Chromatin accessibility profiling methods

Abstract

Chromatin accessibility, or the physical access to chromatinized DNA, is a widely studied characteristic of the eukaryotic genome. As active regulatory DNA elements are generally 'accessible', the genome-wide profiling of chromatin accessibility can be used to identify candidate regulatory genomic regions in a tissue or cell type. Multiple biochemical methods have been developed to profile chromatin accessibility, both in bulk and at the single-cell level. Depending on the method, enzymatic cleavage, transposition or DNA methyltransferases are used, followed by high-throughput sequencing, providing a view of genome-wide chromatin accessibility. In this Primer, we discuss these biochemical methods, as well as bioinformatics tools for analysing and interpreting the generated data, and insights into the key regulators underlying developmental, evolutionary and disease processes. We outline standards for data quality, reproducibility and deposition used by the genomics community. Although chromatin accessibility profiling is invaluable to study gene regulation, alone it provides only a partial view of this complex process. Orthogonal assays facilitate the interpretation of accessible regions with respect to enhancer-promoter proximity, functional transcription factor binding and regulatory function. We envision that technological improvements including single-molecule, multi-omics and spatial methods will bring further insight into the secrets of genome regulation.

Fig. 1 |. Chromatin accessibility profiling in bulk and at single-cell level reveals putative regulatory regions.

a | Representation of a chromatin landscape is shown in which transcription factor (TF)-bound enhancers and the promoter of a gene are nucleosome depleted and thus accessible. The TFs are represented as coloured circles and the arrows represent 3D looping of the enhancers to the promoter of the target gene. b | Bulk and single-cell chromatin accessibility profiles of a heterogeneous sample containing three different cell populations. When performing single-cell chromatin accessibility profiling, sparse single-cell data are used to cluster cells, often followed by aggregating the reads per cluster, thereby reconstituting pseudo-bulk profiles per cluster or cell type. H3K27ac, histone H3 acetylated at lysine 27; Pol II, polymerase II; TSS, transcription start site.

Fig. 2 |. Experimental approaches for measuring chromatin accessibility and nucleosome positioning.

a | In deoxyribonuclease I (DNase I) hypersensitive site sequencing (DNase-seq), the DNase I enzyme (represented as yellow scissors) is used to preferentially cleave accessible chromatin, generating fragments that can then be amplified into sequencing libraries. b | In Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq), a hyperactive version of the Tn5 transposase (represented by the dark grey circle) is used to preferentially insert into accessible chromatin while simultaneously attaching adapters (represented by the red and blue lines on the Tn5 transposase) to the resulting fragments that can be used to directly amplify sequencing libraries. Both DNase-seq and ATAC-seq generate peaks in read coverage over accessible regions in the genome. c | In micrococcal nuclease sequencing (MNase-seq), the MNase enzyme (represented as red scissors) is used to digest DNA that is not protected by bound proteins, leaving intact fragments protected by protein occupancy (primarily nucleosomes). These fragments are then amplified, resulting in increased read coverage over positioned nucleosomes. d | DNA methyltransferase-based approaches rely on the labelling of accessible DNA with DNA methylation modifications (represented by drawing pins), which can either be sequenced using Illumina platforms following bisulfite conversion or via long-read sequencing platforms that directly read the modified bases (unmodified and modified bases are represented as light and dark blue circles, respectively). These single-molecule chromatin accessibility profiling approaches tend to provide a simultaneous read-out of both nucleosome positioning and accessible chromatin regions. Accessible chromatin regions represent themselves as higher peaks due the fact that they have more nucleotides that are accessible to the methyltransferases and are therefore more frequently methylated, compared with the internucleosomal sequences that are thus not methylated in every single-molecule read. In all four panels, bound transcription factors (TFs) are visualized via coloured circles on the accessible chromatin.

Fig. 3 |. Overview of common tasks in the analysis of bulk chromatin accessibility data.

a | Starting from raw sequencing reads. b | Bulk chromatin accessibility data generally undergo several preprocessing steps, including a pre-mapping quality control (QC) and adaptor trimming step. c | Mapping of the trimmed reads to a reference genome for the studied species. d | Mapped reads are filtered. e | An additional post-alignment QC step is recommended through several metrics and data visualizations. f | An important step in chromatin accessibility data analysis is calling peaks, as these usually form the basis of several downstream analyses. g | Differentially accessibility analysis can be performed pairwise (condition A versus B) or across multiple conditions. h–m | Additional downstream analyses include annotation and enrichment analysis for the identified peaks (part h), visual inspection of chromatin accessibility data tracks (part i), motif enrichment within peaks (sets) using predefined databases or de novo (part j), transcription factor (TF) footprinting analysis (part k), mapping of nucleosome positions (part l) and integration with RNA sequencing (RNA-seq) or chromatin immunoprecipitation followed by sequencing (ChIP–seq) data to link different omics layers or to generate gene regulatory networks (part m). TSS, transcription start site.

Fig. 4 |. Overview of common tasks in the analysis of scATAC-seq data.

a | Outline of key steps in processing single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) data sets, six of which are illustrated in panels b–g. b | An important step in the analysis of scATAC-seq data is clustering the cells via dimensionality reduction of the feature by cell matrix (via UMAP, for example) to discover the different cell populations. In the given example, dots represent the single cells, and their colours and numbers represent the nine identified cell clusters or cell populations. c | Identification of marker genes and/or peaks for each of the cell clusters allows further study of the putative cell populations. d | By aggregating the accessibility profiles of all cells within a cluster, pseudo-bulk genome browser tracks can be generated for each cell population. e | Specific tools allow the identification of peak to gene links, which can reveal putative target genes of identified peaks. f | Assessing peak co-accessibility allows grouping peaks into sets of co-regulated regions. g | When analysing scATAC-seq from a time-series or differentiation experiment, trajectory analysis allows study of the dynamic changes in chromatin accessibility along a pseudo-time axis. QC, quality control.

Fig. 5 |. Schematic overview of future roads and opportunities for chromatin accessibility profiling.

In the coming years, our ability to measure chromatin accessibility concurrently with multiple regulatory layers in the same single cell will continue to expand. New insights into regulatory biology will be gained by applying these methods in the native spatial context and in systems undergoing perturbations. Development of computational tools that can dive into the complexity of the emerging data sets will be crucial for the success of these endeavours. Ultimately, these approaches will empower us to functionally dissect the role of regulatory elements and their relationship to gene expression. TF, transcription factor.