(B-47) Analyzing mouse embryonic stem cell H3.2 single-cell combinatorial indexing sequencing dataset

Introduction:: Histones are basic proteins that have a structural role in cells that include controlling DNA replication and the accessibility of chromatin. H3.1, H3.2, H3.3 though similar in amino structure all have different roles in the cell. Unlike some of the other H3 histones, H3.2 is replication-dependent. This means we can attempt to use H3.2 to examine the S sub-phases and replication timing of the cell. We used in-house H3.2 scTIP-Seq data from mouse embryonic stem cells (mESC) to further explore the incorporation of histone H3.2 at single-cell level. We implemented pseudotime analysis and standard single-cell methodologies to further explore H3.2 incorporation. Additionally, we attempted to understand more about the sub-phase patterns of H3.2.

Materials and Methods:: The  mouse embryonic stem cell H3.2 sciTIP-seq dataset was collected by using single-cell combinatorial indexing (sciTIP-Seq). This technique is a new technology for mapping chromatin binding proteins at the single-cell level. This method was used to characterize the incorporation of histone H3.2 in single mouse embryonic stem cells. The data being at a single cell level can allow for a way to research complex groups of heterogeneous cells. BED files were binned into 2.5-kb genomic windows using BEDTools coverage. The BED files by using bedgraphToBigWig, were then converted to BigWig files. This allowed us to view the genomic ranges in Integrative Genomics Viewer (IGV). Datasets from Gene Expression Omnibus (GEO) were processed using Cell Ranger and Seurat(Bartlett, Daniel A., et al). By using the Seurat package we have applied standard single-cell workflow. This categorized the data into 6 different Seurat clusters. Additionally we did a  pseudotime analysis on the data using Slingshot(Street, Kelly, et al) .

Results, Conclusions, and Discussions:: The trajectory inference, specifically the pseudotime analysis, was successfully performed by using Slingshot(Figure 1). This allowed us to classify cells into different replication groups(early, mid, and late) based on H3.2 incorporation. We found the percentage of the read count per phase(of the replication timing, early, mid and late) for each cell id.  From that we made a 3D scatter plot, Early Percentage vs Mid Percentage vs Late Percentage scaled by pseudotime score(Figure 2) that showed the trajectory inference. Trajectory inference recapitulated the replication timing coupled H3.2 incorporation, however after performing read count analysis, such as UMAPs of H3.2 read counts, group by early, mid and late, showing at least two fold change from the other phases . The read count hasn't supported the trajectory inference as in the UMAP there is not a clear connection between the Seurat clusters and read count (Figure 3). We have successfully explored the mESC H3.2 sciTIP-seq dataset and understand more about the incorporation of H3.2 in mESC at a single cell level.  

Acknowledgements (Optional): :

References (Optional): :

Street, Kelly, et al. “Slingshot: Cell Lineage and Pseudotime Inference for Single-Cell Transcriptomics - BMC Genomics.” BioMed Central, 19 June 2018, bmcgenomics.biomedcentral.com/articles/10.1186/s12864-018-4772-0. 

Bartlett, Daniel A., et al. “High-Throughput Single-Cell Epigenomic Profiling by Targeted Insertion of Promoters (Tip-Seq).” Journal of Cell Biology, 6 Dec. 2021, rupress.org/jcb/article/220/12/e202103078/212821/High-throughput-single-cell-epigenomic-profiling.