Telomemore enables single-cell analysis of cell cycle and chromatin condensation

Abstract

It is now easy to perform multiome single-cell analysis, including both RNA and ATAC readouts from the same cell. This enables a closer linkage between the two types of modalities, but it remains an open question what more information can be extracted from this type of data. ATAC-seq is normally only used to assay transcription factor binding to open regions. By reanalyzing several large datasets, and generating an atlas of B cells, we show that telomere accessibility can better pinpoint processes related to cell cycle and chromatin condensation. We provide Telomemore, a tool that can extract telomeric reads, and give examples of new findings it enables. Our new findings will aid in the annotation and analysis of single-cell ATAC or multiome datasets.

The CD4 T cell epigenetic JUNB+ state is associated with proliferation and exhaustion

“…It appears that the chromatin is particularly and continuously condensed in both JUNB/Tfh and JUNB 17 , (Figure 2i), and it is possibly due to a higher order change in TAD structure, with increased insulation, driven by CTCF and other factors. Telomemore also predicts high condensation of naive T conv and naive Tregs, in line with previous time course ATAC-seq data 5,26 . The large difference in chromatin accessibility in JUNB+ vs other clusters opens questions about the nature of this state.…”

“…Cells from respective donors were computationally separated using the available SNP information 47 . Telomemore was invoked to count telomeric reads in the ATAC-seq data, indicating primarily chromatin condensation level 26 .…”

“…The overlap was however poor. Our previous attempts at multiome B cell data 26 label transfer has also failed, suggesting that data using the multiome chemistry is too different from plain scRNAseq for this approach to be viable.…”

“…We generally had poor experience of several single cell batch integration methods. We have had similar issues comparing multiome B cell data 26 , and suspect it may partially be due to differences in cDNA generation between the 10x Genomics multiome RNA-seq vs regular single-cell RNA-seq kits. We thus instead only compared the average gene expression for each cluster across all atlases.…”

“…Interestingly, CTCF is one of the top regulators of FOXP3 expression (Figure 2h), which is down in JUNB+ (Figure 2a). To get an idea of the overall chromatin state, we analyzed the ATAC-seq data using Telomemore, based on the principle that high ATAC telomere abundance frequently indicates low general accessibility 26 . It appears that the chromatin is particularly and continuously condensed in both JUNB/Tfh and JUNB 17 , (Figure 2i), and it is possibly due to a higher order change in TAD structure, with increased insulation, driven by CTCF and other factors.…”