The present study demonstrated that, when each DN gel sheet or plug was implanted at the bottom of the defect so that a 2- to 3-mm deep vacant space was left in the defect, the evaluation scores were found to be the greatest. In this implantation condition, all the evaluation scores and the expression level of the type-2 collagen gene in the two groups with implantation of a 1.0-mm thick gel sheet or a 5.0-mm thick gel plug were significantly greater than those of the untreated control group, and there were no significant differences in those parameters between the two groups. These results suggested that the in vivo hyaline cartilage regeneration phenomenon induced by DN gel is not influenced by the thickness of the gel implanted at the bottom of a large osteochondral defect, at least, in a range between 1.0 mm and 5.0 mm. Concerning the 0.5-mm thick sheet implantation, however, the gross observation score and expression level of type-2 collagen gene were slightly but significantly lower as compared with the 1.0-mm thick gel sheet implantation, while there were no significant differences in comparison with the 5.0-mm thick gel implantation. These results suggested that, when the thickness of the implanted gel is too thin, below 1.0 mm, the thickness of the implanted DN gel slightly but significantly reduces induction ability of the in vivo hyaline cartilage regeneration phenomenon.
We have studied the mechanism of the in vivo cartilage regeneration induced by the PAMPS/PDMAAm DN gel. Our previous in vitro study  has shown that the single-network PAMPS gel can differentiate chondrogenic ATDC5 cells into chondrocytes even in an insulin-free medium. The PAMPS network in this DN gel is negatively charged and has a sulphonic acid base, being similar to proteoglycans in normal cartilage. The PAMPS network may work as an effective reservoir of signaling molecules and growth factors like proteoglycans. This PAMPS function may create an appropriate biological environment for immature cells proliferating in the defect space, leading to chondrogenic differentiation. Another possible mechanism of the induction effect by the PAMPS/PDMAAm DN gel is related to the biomechanical condition. It is also known that mechanical microenvironment significantly affects cartilage differentiation of bone-marrow derived stem cells [21, 22]. Recently, Engler et al.  reported that elasticity of the material on which the cultured cells are attached to directs stem cell differentiation. For example, elastic materials induce differentiation to the cartilage tissue, and stiff materials induce differentiation to the bone tissue. Therefore, the mechanical properties of the DN gel improved by the PDMAAm gel [18, 19] may enhance the effect of the PAMPS gel on cartilage regeneration. In addition, we have elucidated that a dynamic physical environment is a critical factor to induce cartilage regeneration . However, the mechanism of the in vivo cartilage regeneration induced by the PAMPS/PDMAAm DN gel has not been sufficiently clarified.
The present study has added new information in order to increase the database on the hyaline cartilage regeneration induced by the PAMPS/PDMAAm DN gel. In the present study, first, the 1.0-mm and 5.0-mm thick DN gel implants had the same ability in the in vivo hyaline cartilage regeneration, while the ability of the 0.5-mm thick DN gel implant was slightly but significantly inferior to that of the 1.0-mm thick DN gel implant. This fact implied that DN gel sheets thicker than 1.0 mm may provide the same physical environment under knee motion at the bottom of the defect with the 2- to 3-mm deep vacant space. Theoretically, however, when a very thin gel sheet is compressed by large forces, the apparent elastic modulus of the gel-bony floor (bone beneath the gel sheet) complex is increased by the effect of the extremely high modulus of the bony floor. Therefore, when the thickness of a DN gel sheet was 0.5 mm in this study, the apparent elastic modulus of the gel-bone complex might be too high under repetitive compression loading due to joint motion. Therefore, we speculate that the increased apparent modulus of the gel-bone complex under the repetitive compression loading may reduce the cartilage regeneration.
There are some limitations in this study. First, we used a rabbit patellofemoral joint model in this study, although this model has been commonly used to evaluate new procedures in cartilage regeneration experiments [25–27]. The joint which the most frequently requires a cartilage regeneration therapy is not the patellofemoral joint but weight-bearing regions of the femorotibial joint in the clinical field. Therefore, the results obtained cannot be simply transferred to the clinical field because different results might be obtained in the femorotibial joint model. There is a need to conduct experimental studies with a large animal model in order to fully evaluate the pre-clinical efficacy. The second limitation is that we did not actually examine the mechanical properties of the DN gel sheet or plug implanted in the bone tissue after cartilage repair occurs. The biomechanical changes of the DN gel may affect cartilage regeneration and durability because the implanted DN gel, which was seen as a void space in the histology section, was remained surrounded by bone in the defect as it was but it is unknown whether the DN gel properly works as subchondral bone with similar biomechanical properties after inducing cartilage regeneration. The third limitation is that we performed only short-term observations of the regenerated cartilage, because we intended to obtain the initial evidence that the DN gel’s thickness affected induction of the cartilage regeneration. However, we do not know the long-term effect of the DN gel on cartilage regeneration as well as the biological and biomechanical changes in the DN gel itself. To completely clarify the comprehensive effect of the DN gel thickness on the cartilage regeneration, long-term studies need to be conducted in the future. The fourth limitation is that no biomechanical evaluations of the regenerated cartilage were performed. We recognize that the biomechanical quality of cartilage is of vital importance for cartilage tissue restoration and durability. However, we believe that the biological and histological evaluations are fundamental to evaluate cartilage regeneration induced by synthetic gel implantation. The fifth limitation is that we did not evaluate the cell viability or extracellular matrix change of adjacent cartilage in the present study. As this cartilage repair method creates a large defect with a 2 to 3 mm depth of the vacant space at DN gel implantation, it might cause a significant mechanical change in the adjacent cartilage. We observed only minor influence on the adjacent cartilage in this animal model based on our previous study ; however, long-term studies are needed to clarify the effect of this method.
Concerning the safety of the PAMPS/PDMAAm DN gel as a biomaterial, we previously performed a pellet implantation test into the para-vertebral muscle , according to the guideline for biological evaluation of the safety of biomaterials, which had been published by the Ministry of Health, Labour and Welfare, Japan. Although this DN gel implantation induced a mild cell infiltration at 1 week, the degree of the inflammation significantly decreased to the same degree as that of the negative control at 4 and 6 weeks. We also cultured ATDC5 cells on the PAMPS/PDMAAm DN gel  as well as the single-network PAMPS and PDMAAm gels . No harmful effects due to these gel surfaces were detected. We believe that the PAMPS/PDMAAm DN gel is a safe biomaterial. However, we have not completed all protocols to establish the clinical safety of this DN gel as an implant. Further studies are needed to establish the clinical safety of this gel in the near future.
As to clinical relevance, the innovative strategy with the DN gel implantation for cartilage repair without cell culture has been proposed [15, 16]. We should note that this strategy is in sharp contrast to current prevailing strategies that completely fill the defect with the tissue-engineered cartilage tissue, cell-seeded scaffold material implantation, or acellular polymer scaffolds with signaling molecules [11–14]. In addition, we can expect this strategy to solve the various problems and concerns in current strategies, including donor site morbidity, two surgeries being needed, a long non-weight bearing period, a potential risk of zoonotic transmission, an enormous amount of money to establish such a tissue-engineering industry system, an extremely high medical fee for patients having the treatment, and other factors [29–36]. The results of the present study have provided important information to realize benefits of this innovative strategy with the DN gel in the near future. As described in the introduction section, a synthetic implant in the bone should be as thin and small as possible to avoid loosing the bone tissue due to the implantation. However, we do not know whether this strategy using a DN gel implantation is superior to other cartilage repair procedures such as the current autologous chondrocyte implantation technique. For the possible clinical use of this treatment strategy, further studies should be conducted to clarify these issues in the near future.