Improving prediction of core transcription factors for cell reprogramming and transdifferentiation

Identification of transcription factors (TFs) that could induce and direct cell conversion remains a challenge. Though several hundreds of TFs are usually transcribed in each cell type, the identity of a cell is controlled and can be achieved through the ectopic overexpression of only a small subset of so-called core TFs. Currently, the experimental identification of the core TFs for a broad spectrum of cell types remains challenging. Computational solutions to this problem would provide a better understanding of the mechanisms controlling cell identity during natural embryonic or malignant development, as well as give a foundation for cell-based therapy. Herein, we propose a computational approach based on over-enrichment of transcription factors binding sites (TFBS) in differentially accessible chromatin regions that could identify the potential core TFs for a variety of primary human cells involved in hematopoiesis. Our approach enables the integration of both transcriptomic (single-cell RNA sequencing, scRNA-seq) and epigenenomic (single-cell assay for transposable-accessible chromatin, scATAC-seq) data at the single-cell resolution to search for core TFs, and can be scalable to predict subsets of core TFs and their role in a given conversion between cells.

Single-Cell RNA Sequencing and Assay for Transposase-Accessible Chromatin Using Sequencing Reveals Cellular and Molecular Dynamics of Aortic Aging in Mice

Objective: The impact of vascular aging on cardiovascular diseases has been extensively studied; however, little is known regarding the cellular and molecular mechanisms underlying age-related vascular aging in aortic cellular subpopulations. Approach and Results: Transcriptomes and transposase-accessible chromatin profiles from the aortas of 4-, 26-, and 86-week-old C57/BL6J mice were analyzed using single-cell RNA sequencing and assay for transposase-accessible chromatin sequencing. By integrating the heterogeneous transcriptome and chromatin accessibility data, we identified cell-specific TF (transcription factor) regulatory networks and open chromatin states. We also determined that aortic aging affects cell interactions, inflammation, cell type composition, dysregulation of transcriptional control, and chromatin accessibility. Endothelial cells 1 have higher gene set activity related to cellular senescence and aging than do endothelial cells 2. Moreover, construction of senescence trajectories shows that endothelial cell 1 and fibroblast senescence is associated with distinct TF open chromatin states and an mRNA expression model. Conclusions: Our data provide a system-wide model for transcriptional and epigenetic regulation during aortic aging at single-cell resolution.

Chemical, molecular, and single cell analysis reveal chondroitin sulfate proteoglycan aberrancy in fibrolamellar carcinoma

Fibrolamellar carcinoma (FLC) is an aggressive liver cancer with no effective therapeutic options. The extracellular environment of FLC tumors is poorly characterized and may contribute to cancer growth and/or metastasis. To bridge this knowledge gap, we assessed pathways relevant to proteoglycans, a major component of the extracellular matrix. We first analyzed gene expression data from FLC and non-malignant liver tissue to identify changes in glycosaminoglycan (GAG) biosynthesis pathways. We then implemented a novel LC-MS/MS based method to quantify the abundance of different types of GAGs in patient tumors, followed by measurement of the levels of different GAG-associated proteins. Finally, we performed the first single-cell assay for transposon-accessible chromatin-sequencing on FLC tumors, to identify which cell types are linked to the most dominant GAG-associated protein in FLC. Our results reveal a pathologic aberrancy in chondroitin (but not heparan) sulfate proteoglycans in FLC and highlight a potential role for activated stellate cells.

ATAC-STARR-seq v1

Transcriptional enhancers control cell-type specific gene expression in humans and dysfunction can lead to debilitating diseases, including cancer. Identifying bona-fide enhancers is difficult due to a lack of spatial or sequence constraints. In addition, only a small percentage of the genome is accessible in matured cell types; and therefore, most enhancers are inactive due to their chromatin context rather than intrinsic properties of the DNA sequence itself. For this reason, we decided to assay regulatory activity exclusively within accessible chromatin. To do this, we combined assay for transposase-accessible chromatin using sequencing (ATAC-seq) with self-transcribing active regulatory region sequencing (STARR-seq); we call this method ATAC-STARR-seq. With ATAC-STARR-seq, we identify both active and silent regulatory elements in GM12878 B cells; these active and silent elements are enriched for transcription factor motifs and histone modifications associated with activating and repressing regulation, respectively. We also show that ATAC-STARR-seq quantifies chromatin accessibility and transcription factor binding. We integrate this information and subset active regions based on transcription factor binding profiles. Depending on the transcription factors bound, subsets are enriched for distinct reactome pathways. Altogether, this highlights the power of ATAC-STARR-seq to investigate the transcriptional regulatory landscape of the human genome.

ATAC-STARR-seq v2

Massively parallel reporter assays test the capacity of putative cis-regulatory elements (CREs) to drive transcription on a genome-wide scale. In nearly all cases, chromatin accessibility is necessary to drive activity, so most CREs are inactive due to chromatin context rather than intrinsic DNA sequence properties. Here, we combined assay for transposase-accessible chromatin (ATAC-seq) with self-transcribing active regulatory region sequencing (STARR-seq) to selectively assay the regulatory potential of nucleosome-free DNA genome-wide. Our approach enabled high-resolution testing of ~50 million unique DNA fragments tiling ~101,000 accessible chromatin regions in human lymphoblastoid cells. To illustrate the application of our approach, we show that 30% of all accessible regions contain an activator, a silencer or both. Benchmarking against standard ATAC-seq, our approach faithfully captures chromatin accessibility and transcription factor (TF) footprints with high signal-to-noise. Integrating three layers of genomic information (accessibility, TF occupancy, and activity) provided by ATAC-STARR-seq, we stratified active and silent CREs by the presence of several TF footprints and show that CREs with specific TF combinations are associated with distinct gene regulatory pathways. Altogether, these data highlight the power of ATAC-STARR-seq to comprehensively investigate the regulatory landscape of the human genome from a single DNA source.

One-pot universal NicE-seq: all enzymatic downstream processing of 4% formaldehyde crosslinked cells for chromatin accessibility genomics

Background Accessible chromatin landscape allows binding of transcription factors, and remodeling of promoter and enhancer elements during development. Chromatin accessibility along with integrated multiomics approaches have been used for determining molecular subtypes of cancer in patient samples. Results One-pot Universal NicE-seq (One-pot UniNicE-seq) is an improved accessible chromatin profiling method that negate DNA purification and incorporate sonication free enzymatic fragmentation before library preparation and is suited to a variety of mammalian cells. One-pot UniNicE-seq is versatile, capable of profiling 4% formaldehyde fixed chromatin in as low as 25 fixed cells. Accessible chromatin profile is more efficient on formaldehyde-fixed cells using one-pot UniNicE-seq compared to Tn5 transposon mediated methods, demonstrating its versatility. Conclusion One-pot UniNicE-seq allows the entire process of accessible chromatin labeling and enrichment in one pot at 4% formaldehyde cross-linking conditions. It doesn’t require enzyme titration, compared to other technologies, since accessible chromatin is labelled with 5mC incorporation and deter degradation by nicking enzyme, thus opening the possibility for automation.

Epigenetic loss of heterogeneity from low to high grade localized prostate tumours

Identifying precise molecular subtypes attributable to specific stages of localized prostate cancer has proven difficult due to high levels of heterogeneity. Bulk assays represent a population-average, which mask the heterogeneity that exists at the single-cell level. In this work, we sequence the accessible chromatin regions of 14,424 single-cells from 18 flash-frozen prostate tumours. We observe shared chromatin features among low-grade prostate cancer cells are lost in high-grade tumours. Despite this loss, high-grade tumours exhibit an enrichment for FOXA1, HOXB13 and CDX2 transcription factor binding sites, indicating a shared trans-regulatory programme. We identify two unique genes encoding neuronal adhesion molecules that are highly accessible in high-grade prostate tumours. We show NRXN1 and NLGN1 expression in epithelial, endothelial, immune and neuronal cells in prostate cancer using cyclic immunofluorescence. Our results provide a deeper understanding of the active gene regulatory networks in primary prostate tumours, critical for molecular stratification of the disease.

Genome-wide cancer-specific chromatin accessibility patterns derived from archival processed xenograft tumors

Chromatin accessibility states that influence gene expression and other nuclear processes can be altered in disease. The constellation of transcription factors and chromatin regulatory complexes in cells results in characteristic patterns of chromatin accessibility. The study of these patterns in tissues has been limited because existing chromatin accessibility assays are ineffective for archival formalin-fixed, paraffin-embedded (FFPE) tissues. We have developed a method to efficiently extract intact chromatin from archival tissue via enhanced cavitation with a nanodroplet reagent consisting of a lipid shell with a liquid perfluorocarbon core. Inclusion of nanodroplets during the extraction of chromatin from FFPE tissues enhances the recovery of intact accessible and nucleosome-bound chromatin. We show that the addition of nanodroplets to the chromatin accessibility assay formaldehyde-assisted isolation of regulatory elements (FAIRE), does not affect the accessible chromatin signal. Applying the technique to FFPE human tumor xenografts, we identified tumor-relevant regions of accessible chromatin shared with those identified in primary tumors. Further, we deconvoluted non-tumor signal to identify cellular components of the tumor microenvironment. Incorporation of this method of enhanced cavitation into FAIRE offers the potential for extending chromatin accessibility to clinical diagnosis and personalized medicine, while also enabling the exploration of gene regulatory mechanisms in archival samples.

Cis-regulatory architecture of human ESC-derived hypothalamic neuron differentiation aids in variant-to-gene mapping of relevant complex traits

The hypothalamus regulates metabolic homeostasis by influencing behavior and endocrine systems. Given its role governing key traits, such as body weight and reproductive timing, understanding the genetic regulation of hypothalamic development and function could yield insights into disease pathogenesis. However, given its inaccessibility, studying human hypothalamic gene regulation has proven challenging. To address this gap, we generate a high-resolution chromatin architecture atlas of an established embryonic stem cell derived hypothalamic-like neuron model across three stages of in vitro differentiation. We profile accessible chromatin and identify physical contacts between gene promoters and putative cis-regulatory elements to characterize global regulatory landscape changes during hypothalamic differentiation. Next, we integrate these data with GWAS loci for various complex traits, identifying multiple candidate effector genes. Our results reveal common target genes for these traits, potentially affecting core developmental pathways. Our atlas will enable future efforts to determine hypothalamic mechanisms influencing disease susceptibility.

Modeling chromatin state from sequence across angiosperms using recurrent convolutional neural networks

Accessible chromatin regions are critical components of gene regulation but modeling them directly from sequence remains challenging, especially within plants, whose mechanisms of chromatin remodeling are less understood than in animals. We trained an existing deep learning architecture, DanQ, on leaf ATAC-seq data from 12 angiosperm species to predict the chromatin accessibility of sequence windows within and across species. We also trained DanQ on DNA methylation data from 10 angiosperms, because unmethylated regions have been shown to overlap significantly with accessible chromatin regions in some plants. The across-species models have comparable or even superior performance to a model trained within species, suggesting strong conservation of chromatin mechanisms across angiosperms. Testing a maize held out model on a multi-tissue scATAC panel revealed our models are best at predicting constitutively-accessible chromatin regions, with diminishing performance as cell-type specificity increases. Using a combination of interpretation methods, we ranked JASPAR motifs by their importance to each model and saw that the TCP and AP2/ERF transcription factor families consistently ranked highly. We embedded the top three JASPAR motifs for each model at all possible positions on both strands in our sequence window and observed position- and strand-specific patterns in their importance to the model. With our cross-species "a2z" model it is now feasible to predict the chromatin accessibility and methylation landscape of any angiosperm genome.