Integrated multi-omics reveal polycomb repressive complex 2 restricts human trophoblast induction

By Irene Talon · Ph.D. Student at KU Leuven, see authors

The Laboratory of Cellular Reprogramming and Epigenetic Regulation (Prof. Vincent Pasque) at KU Leuven has recently published an article in Nature Cell Biology on a multi-omics study to uncover the proteome, transcriptome, and investigate the cellular plasticity of human stem cells. The VSC has been a key tool for this research.

The lab of Dr. Vincent Pasque at the KU Leuven, together with the labs of Dr. Peter Rugg-Gunn (Babraham Institute, UK), Dr. Hendrik Mark (Radboud University, the Netherlands), and Dr. Maarten Dhaenens (University of Ghent, Belgium) and Dr. Nicolas Rivron (IMBA, Austria) apply an integrated multi-omics approach to comprehensively map the chromatin associated proteome, histone post-translational modifications, and transcriptome of human stem cells. In this study, two types of human stem cells were compared, stem cells that resemble the preimplantation state of the human embryo (naïve stem cells) and the post-implantation state (primed stem cells).

A total of 4576 proteins were identified, of which 1819 significantly changed between the two stem cell states. Out of the 43 individual histone post-translational modifications identified, PRC2-mediated H3K27me3, a modification that promotes gene silencing, was the focus of this study. H3K27me3 is the histone post-translational modification that most changed between the naïve and the primed stem cell states, being significantly more enriched in the naïve state. However, it was an open question why this repressive mark is highly abundant in the naïve state.

The VSC has been a key tool for this research, thanks to the power of the VSC Genius. -Irene Talon (KU Leuven)

We used a method known as calibrated CUT&RUN to determine the genome-wide distribution of H3K27me3 was enriched. With this technology, we obtained several genes that were enriched with H3K27me3, importantly, several of these genes belonged to lineage-determining genes, such as the placenta precursor lineage, known as the trophoblast lineage.

We hypothesized that if we block the deposition of this repressive mark, we could more efficiently get trophoblast cells. We performed this experiment and looked at the gene expression of all genes of 7629 single cells during the conversion of naïve stem cells into trophoblast cells. Here, the VSC Genius Tier-2 cluster was used to analyze gene expression and conclude that when H3K27me3 is chemically inhibited, trophoblast cells are increased. In addition, using different computational analyses, we could observe that these placenta precursor cells were comparable to the in vivo human embryo placenta precursor cells thanks to the power of the VSC Genius.

Moreover, we have also used 3D human embryo models (blastoids), created in the lab from a group of stem cells. Blastoids effectively form and recapitulate the first days of human life with a similar developmental progression. We have shown that we can accelerate the placenta formation in blastoids upon removal of the H3K27me3 repressive mark.

In conclusion, we discovered that PRC2-mediated H3K27me3 acts as a lineage gatekeeper in naïve stem cells. Our work sheds light on how the developmental plasticity of human naïve pluripotent stem cells is regulated.

Acknowledgments We thank J. Ooghe and the Vlaamse Supercomputer Center Leuven ( for computing and the VIB/KU Leuven Center for Brain and Disease Research, in particular to K. Davie and S. Aerts for including the scRNA-seq data on Scope. We thank the UZ/KU Leuven Genomics Core ( for high-throughput sequencing expertise, S. Schlenner and the KU Leuven/VIB FACS Core, the personnel of the KU Leuven Animal Facilities, J. C. Marine’s laboratory for the use of Tapestation and 10X Chromium Controller and in particular G. Bervoets for helpful feedback. We thank A. Smith for providing naive reset H9 hESCs and C. Verfaillie for providing Sigma hiPSCs. We thank L. David for helpful discussions; and B. Thienpont, L. Roderick and T. Voet for helpful feedback during thesis committee meetings. The V.P. laboratory is part of the Leuven Single-Cell Omics Institute (LISCO) and Leuven Stem Cell Institute (SCIL). We thank F. Krueger from Babraham Bioinformatics for sequencing quality control and mapping analysis, P. Kokko-Gonzales and A. Edwards at the Babraham Institute Next Generation Sequencing Facility, the Wellcome–MRC Cambridge Stem Cell Institute Tissue Culture Facility for providing reagents and C. Semprich for help with the CUT&RUN protocol. We thank B. V. Puyvelde