In the first sub-study, we confirmed the basic functions of the custom-made PAMPS and PDMAAm gels used in the present study. Even in the maintenance medium, expression of type-2 collagen and aggrecan genes in the cells cultured on the PAMPS and PDMAAm gels was significantly greater than that in the cells cultured on the polystyrene surface. In addition, the former expression level by the synthetic gels was comparable to the expression level of the cells cultured on the polystyrene surface in the differentiation medium with insulin. These results provided a basis concerning the material function to the second and third sub-studies in the present study. It is remarkable that the synthetic gels can induce chondrogenic differentiation in vitro without the existence of any cytokines. No biomaterials having such effects are reported except for these gels. We can speculate that some molecular signals from each hydrogel surface may be delivered to a nucleus of the ATDC5 cells via certain receptors on the cell surface, resulting in induction of chondrogenic differentiation of the cells. In the present study, the effects of the two synthetic gels on the ATDC5 cells were not identical. Therefore, the PAMPS and PDMAAm gels may deliver molecular signals to different receptors on the ATDC5 cell. However, it is unknown what signal transduction system exists between the synthetic gels and the ATDC5 cell. Further molecular biological studies are needed to clarify the signal transduction.
Benya et al.  reported that dedifferentiated chondrocytes re-express the differentiated collagen phenotype when cultured on a hydrophobic surface. Namely, they observed that decreased cell attachment promoted collagen type II gene expression. This phenomenon is similar to the behavior of the ATDC5 cells cultured on the PDMAAm gel in the present study. However, the PDMAAm hydrogel is not hydrophobic, because the contact angle to water is 47° , while the contact angle a cell culture-treated PS used in the present study is 67°. When the contact angle to water is less than 90°, the surface is hydrophilic. Nevertheless, Yang et al.  showed that the dedifferentiated chondrocytes re-express the differentiated collagen phenotype when cultured on the PDMAAm gel. They also showed that the chondrocytes shows low adhesion to the PDMAAm gel, because it is a neutral hydrogel without charged moieties on its polymer chains . We speculate that the similarity in the results between the present study and the Benya’s study may be explained by the low adhesion property of the PDMAAm gel to the chondrogenic cells. On the other hand, the contact angle of the PAMPS gel is only 12 degrees , and the PAMPS gel showed high adhesion property to the ATDC5 cells in the present study. Nevertheless, the PAMPS gel showed high ability that differentiated the ATDC5 cells to chondrocytes. This ability cannot be explained by the low adhesion property. It is considered that the mechanism that the PAMPS gel differentiated the ATDC5 cells to chondrocytes is different from the mechanism that the PDMAAm gel did it. Further studies are needed to clarify the mechanism of the PAMPS and PDMAAm gels that can induced chondrogenic differentiation of the ATDC5 cells.
In the second and third sub-studies, we found that the in vitro induction effects of the PAMPS and PDMAAm gels on chondrogenic differentiation of ATDC5 cells were significantly affected by HA, depending on the level of concentration. Namely, supplementation of a relatively low concentration (0.01 and 0.10 mg/mL) of HA significantly enhances the chondrogenic differentiation phenomenon induced by the PAMPS gel. On the other hand, supplementation of a relatively high concentration (1.00 mg/mL) of HA significantly reduced the chondrogenic differentiation phenomenon by the PDMAAm gel. It is known that isolated application of HA to the ATDC5 cells cultured on a polystyrene dish does not induce chondrogenic differentiation. Therefore, these sub-studies suggested that HA supplementation significantly affected the cells which had received certain signals concerning chondrogenic differentiation from the PAMPS or PDMAAm gel. The in vitro effects of HA supplementation on the chondrogenic differentiation phenomenon, which was shown in the first sub-study, were different between the PAMPS and PDMAAm gel surfaces. It is known that HA affects chondrocytes through signal transduction receptors that exist on the cell surface, depending on the concentration [15, 16]. Because the PAMPS and PDMAAm gels are considered to deliver molecular signals to different receptors on the ATDC5 cell, as implied in the first sub-study, the receptor difference may explain the difference of the HA supplementation effect between these gels.
HA differs from the other glycosaminoglycans, because it has a high molecular weight and it is capable of forming entangled networks. Concerning the effect of HA on chondrogenesis, previous studies reported that HA significantly affects not only gene expression of type-2 collagen and aggrecan but also proteoglycan synthesis in in vitro and in vivo conditions, depending on the concentration. For example, Akmal et al.  reported that supplementation of a low concentration of HA enhances the glycosaminoglycan synthesis in cultured chondrocytes. On the other hand, Allemann et al.  studied the effects of HA soaked into a three-dimensional scaffold on bovine chondrocyte function in vitro, and reported that a high concentration of HA inhibits the cellularity and matrix accumulation. In addition, recent studies reported that HA has a regulatory role in the maintenance of ESCs in their undifferentiated state in vitro[19, 20]. Frean et al.  demonstrated that HA enhances proteoglycan synthesis of the cartilage tissue, depending on the concentration: a low concentration of HA has stimulatory effects on matrix production, while a high concentration of HA does not show such stimulatory effects. They concluded that there is an optimal concentration in the effect of HA on the chondrogenesis. These reports appear to support the results obtained in the second and third sub-studies.
HA has not only pharmacological effects but also physiochemical effects, the concentration of HA affects its physiochemical property . Concerning the reduction effect of HA with a high concentration, as shown in the third substudy, we can consider a few physical effects of HA. For example, it is known that HA affects cell-substrate adhesion, cell migration, and cell proliferation through signal transducing receptors such as CD44 [23–27]. Therefore, there is a possibility that HA of a high concentration might physically coat the cells and reduce the sensitivity of these receptors, resulting in a reduction of transduction of important molecules related to cell differentiation. Also HA of a high concentration might change the physical conditions of the PDMAAm gel surface, the stiffness of which is closer to the normal cartilage than that of the PAMPS gel. The physical changes might reduce the capability of the PDMAAm gel to induce chondrogenic differentiation of the ATDC5 cells, because the mechanical properties of a material surface significantly affect culture results . Moreover, recent studies reported that HA has a regulatory role in the maintenance of ESCs in their undifferentiated state in vitro[19, 20]. The regulatory function of HA might maintain or reduce the chondrogenic differentiation of the ATDC5 cells in the case of induction with the PDMAAm gel.
There are some limitations in this study. First, we cultured the cells for only 7 days. As the period for cell differentiation depends on the type of cells, the long term culture may show a different result. Based on our previous study , however, we determined that day 7 was the best time to compare the effects of HA supplementation on the cells, as described in the study design section. Therefore, we believe that the 7th day of observation is appropriate for this study. Secondly, we performed PCR analyses concerning only type-2 collagen and aggrecan in the cultured ATDC5 cells, because the over-expression of these mRNAs were accompanied by over-expression of type-2 collagen and aggrecan molecules in the same culture system used in our foregoing study . However, we cannot refer to the effect of HA supplementation on other important molecules, such as SOX9, type-10 collagen, and so on, from this study. Thirdly, we did not count the number of cells cultured on the different discs. For example, although the cultured cells attached onto the PDMAAm gel formed nodules, many cells did not attach onto the gel surface. The data may be explained by the low adhesion property of the PDMAAm gel surface to the ATDC5 cells. Also there is a possibility that HA is inducing more cell detachment from the substrate. Fourthly, we did not assess other culture conditions that might affect chondrogenic differentiation of ATDC5 cells; eg, biomechanical factors, and other stimulating factors. In this study, however, because we intended to focus on the effect of HA concentration on chondrogenic differentiation, we believe that this study design is acceptable. Fifthly, we did not clarify the effect of HA on other culture systems including different cells. Thus, further studies will be needed to clarify the effect of HA supplementation on chondrogenic differentiation induced not only by the PAMPS and PDMAAm gels but also other synthetic materials in the near future. However, the present study has provided important evidence that in vitro HA supplementation can significantly modify the chondrogenic differentiation induced by the PAMPS and PDMAAm gels in a complex manner, depending on the concentration.