Lubricin is believed to be an effective therapeutic agent for osteoarthritis and synovitis. However, clinical application of recombinant or synthesised lubricin has not been attempted yet. In this study, we successfully developed a protocol to induce lubricin-expressing cells from pluripotent stem cells and confirmed their differentiation into Prg4-expressing cells that shared the characteristics of SFZ chondrocytes and FLSs from the joint interzone.

Prg4 is expressed in SFZ chondrocytes, synovial cells, synovial fluid, tendons, and tendon sheath [3, 4]. Such localisation was confirmed in our Prg4-mRFP1 transgenic mice. First, iPSCs were established by transducing ear tip mouse fibroblasts with all four Yamanaka factors [26]. Second, we devised a protocol for the differentiation of iPSC-derived Prg4-positive cells, focusing on differentiation into SFZ cells and FLSs, which formed the synovial joint rather than cartilage. To induce the primitive streak, and subsequent differentiation into paraxial mesoderm and somites, we followed previous reports on embryoid body formation in suspension or adhesion for 4–6 days [11, 12, 15]. To maximise differentiation, we chose embryoid body formation in suspension for 5 days. Based on known cytokines involved in this process [12, 14, 15], we selected Wnt3a, activin A, and bFGF to promote differentiation of the sclerotome as a precursor to the synovial joint. Although useful for induction of the primitive streak, bone morphogenetic protein (BMP) 4 was not applied in this case as it could not lead to differentiation of the paraxial mesoderm.

For chondrogenic differentiation, most studies have employed serum-free medium [13,14,15] or 1% foetal bovine serum [12], as well as the cytokines ITS, bFGF, BMP2, TGF-β1, TGF-β3, growth differentiation factor 5, TD-198946, and ascorbic acid. Here, we cultured iPSCs in medium containing 10% foetal bovine serum, ITS, bFGF, and TGF-β1 to differentiate Prg4-positive SFZ cells and FLSs of the synovial joint.

To confirm successful differentiation, we analysed mRNA expression of iPSC-derived differentiated cells over time using qRT-PCR. Wnt9a and Wnt16 are important in the joint interzone, for the development of synovial joints [41,42,43]. Wnt16 also supports the phenotype of SFZ progenitor cells and lubricin expression [56]. During differentiation, Wnt9a and Wnt16 showed significant upregulation on day 12, followed by a drop on day 21. This expression dynamics suggested that iPSCs differentiated in the joint interzone on day 12 and advanced to the next differentiation stage thereafter.

Besides Prg4, other genes expressed in SFZ chondrocytes include Erg, Tnc, and Runx1 [44,45,46,47,48]. We confirmed increased expression of Prg4, Erg, and Tnc on day 21, but could hardly detect them on day 12. Cdh11, Col4a, Cd55, Vcam1, Icam1, and Cd248 are expressed in FLSs of synovial joints [49,50,51,52,53,54,55]. Here, SFZ chondrocyte markers Prg4, Erg, and Tnc exhibited a similar expression pattern as FLS markers Cdh11, Col4a, and Vcam1. These results suggest that iPSCs differentiated into Prg4-expressing cells via the joint interzone. Analysis of mRNA expression in ESC-derived differentiated cells produced a similar temporal pattern as in iPSCs, suggesting that our protocol was applicable to all pluripotent stem cells. RNA-sequencing confirmed that iPSC-derived Prg4-positive cells (clone #2–9) were more similar to SFZ chondrocytes and FLSs than to deep zone chondrocytes, mesodermal cells, myogenic cells, and MLSs, suggesting that iPSC-derived Prg4-positive cells shared the characteristics of both SFZ chondrocytes and FLSs in synovial joints. Indeed, SFZ chondrocytes and FLSs are representative lubricin-expressing cells, showing higher expression than deep chondrocytes, mesoderm cells, myogenic cells, and MLSs. Therefore, the results of this study are significant in that high lubricin-expressing cells with SFZ chondrocytes and FLS characters were induced.

Recent strategies in regenerative therapy favour the use of stem cells rather than autologous cartilage transplantation to replenish missing or damaged cartilage [10, 11]. Examples of such an approach include the generation of articular cartilage using human ESCs [15], generation of scaffold-free hyaline cartilage from human iPSCs [12], generation of Col2a1-EGFP iPSCs for monitoring chondrogenic differentiation [13], the formation of stable human articular cartilage using human MSCs [16, 17], cartilage repair using a scaffold-free construct derived from porcine synovial MSCs [18], and specification of chondrocytes and cartilage tissue from murine ESCs [14]. The present study aimed to provide lubrication for arthritic joints and tenosynovitis by transplanting Prg4-expressing cells rather than regenerating damaged cartilage tissue. Hence, we injected iPSC-derived Prg4-mRFP1-positive cells into the paratenon surrounding the Achilles tendon and knee joints of SCID mice and confirmed the cells’ survival and lubricin expression in vivo. A recent study revealed that Prg4-lineage cells in the joint SFZ differentiated into articular chondrocytes [57], suggesting that injection of iPSC-derived Prg4-expressing cells might also have the potential to regenerate articular cartilage. Major advantages of our approach are the potential to obtain a large number of Prg4-expressing cells by using iPSCs, that harbour high capacity of self-renewal. Regarding the cell source, we can use mesenchymal stem cells with Prg4-mRFP1 reporter system and develop differentiation protocol from MSC to Prg4-positive cells. Differentiation efficiency of MSCs to Prg4-positive cells could become better because usage of MSCs can skip the process of iPSC to mesenchymal lineage differentiation. Mature Prg4-expressing cells could also be available, however, it may be difficult to obtain and expand a large number of Prg4-expressing cells in clinical practice, especially in the elderly patients.

Finally, several hurdles need to be overcome prior to the clinical application of Prg4-expressing cells. First, disadvantage of our approach is the insufficient induction efficiency of Prg4-mRFP1 positive cells, with average 4.9% (maximum 9.3%). Also, using the same protocol, we previously induced tenocyte-like cells from Scx-EGFP iPSCs [58]. Tenocytes share the same developmental process through somite to sclerotome with chondrocytes and synovial cells. Given that Prg4-expression has been confirmed in tendons, iPSC-derived Prg4-positive cells in this study might contain tenocytes. Therefore, the differentiation protocol should be optimised to improve the induction efficiency and purity of Prg4-epressing cells presenting SFZ and FLS characteristics. Second, we need to prolong the survival of these cells after injection. Transplanted Prg4-mRFP1-positive cells were detected in the paratenon surrounding the Achilles tendons and knee joints, but only for up to 3 days. Use of scaffold materials or platelet-rich plasma might offer a physical means for improved temporal survival of iPSC-derived Prg4-expressing cells [59, 60]. Also, condition media containing cytokines, growth factors, and exosomes has been used for regenerative medicine and many studies have shown their efficacy [12]. Indeed, our previous study regarding Scx-EGFP expressing iPSC-derived tenocyte transplantation revealed that cell transplantation had dual role, a direct contribution for regenerative tendon and an indirect effect (paracrine effect) reducing scar formation [58]. Of note, paracrine effect was presumed to contribute more preferentially for regeneration, indicating that cytokines, growth factors, and exosomes released from induced cells have an important role in regeneration. Therefore, the condition media of Prg4-mRFP1 expressing cells could become an alternative regenerative agent against osteoarthritis and synovitis, and might also prolong the survival of Prg4-mRFP1 positive cell after transplantation. Moreover, transplantation of Prg4-expressing cells might be beneficial in terms of such paracrine effect other than lubricin compared to rhPRG4 administration therapy. Third, to confirm engraftment of iPSC-derived Prg4-expressing cells in vivo, we used SCID mice for the transplantation experiment to avoid immune rejection. However, transplantation for osteoarthritis or synovitis model could be required to confirm the treatment effect of iPSC-derived Prg4-expressing cells. We would need further experiments to reveal its therapeutical benefit in the future.