This study found improvement in osteochondral healing mediated by an in situ photocrosslinked hydrogel containing leukocyte-reduced PRP or allogeneic ASCs, with a synergistic effect of combining PRP and ASCs, as assessed by gross appearance, biomechanical properties, histological and immunohistochemical characteristics, and subchondral bone volume. These findings are consistent with previous studies. Notably, Hsu et al.[24] reported that intra-articular injection of both ASCs and platelet-rich fibrin releasate, compared to either alone, improved the gross and histological appearance of healing osteochondral defects in a rabbit model. Likewise, Spakova et al.[25] found improved gross and histological characteristics of healing osteochondral defects in a rabbit model when treated with locally adherent ADSCs and subsequent PRP injections, as compared to microfracture with PRP injections or no treatment. While these results were promising, biomechanical properties, immunohistochemical characteristics, and subchondral bone quality were not assessed. This study is the first to examine the individual and combined effects of PRP and ASCs encapsulated in a photocrosslinkable hydrogel as single-stage treatment for osteochondral defects.
Hydrogels have been used widely for osteochondral tissue engineering, given their ability to deliver high concentrations of homogeneously distributed cells while being amenable for in situ gelation within irregularly shaped defects [39–42]. We chose to use the photocrosslinked methacrylated gelatin hydrogel (GelMA) as a bioscaffold in this study. The advantages of GelMA were highlighted in in vitro experiments by Lin et al. [11], demonstrating that visible light (VL)-based photocrosslinking of the GelMA hydrogel permitted cell encapsulation that led to desirable properties for articular cartilage repair, including cytocompatibility, chondroconductivity, and tunable mechanical properties. In addition, GelMA hydrogels can be rapidly photocrosslinked by VL, an easily adaptable technique for arthroscopic procedures. The promotion of chondrogenesis on MSC-seeded GelMA was previously demonstrated in vitro by Rothrauff et al. [10], who showed upregulated gene expression of chondrogenic markers ACAN, COL2, and SOX9.
In addition to GelMA, other photocrosslinked hydrogel products have been investigated. For example, Pascual-Garrido et al. manufactured a photopolymerizable cartilage mimetic hydrogel modified by chondroitin sulfate and arginyl-glycyl-aspartic acid to enhance cell adhesion and provide chondrogenic cues [43]. Qi et al. designed a sericin methacryloyl (SerMA)-based UV crosslinking hydrogel, which was adhesive to chondrocytes and promoted the proliferation of attached chondrocytes even under a nutrition-deficient condition [44]. In vivo implantation of chondrocyte loaded SerMA hydrogels promoted cartilage formation. Absent direct comparison of differing hydrogels, identification of the optimal biomaterial for osteochondral tissue regeneration remains to be elucidated. Given our laboratory’s extensive experience with GelMA, we used this established biomaterial to explore the additive effects of PRP and/or ASCs for osteochondral regeneration.
PRP, with its myriad bioactive factors that participate in the innate wound healing process, has been investigated as an orthobiologic for osteochondral regeneration [45]. While intraarticular injection of PRP has been recently shown to clinically improve pain and knee function in early osteoarthritic knees [46], likely mediated by the anti-inflammatory and immunomodulatory effects of PRP, its efficacy in promoting neotissue formation in cartilage or osteochondral lesions remains uncertain and has been largely explored only in in vitro and animal models. Notably, while some studies have found PRP to promote chondrogenic differentiation of MSCs in vitro [13], others have found a neutral [16, 47] or inhibitory effect on chondrogenesis [15]. Similarly, treatment of osteochondral defects by PRP alone has yielded equivocal results [13, 48], as this study found GelMA + PRP tended to improve osteochondral healing but seldom reaching statistically significant differences compared to controls.
Rather, the combination of ASCs + PRP produced superior effects, suggesting that the ASCs improved healing either by contributing to neotissue formation and/or secreting paracrine factors that further facilitated endogenous repair beyond the bioactive mediators found in PRP. While the mechanism underlying the added benefit of including ASCs was not investigated, MSCs are principally known to improve healing through paracrine signaling, yet a related study in which MSCs were delivered to an osteochondral defect by encapsulation in a PRP gel did find these exogenous cells still present at 9 weeks post-implantation [13]. Despite the additive benefit of including both ASCs and PRP in the photocrosslinked hydrogels, the regenerated osteochondral tissue did not fully recapitulate the appearance, biomechanical properties, histological characteristics, or bone volume of healthy, uninjured tissue. While ASCs were utilized herein because they are easier to isolate and present at a higher density per tissue volume than BMSCs, they may not represent the ideal MSC source for osteochondral repair. In related studies, Xie et al. [13] found BMSCs to be superior to ASCs in terms of PRP-modulated in vitro chondrogenesis and in vivo osteochondral repair, while Wang et al. [49] found cartilage-derived chondrogenic progenitor cells to be superior to both BMSCs and chondrocytes.
In addition to identification of the ideal cell source for osteochondral regeneration, modification of PRP composition may be necessary. Xu et al. [14] found leukocyte-reduced PRP was better than leukocyte-rich PRP in improving cartilage regeneration in a similar rabbit model, in part motivating the decision to used leukocyte-reduced PRP in this study. Furthermore, while PRP contains numerous growth and bioactive factors involved in innate wound healing, enrichment or neutralization of particular components may permit optimization of PRP for a specific tissue. For instance, TGF-β1 adversely contributes to fibrosis during muscle healing, with neutralization of TGF-β1 in PRP shown to enhance muscle regeneration compared to unaltered PRP [50]. Given the avascular nature of cartilage and the role of angiogenesis in osteoarthritis pathogenesis, inhibition of angiogenic promoters, such as VEGF-1, may further enhance PRP-mediated cartilage regeneration, as is being explored in ongoing work.
Although articular cartilage is known to be a poroelastic material [51, 52], a variety of models have been applied to elucidate the mechanical properties of cartilage in experimental conditions similar to our study [32]. For a rigid flat-ended cylindrical indenter, the commonly applied Hertz theory of linear contact mechanics predicts a linear load-displacement relationship during the loading phase [53]. Our indentation of repair tissue at defect sites demonstrated linear loading profiles and could be suitably fitted using linear elastic theory. Based on these findings, we calculated tangent elastic moduli using the last 35% of the loading phase curve. We additionally sought to capture the behavior of repair tissue after stress relaxation to an equilibrium state, utilizing poroelastic theory to calculate equilibrium elastic moduli after the completion of stress relaxation during the hold phase. Tangent elastic moduli are usually reported to be higher than equilibrium elastic moduli [32], as they represent the dynamic response of cartilage to increasing strain before the material is able to relax. These values are also specific to the rate of loading (25 µm/s in this study). By contrast, equilibrium elastic moduli represent the way cartilage responds to a prolonged static load, such as during standing. To gain a more clinically relevant picture of cartilage mechanics, both dynamically measured tangent moduli and static equilibrium moduli should be reported when possible. Here, our results showed no significant differences in tangent elastic moduli between the Sham group and the GelMA + ASCs and GelMA + ASCs + PRP groups, suggesting that incorporation of ASCs assists in restoring the ability of the neotissue to bear stresses similarly to native tissue. However, all repair groups showed lower equilibrium elastic moduli than the Sham group, suggesting that repair matrices are less able to bear stress once relaxation has occurred and water has evacuated the matrix. Modulation of the mechanical microenvironment of the healing osteochondral tissue, such as the use of biomaterials with different mechanical properties and/or controlling in vivo loading through the use of bracing or time-dependent rehabilitation protocols, may further improve the restoration of native osteochondral biomechanical properties [54, 55].
This study was not without limitations. While the rabbit model used herein has been commonly employed in similar studies, it does not fully recapitulate the mechanical forces and healing microenvironment of the human, which may be better accomplished through the use of a larger animal model. Furthermore, only a single timepoint was explored, limiting our understanding of the effect of the studied constructs on osteochondral regeneration over time. A single ASC density (20 x 106 cells/mL) was used; different concentrations may impact osteochondral regeneration, although the studied concentration was previously shown to produce abundant cartilage matrix [56]. The platelet concentration in PRP, while enriched over whole blood, was not controlled per se, and differing platelet concentrations may affect PRP bioactivity. Finally, the mechanism of efficacy of ASCs and/or PRP was not explored.