Tracking chromatin structure changes by single-cell multi-epigenomics with RNA polymerase II binding profiles

doi:10.21203/rs.3.rs-4503255/v1

Download PDF

Article

Tracking chromatin structure changes by single-cell multi-epigenomics with RNA polymerase II binding profiles

https://doi.org/10.21203/rs.3.rs-4503255/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Transcription factor-bound chromatin structures regulate cell lineages in multicellular organisms. Single-cell epigenomics has the potential to reveal lineage determination on chromatin structure, but the methodology is still in development. Here, we develop single-cell combinatorial-indexing multi-target Chromatin Integration Labeling followed by sequencing (sci-mtChIL-seq) as a single-cell multi-epigenomics approach, which enables simultaneous single-cell analysis of both RNA polymerase II binding to chromatin and epigenomic factors such as transcription factors/histones. We apply sci-mtChIL-seq to analyze the binding dynamics of skeletal-muscle-specific transcription factor MyoD during mouse embryonic myogenesis. Based on RNA polymerase II-bound gene profiles, single-cells are efficiently classified into myogenic-clusters and ordered pseudotemporally. MyoD exhibits genome-wide binding in the muscle-progenitor-cell population, but in myocytes, this transitions toward enrichment in muscle-specific genes on active chromatin. Thus, sci-mtChIL-seq can be a powerful tool to analyze epigenomic dynamics in cell fate determination.

Biological sciences/Biological techniques/Epigenetics analysis/Chromatin analysis

Biological sciences/Developmental biology/Embryogenesis/Cell lineage

Biological sciences/Biotechnology/Sequencing/Next-generation sequencing

Biological sciences/Molecular biology/Chromatin/Chromatin structure

Yes there is potential Competing Interest. The authors declare no competing financial interests except A.H., H.Ku., Yu.S., H.Ki. and Y.O., who are involved in a patent related to ChIL.

Supplementaryinformation.docx
Supplementarytables.xlsx
ExtendedDataFig1240530.pdf
Extended Data Fig. 1 Overview of sci-ChIL-seq. a, sci-ChIL-seq scheme. A total of 10-30k biotinylated cells per well were combined in 96-well avidin-coated plates and stained with primary antibody (1st Ab) and sci-ChIL probe (see Fig. 1b). Tn5 transposition via Tn5 transposase integrates the sci-ChIL probe’s indexed DNA into the genome around the primary antibody of the cell (cells are indexed with a unique sci-ChIL probe per well). The cells in the 96-well plate are then pooled, and 25 cells per well are added to another 96-well plate with a cell sorter, so that cells with different primary indexes are placed in the wells. The integrated regions are then amplified as RNA by in vitrotranscription using T7 RNA polymerase. Transcribed RNA is purified and converted to cDNA for sequencing library preparation; a second index is added by PCR prior to sequencing. PCR is performed with a set of primers with unique i5 and i7 sequences per well. A library with a different primary index and the same secondary index was generated for each well. Overall, a library of approximately 2400 cells with different primary and secondary indices is produced. b, Example of sci-ChIL probe index usage (extensibility). One single-target sequence includes one set of 96 sci-ChIL probes with T5 barcodes, and two sets of multi-target sequences is used.
ExtendedDataFig22405292.pdf
Extended Data Fig. 2 Quality control of sci-mtChIL-seq. a, Scatter plots showing the number of reads obtained for the two epigenomic factors relative to the index used in sci-mtChIL-seq. The horizontal axis shows read counts for RNAPII, and the vertical axis shows the read counts for H3K4me3 (top) or H3K27me3 (bottom). Scatter plots are shown by embryo day. b, Genomic feature distributions. Each graph shows the RNAPII (top left) and H3K4me3 reads (top right) of sci-mtChIL-seq RNAPII-H3K4me3 or the RNAPII reads (bottom left) and H3K27me3 reads (bottom right) of sci-mtChIL-seq RNAPII-H3K27me3 dataset for each embryo day, and shows the percentage of reads in the promoter, CpG-island, intron, exon, enhancer, and intergenic regions. c, Violin plot showing the number of genes detected in RNAPII-H3K4me3 or RNAPII-H3K27me3. Plots of RNAPII (top left) and H3K4me3 (top right) in RNAPII-H3K4me3 and RNAPII (bottom left) and H3K27me3 (bottom right) in RNAPII-H3K27me3 are shown for each embryo stage. The medians, 25th/75th percentiles, and 1.5 interquartile range (IQR) were used to construct the box plots. Source data are provided as a Source Data file.
ExtendedDataFig3240529.pdf
Extended Data Fig. 3 Classification of cell types by RNAPII is independent of paired factors. a, UMAP visualization of 3 sci-mtChIL-seq datasets using RNAPII signals. The right panel (pair = MyoD) corresponds to Fig 2d. The color represents each cell type. Each cell type was determined in the same way as in Figs. 2d–2e. b, IGV track view showing the distribution of reads. For each dataset, reads were aggregated by cell type, showing the region around the marker gene for each cell type. The colors are the same as in a. c, Correlation of RNAPII binding to genes. For each dataset, the mean RNAPII binding per gene per cluster was calculated, and Pearson’s correlation coefficients were calculated. The sci-mtChIL-seq RNAPII and cell types of H3K4me3, H3K27me3, and MyoD are shown in the indicated colors. d, Plot showing the proportion of cell types per stage of the sampled embryos. The colors correspond to the explicitly indicated cell type.
ExtendedDataFig4240529.pdf
Extended Data Fig. 4 Efficiency of muscle trans-differentiation with the Tet-On 3G system. a, Representative cell immunostaining fluorescence microscopy images showing MyoD (left) and Myog (right) expression in each cell stage. Each stain is shown in red; nuclear staining with Hoechst is shown in blue. The scale bar indicates 50 mm. Bars show the percentage of MyoD- and Myog-positive cells relative to the number of cells indicated by each stain. The error bars indicate mean ± standard deviation of the nine fields. b, Violin plot showing the number of genes detected by sci-mtChIL-seq in NIH3T3 cells. The medians, 25th/75th percentiles, and 1.5 interquartile range (IQR) were used to construct the box plots. Source data are provided as a Source Data file.
ExtendedDataFig5240529.pdf
Extended Data Fig. 5 Pseudotime analysis of NIH3T3 muscle trans-differentiation using RNAPII data from each sci-mtChIL-seq dataset. a, Various sci-mtChIL-seq data visualized by UMAP and PHATE. The colors indicate the samples. Points connected by lines indicate identical cells. b, Pseudotime trajectory estimated using PHATE. The pairs of antibodies are clearly indicated in the diagram. c, Density plot showing the distribution of the PHATE coordinates in D0h (top), D24h (middle), and D72h (bottom) cells. The data are in the same order as in b.
ExtendedDataFig6240529.pdf
Extended Data Fig. 6 Consistency of estimated pseudotime of each RNAPII signal. a, Dot plot showing GO enrichment analysis of differentially expressed genes (DEGs) using sci-mtChIL-seq RNAPII counts over time. For each gene group, the two biological processes with the highest gene ratios were extracted and shown. The size of the dots represents the gene ratio, and the color represents the adjusted p-value. The numbers below represent the numbers of genes analyzed. b. Examples of DEGs identified using sci-mtChIL-seq RNAPII counts over time (left; genes with increased RNAPII binding; right, genes with decreased RNAPII binding). The color indicates the pair of antibodies.

Download PDF

Version 1

posted

You are reading this latest preprint version

Tracking chromatin structure changes by single-cell multi-epigenomics with RNA polymerase II binding profiles

Status:

Version 1

Abstract

Figures

Full Text

Additional Declarations

Supplementary Files

Status:

Version 1