Affiliation:
1. Leiden University Medical Center Department of Biomedical Data Sciences: Leids Universitair Medisch Centrum Afdeling Biomedical Data Sciences
2. Leiden University Medical Center Department of Anatomy and Embryology: Leids Universitair Medisch Centrum Afdeling Anatomie en Embryologie
3. Leiden University Medical Center Department Cell and Chemical Biology
4. Leiden University Medical Center Department of Orthopeadics
5. Technical University Eindhoven Department of Biomedical Engineering
6. LUMC: Leids Universitair Medisch Centrum
Abstract
Abstract
Background
Lack of insight into factors that determine purity and quality of human iPSC (hiPSC)-derived neo-cartilage precludes applications of this powerful technology towards regenerative solutions in the clinical setting. Here we set out to generate methylome-wide landscapes of hiPSC derived neo-cartilages from different tissues-of-origin and integrated transcriptome-wide data to identify aberrant set-points of transcription.
Methods
We applied in vitro chondrogenesis using hiPSCs generated from two different tissue sources; skin fibroblasts and articular cartilage. Upon differentiation towards chondrocytes these are referred to as hFiCs and hCiC, respectively. Genome-wide DNA methylation and RNA sequencing datasets were generated of the hiPSC-derived neo-cartilages, and the epigenetically-regulated transcriptome was compared to that of neo-cartilage deposited by human primary articular cartilage (hPAC).
Results
Methylome-wide landscapes of neo-cartilages of hiPSCs reprogrammed from two different somatic tissues were 85% similar to that of hPACs. By integration of transcriptome-wide data, changes in transcriptionally active CpGs between hCiC relative to hPAC were prioritized. Among CpG-gene pairs lower expressed in hCiCs relative to hPACs, we identified genes such as MGP, GDF5, and CHAD enriched in closely related pathways and involved in cartilage development that likely mark phenotypic differences in chondrocyte states. Vice versa, among CpG-gene pairs higher expressed, we identified genes such as KIF1A or NKX2-2 enriched in neurogenic pathways and likely reflected off target differentiation.
Conclusions
We did not find significant variation between the different tissue sources of the hiPSCs, suggesting that application of a robust differentiation protocol such as we applied here is more important as compared to the epigenetic memory of the cells of origin. Results of our study could be further exploited to improve quality, purity, and maturity of hiPSC derived neo-cartilage matrix, ultimately to realize introduction of sustainable, hiPSC derived neo-cartilage implantation into clinical practice.
Publisher
Research Square Platform LLC