assays is interpreted to be the result of the heterogeneity of the preparations. The lack of correlation between the parameters studied shows the importance of the conditions intrinsic to the different assays. These results serve to emphasize that any experimental design involving rabbit progenitor cells must include a sufficiently large sample size to allow statistically significant and rigorous conclusions.

Hyaluronic acid-based polymers as cell carriers for tissue-engineered repair of bone and cartilage.

Culture-expanded bone marrow-derived mesenchymal progenitor cells differentiate into chondrocytes or osteoblasts when implanted subcutaneously in vivo in combination with an appropriate delivery vehicle. This in vivo implantation technique is used to test new materials as putative delivery vehicles in skeletal tissue-engineering models. HYAFF 11 and ACP sponges, two biomaterials based on hyaluronic acid modified by esterification of the carboxyl groups of the glucuronic acid, were tested as osteogenic or chondrogenic delivery vehicles for rabbit mesenchymal progenitor cells and compared with a well characterized porous calcium phosphate ceramic delivery vehicle. The implant materials were examined by scanning electron microscopy for differences in pore structure or cellular interactions, were quantified for their ability to bind and retain mesenchymal progenitor cells, and were examined histologically for their ability to support osteogenesis and chondrogenesis after subcutaneous implantation into nude mice. The ACP sponge bound the same number of cells as fibronectin-coated ceramic, whereas the HYAFF 11 sponge bound 90% more. When coated with fibronectin, ACP and HYAFF 11 bound, respectively, 100 and 130% more cells than the coated ceramics. HYAFF 11 sponge composites retained their integrity after the 3 or 6-week incubation period in the animals and were processed for histomorphometric analysis. As a result of rapid degradation or resorption in vivo, ACP sponges could not be recovered after implantation and could not be analyzed. HYAFF 11 sponges presented more area available for cell attachment and more available volume for newly formed tissue. Following loading with mesenchymal progenitor cells and implantation, the pores of the sponges contained more bone and cartilage than the pores of ceramic cubes at either time point. Thus, relative to ceramic, HYAFF 11 sponges allow incorporation of twice as many cells and produce a 30% increase in the relative amount of bone and cartilage per unit area. Hence, the hyaluronic acid-based delivery vehicles are superior to porous calcium phosphate ceramic with respect to the number of cells loaded per unit volume of implant, and HYAFF 11 sponges are superior to the ceramics with regard to the amount of bone and cartilage formed. Additionally, hyaluronic acid-based vehicles have the advantage of degradation/resorption characteristics that allow complete replacement of the implant with newly formed tissue.

Chondroprogenitor cells of synovial tissue.

OBJECTIVE: To assess the chondrogenic potential of cells within the synovium. METHODS: Explants of synovium taken from various sites in the joint were embedded in agarose and cultured with transforming growth factor beta1 (TGFbeta1) to assess their chondrogenic potential. Isolated synovial cells were also tested for their chondrogenic potential by culturing them as aggregates in a chemically defined medium with TGFbeta1. Cartilage formation was determined with histologic staining and immunohistochemistry. The osteochondral potential of the isolated cells was also assessed after subcutaneous implantation of the cells, loaded into porous calcium phosphate ceramic cubes, in athymic mice. RESULTS: A total of 48 synovial explants were cultured in agarose with TGFbeta1. The formation of cartilage was observed in the outer region of 21 explants, and type II collagen was localized in that region by immunohistochemistry. A larger percentage of TGFbeta1+ explants from the inner synovium sites formed cartilage compared with those from the outer synovium sites. Chondrogenesis occurred in aggregates incubated with TGFbeta1 as early as day 7, and by day 14, all TGFbeta1+ aggregates demonstrated chondrogenesis. In contrast with the results of the in vitro aggregate assay for chondrogenesis, no formation of cartilage or bone was evident in any section containing synovial cell-loaded ceramic cubes that were harvested at either 3 or 6 weeks after implantation subcutaneously in athymic mice. CONCLUSION: Synovial explants and isolated synovial cells will undergo chondrogenesis when cultured in the presence of TGFbeta1. The data indicate a possible synovial origin for the chondrocytic cells found in rheumatoid pannus. Furthermore, these data are consistent with the clinical findings of synovial chondrogenesis leading to synovial chondromatosis.

Different response to osteo-inductive agents in bone marrow- and periosteum-derived cell preparations.

Rabbit bone marrow- and periosteum-derived cells were cultured in medium containing dexamethasone, bone morphogenetic protein-2 (BMP2) or both. The response of bone marrow-derived cells, measured as alkaline phosphatase expression, depended on the stage of growth. In subconfluent cultures, BMP2 had a greater effect than dexamethasone and treatment with both further increased enzyme activity. In confluent cultures, the effect of dexamethasone was greater than that of BMP2 and, when used together, they had an additive effect. The mineral deposition observed in these cultures did not have the typical structure of bone nodules. For periosteum-derived cells, dexamethasone did not increase the expression of alkaline phosphatase, while BMP2 did; treatment with both was less effective than treatment with BMP2 alone. Typical bone nodules were observed in cultures of periosteum-derived cells treated with dexamethasone and BMP2. These findings indicate that either osteoprogenitor cells from these two sources are intrinsically different or else non-progenitor cells in the preparations directly or indirectly affect the responsiveness to osteo-inductive agents.

The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells.

Mesenchymal progenitor cells provide a source of cells for the repair of musculoskeletal tissue. However, in vitro models are needed to study the mechanisms of differentiation of progenitor cells. This study demonstrated the successful induction of in vitro chondrogenesis with human bone-marrow-derived osteochondral progenitor cells in a reliable and reproducible culture system. Human bone marrow was removed and fractionated, and adherent cell cultures were established. The cells were then passaged into an aggregate culture system in a serum-free medium. Initially, the cell aggregates contained type-I collagen and neither type-II nor type-X collagen was detected. Type-II collagen was typically detected in the matrix by the fifth day, with the immunoreactivity localized in the region of metachromatic staining. By the fourteenth day, type-II and type-X collagen were detected throughout the cell aggregates, except for an outer region of flattened, perichondrial-like cells in a matrix rich in type-I collagen. Aggrecan and link protein were detected in extracts of the cell aggregates, providing evidence that large aggregating proteoglycans of the type found in cartilaginous tissues had been synthesized by the newly differentiating chondrocytic cells; the small proteoglycans, biglycan and decorin, were also detected in extracts. Immunohistochemical staining with antibodies specific for chondroitin 4-sulfate and keratan sulfate demonstrated a uniform distribution of proteoglycans throughout the extracellular matrix of the cell aggregates. When the bone-marrow-derived cell preparations were passaged in monolayer culture as many as twenty times, with cells allowed to grow to confluence at each passage, the chondrogenic potential of the cells was maintained after each passage.

[Repair of articular cartilage with biological tissues. An experimental study in sheep]. / Réparation du cartilage articulaire par matériaux biologiques. Etude expérimentale sur le mouton.

UNLABELLED: The purpose of this study was to assess the joint cartilage's capacity for repair and the potential of various biological tissues as replacements for damaged cartilage. METHODS: We operated 30.3 months old, lambs, creating a chondral lesion which was left untreated in group I and treated with a fresh chondral implant in group III, a frozen chondral implant in group IV, and a frozen periostal implant in group V; in group II the lesion extended as far as the subchondral bone. The lesions were performed in the loading area of the medial condyle of the knee. Follow-up time was 6 months, and the results were assessed histologically. RESULTS: In the chondral lesions which remained untreated (group I), degeneration of the exposed layers occurred, and loss of both cartilage thickness and homogeneity of the matrix was noted. Where the lesion extended as far as the subchondral bone (group II), repair was found to have taken place with a fibrous tissue indistinguishable from cartilage. When cartilage was implanted (group III and IV), the integration of the implant depended on wether there was any contact between the implant and the surrounding tissue. DISCUSSION: The integrity of the fresh implants was maintained better than that of the frozen ones, which were found to contain cells with a proliferative capacity. When periosteum was placed over the chondral lesion, we observed the formation of a very loose fibrous tissue in which the initial stages of differentiation could be appreciated in the deepest layers.

The effect of shockwaves on mature and healing cortical bone.

It has been proposed that high energy shockwaves could be used to create microfractures in cortical bone. This quality might be exploited clinically to perform closed osteotomies and promote healing in nonunion (15). However, no study has previously documented the effect of shockwaves on cortical bone "in vivo". We report an investigation designed to demonstrate the effect of shockwaves on mature cortical and healing bone. An osteotomy was performed on the tibiae of 37 lambs; two weeks later the operation site was exposed to shockwaves. Three weeks later the lambs were killed and specimens of the bone examined histologically and radiographically. Shockwaves had no effect on the periosteal surface of mature cortical bone, but on the endosteal surface some new trabecular bone was seen. Healing of bone was delayed by the shockwave therapy. We conclude that there is currently little place for shockwave treatment in clinical orthopaedics.