Skip to main content

Decoding the Epigenomic Regulatory Code by the Use of Single Cell Technologies

Periodic Reporting for period 2 - SC-EpiCode (Decoding the Epigenomic Regulatory Code by the Use of Single Cell Technologies)

Reporting period: 2019-04-01 to 2020-09-30

Chromatin Regulators (CRs) are key players in mediating chromatin states. CRs can read, write, and remove chemical modifications on histone tails and hence modify chromatin accessibility to transcription factors. This function allows rigorous maintenance of differentiation states and is central in regulating the chromatin changes that occur during Embryonic Stem Cell (ESC) differentiation. CRs are known to act in concert, with modules interacting primarily with specific chromatin regions and functional elements. Although some effects of specific CR mutations on ESC proliferation and differentiation potential are known, the significance of the interactions between CRs of different modules during differentiation remains elusive. Cellular variability has been observed at the earlier stages of embryonic development, where subtle differences within the populations of pluripotent cells may eventually lead to different developmental outcomes. Some exciting questions are: 1. What triggers differentiation at the epigenomic level? 2. How do chromatin and mRNA expression interact to co-regulate cellular identity? 3. What is the impact of cellular heterogeneity on differentiation? 4. What is the functional role of CRs and TFs, and how do their interactions affect decisions concerning cell fate? 5. Can we uncover the universal organization principles guiding such regulatory networks? The single-cell ChIP-seq experimental approach and corresponding analysis tools are an essential pre-requisite in addressing these questions, and our preliminary results (Rotem et al., 2015b, 2015a) provide a proof-of-principle for the successful implementation of these techniques. To this end, we set out in the first part of this project to develop an integrative microfluidic-based single-cell system (Aim 1). It includes improvements in modifying microfluidic devices, developing an advanced barcoding scheme, and other optimizations to increase the sensitivity and throughput capabilities of the protocol. In the second part of the project, we develop a robust method for single and multiple perturbations (Aim 2) that enables dissecting out the multifactorial effect of CRs during ESCs differentiation using a CRISPR/CAS9- based multi-perturbation screen. In the third part, we would apply this technology to study the early differentiation of mouse and human ES cells (Aim 3).
"We have advanced significantly in essentially each aim independently. For Aim 1, we have successfully established a robust single-cell RNA-seq using drop based microfluidics in our lab and applied it to a variety of different cell types (ex. mouse ES cells, human lung adenocarcinoma cells, etc.). We also designed and produced polyacrylamide beads carrying double-strand primers which will allow capturing chromatin from a single cell. Barcoding ligation efficiency was improved by changing from the current PCR-based amplification approach, which requires two-side ligation, to a T7 in-vitro transcription-based-system which needs only one side of the nucleosomal DNA to be ligated.
For Aim 2 and 3, we have generated mouse ES cell lines perturbed in multiple chromatin regulator genes rendering the cells viable and able to differentiate. This was done by cloning of a lentiviral plasmid bearing tet-on Cas9 fused to a GFP reporter (TRE-Cas9-GFP-tetON), it yielded a cell line expressing Cas9 upon Doxycycline activation. By analyzing the resulting genetic background of the perturbed cell lines, we verified the effectiveness of the perturbations, while the robustness of perturbations needed to be refined.

Finally, we just published a paper ""Transcription Factor Binding in Embryonic Stem Cells Is Constrained by DNA Sequence Repeat Symmetry"" in Biophysical Journal for which we reveal that DNA sequence repeat symmetry plays a central role in defining TF-DNA-binding preferences. Using an in vivo reporter assay, we show that gene expression in embryonic stem cells can be positively modulated by the presence of genomic and computationally designed DNA oligonucleotides containing identified nonconsensus-repetitive sequence elements. Our findings show that despite the enormous sequence complexity of the TF-DNA binding landscape in differentiating embryonic stem cells, this landscape can be quantitatively characterized in simple terms using the notion of DNA sequence repeat symmetry."
During the development of the project, we also had some progress beyond the state of the art. We chose to use single-cell rather than bulk population to profile transcriptome and map epigenomic states because bulk population measurements average signals over the entire population, hiding transcriptionally and epigenetically diverse sub-populations, while single-cell assays have revealed the scope and importance of heterogeneity in many biological systems. However, low sensitivity makes it difficult to delineate biological variation from noise, and as a result, the capacity of such assays to identify cellular state differences that arise from genes expressed at low-to-mid abundance is highly limited. To overcome this limitation, we developed CloneSeq, a 3D clone-based RNA-seq approach. Our hypothesis was that clones are composed of cells more similar to each other than cells picked at random, and that analysis of clonal cells would have improved sensitivity and broader coverage relative to single-cell RNA-seq. Our results support this hypothesis, as cells originating from a given clone had more similar transcriptional profiles than cells across clones. The small clones in our 3D system also had detectably different phenotypes. We leveraged this observation to perform an in depth dissection of cellular heterogeneity in lung adenocarcinoma PC9 cells. We were able to characterize different cellular states including cancer stem-like cells (CSCs), high and low replicative cancer cellular states, and different levels of invasiveness. Such features cannot be detected using scRNA-seq due to its low mapping resolution. We also showed that this 3D system induces embryonic stem cell formation without standard supplements, such as LIF and MEK and GSK3 inhibitors (2i), and improves efficiency of induced pluripotent stem cell (IPSC) production, making it superior than standard ESC culturing methods. This work is currently under revision in Nature Biotechnology.
Meanwhile, we constated that the available technologies for single-cell sequencing nowadays are either microfluidics-based that allow routine profiling of thousands of single cells in an experiment but sequence only the 3’ end and cannot recover splicing patterns or sequence variants, or well plate-based that allow a better sensitivity per cell and enable sequencing of the full transcripts with an individual amplification strategy but suffer from high-cost, labour-intensity and small-scale. Therefore, we are developing a novel 3D clone-based full-length RNA-Seq profiling technology, which will be the first time a full-length RNA-seq is done in drop-based microfluidics. We modify the conventional DGE (Differential Gene Expression) 3’-sequencing into full-length RNA sequencing in droplets by using Not So Random (NSR) primers, specific 408 hexamers targeted to human mRNA but depleted from rRNA and tRNA will be constructed within our barcoded hydrogels using a dedicated robotic liquid handler. This work is currently under revision for ERC Proof of Concept Pilot Lump Sum Grant 2020.
Biophysical journal cover