Functional Genomics in Embryonic Stem Cells
Embryonic stem cells (ES cells) and induced pluripotent stem cells (iPS cells) are likely to transform personalized medicine because they can be infinitely expanded and differentiated in several tissue types. Hence they permit in vitro modeling of disease and possibly even tissue replacement in patients. Our group tries to clarify the genetic cascades governing cell state and lineage decisions of early development by systematic genetic approaches employing genome-wide screens, facilitated by the generation of haploid ES cells.
Embryonic stem cells represent the fascinating in vitro capture of a transient developmental state of pluripotency found in blastocysts prior to implantation in the uterine wall (Figure A). Such immortal cells can be infinitely expanded in cell culture and differentiated in virtually all cell types of our body. While we understand much about the governing transcription factors defining the state, we know much less about the genetic factors controlling entrance to and exit from pluripotency.
Using genome-wide genetics, our team systematically investigates the genetic framework governing early embryonic development, including differentiation, dedifferentiation, reprogramming, lineage decisions, and epigenetic modifications. Some of the questions we address are the specific genes required for ES cell maintenance and those needed for differentiation. What genes are involved in lineage decisions and control of the epigenetic environment?
Even today, random mutagenesis is by far the most efficient way of generating a multitude of mutations to be used in pooled screening approaches. However, such mutations are masked in diploid cells due to the presence of a second allele (Figure C). In order to achieve systematic saturating screens in ES cells based on random mutagenesis, we generated haploid murine embryonic stem cells via parthenogenesis and developed genetic tools for forward and reverse genomic approaches. Such cells display all features of pluripotent ES cells and contain a precisely haploid chromosome set (Figure B). This setup combines the power of “yeast genetics” with the pluripotency of embryonic stem cells. We complement our haploid genetic platform with small RNA-based approaches such as RNAi and CRISPR for tool generation, validation, and targeted mutagenesis to subsets of genes. Our goal is to identify genetic triggers for more efficient lineage transition and thus create a better experimental regimen of ES cell dedifferentiation and differentiation.
ES cells may be derived directly from blastocysts or generated by dedifferentiation of differentiated cell types; these are known as induced pluripotent stem cells (iPS cells). Dedifferentiation is achieved by nuclear transfer to oocytes or via the expression of 4 transcription factors. While overexpression of these lineage-defining activators may be sufficient to reawaken pluripotency, the process is highly inefficient and rare. We aimed to understand the nature of epigenetic roadblocks inhibiting the generation of induced pluripotent stem cells arising from the expression of these 4 transcription factors, namely Oct4, Sox2, Klf4, and Myc. By employing a very rigorous screen using optimized shRNA technology, we were able to identify new epigenetic barriers to the change of lineage identity, namely the CAF-1 complex depositing core histones to nascent DNA, as well as the sumoylation pathway (Figure D). The project was performed in close collaboration with the laboratories of Johannes Zuber (IMP) and Konrad Hochedlinger (Harvard).
For technical reasons, most genome-wide screens are based on a positive selection of hits. Dropout screens are a standing challenge for functional genomics because of the high demands in terms of signal-to-noise ratios and the quantity of data required for analysis. However, they provide a more direct genetic understanding of pathways. Haploid genomics combined with optimized mutagenesis tools such as CRISPR are likely to produce genome-wide depletion screens. We intend to raise and answer questions beyond what is currently possible by traditional means, such as what mutations are lethal in synergy with an oncogenic lesion? What mutations hypersensitize to compounds with unknown specificity, i.e. destabilize the drug target pathway? What genes are required for specific cell states and cellular responses? We use improved massive parallel sequencing protocols as well as genetic and bioinformatics tools to address these questions.
Our lab focuses on developing new tools with wide applicability. This fosters academic interaction and collaboration within and beyond the campus. In close collaboration with several industrial partners, we also use genomics for drug target prediction. Our team interacts closely with Haplobank, an archive of mutated and sequenced conditional ES cell lines we have jointly set up and operated over the last few years. Of the 100,000 available cell lines, testing those that harbor specific conditional mutations allows for rapid validation of hits identified in genetic screens as well as candidate approaches.