Simple combination of a wnt activator and a retinoic acid receptor agonist robustly induces human pluripotent stem cells into chondrocytes

Purpose: Human pluripotent stem cells (hPSCs), such as human embryonic stem cells and human induced pluripotent stem cells (hiPSCs), are a promising cell source for treating degenerative diseases such as osteoarthritis, because of their unlimited expansion potential. They are also valuable experimental tools for in vitro studies of cell differentiation including chondrogenesis. While some efficient methods using several kinds of cytokines have been developed for chondrocyte induction from hPSCs, a simple induction protocol to differentiate chondrocytes from hPSCs using small-molecule compounds is beneficial both for cartilage regenerative medicine and for mechanistic studies of chondrogenesis. Because canonical Wnt signaling regulates mesendoderm induction of PSCs, treatment of PSCs with Wnt activator such as CHIR99021, a glycogen synthase kinase 3 inhibitor, will be useful in the initial step of chondrogenesis. Moreover, while retinoic acid (RA) and retinoids are reported to suppress collagen production or cartilage matrix synthesis in cultured chondrocytes, they are also known to play critical roles in limb bud formation and subsequent chondrogenesis during skeletal development. Therefore, RA and retinoids may be useful for inducing chondrocytes from PSCs under a certain condition. The purpose of the present study is to demonstrate that chondrocytes are robustly induced from hPSCs by simple combination of two compounds (2C), CHIR99021 and TTNPB, a retinoic acid receptor (RAR) agonist, under serum- and feeder-free conditions within 5-9 days. We also aim to reveal molecular functions of RA and Wnt/β-catenin signaling in the process of chondrogenesis by genome-wide analysis of RAR and β-catenin association.

Methods: To confirm reproducibility of results, multiple clones of hiPSCs adapted to feeder-free defined culture conditions were used for experiments. At least three independent experiments were performed for gene expression analyses, and P-values < 0.05 by Dunnett’s test were considered as significant for multiple group comparisons. FACS analysis was performed with a FACSAria Fusion cell sorter, and data were analyzed by BD FACSDiva software. To form a hiPSC-derived particle for transplantation, 1.7 × 10⁶ hiPSC-derived cells that had undergone the present protocol for 9 days were seeded into cloning rings set on a permeable membrane insert. After 1 week of culture, the formed particles were transplanted into subcutaneous spaces or knee cartilage defects of 8-week-old male NOD/SCID mice. The mice were sacrificed after 8 weeks for subcutaneous transplantation, and after 6 months for transplantation in knees. Microarray analysis was performed using n = 3 sequential samples from independent experiments during the 2C differentiation, and gene ontology (GO) terms with corrected P-values of < 0.01 were considered as significant in GO analysis. For open chromatin analysis, assay for transposase-accessible chromatin using sequencing (ATAC-seq) was performed using two biological replicates of hiPSC-derived cells at each stage of the 2C protocol. ChIP-seq of RARα and Wnt/β-catenin were also performed at each differentiation stage with two biological replicates per condition. A P-value cut-off of 1 × 10^-5 compared with input controls was used for peak calling, and peaks with a false discovery rate of < 0.01 were incorporated into further analysis.

Results: The optimized standard protocol was combined treatment with 10 μM CHIR99021 for the initial 2 days, and 100 nM TTNPB for the whole period. Under this protocol, expression of mesendoderm markers T and MIXL1 was upregulated from day 1 to 2, and expression levels of mesoderm markers, such as TBX6, MEOX1 and HAND1, were elevated from day 2 to 4. Then, chondrogenic markers including SOX9, SOX5, SOX6, COL2A1 and COL11A2 were upregulated after about day 5. FACS analysis of hiPSC-derived differentiated cells after 9 days of culture under the present protocol showed that 97.0% of cells were positive for SOX9 while NANOG- or OCT4-positive cells were not totally detected. We compared marker gene expression between the hiPSC-derived cells on day 9 of the 2C protocol and those after differentiation by two other reported protocols using cytokines for about 2 weeks; expression levels of chondrogenic markers were significantly increased by 2C differentiation compared with the other protocols, while those of pluripotent and other lineage markers were significantly decreased. Particles prepared from hPSC-derived cells differentiated by the 2C protocol formed hyaline cartilaginous tissues when transplanted into subcutaneous spaces and articular cartilage defects in NOD/SCID mouse knee joints, and no signs of teratoma or other tumor formations were seen. GO analysis using upregulated genes in day 9 samples compared with those of day 0 in microarray datasets showed that terms related to skeletal and cartilage formation were enriched at highest ranks. ATAC-seq of sequential samples under the present protocol demonstrated that enhancer regions of key marker genes for mesendoderm, mesoderm, and chondrocyte were activated at each differentiation stage. Finally, ChIP-seq analysis for RAR and β-catenin detected peaks in the enhancer regions of the key marker genes for the respective stages, and some of which were common between the signals.

Conclusions: The 2C protocol induces chondrocytes more efficiently compared with reported protocols using cytokines. The present method provides a promising cell source for cartilage regenerative medicine and may facilitate elucidation of molecular mechanisms underlying chondrocyte differentiation. The findings displayed by the RAR and β-catenin ChIP-seq datasets show that RA and Wnt/β-catenin signals are cooperatively involved in direct regulation of chondrogenesis.