Integrated Single-Cell Analysis Maps the Continuous Regulatory Landscape of Human Hematopoietic Differentiation.

Human hematopoiesis involves cellular differentiation of multipotent cells into progressively more lineage-restricted states. While the chromatin accessibility landscape of this process has been explored in defined populations, single-cell regulatory variation has been hidden by ensemble averaging. We collected single-cell chromatin accessibility profiles across 10 populations of immunophenotypically defined human hematopoietic cell types and constructed a chromatin accessibility landscape of human hematopoiesis to characterize differentiation trajectories. We find variation consistent with lineage bias toward different developmental branches in multipotent cell types. We observe heterogeneity within common myeloid progenitors (CMPs) and granulocyte-macrophage progenitors (GMPs) and develop a strategy to partition GMPs along their differentiation trajectory. Furthermore, we integrated single-cell RNA sequencing (scRNA-seq) data to associate transcription factors to chromatin accessibility changes and regulatory elements to target genes through correlations of expression and regulatory element accessibility. Overall, this work provides a framework for integrative exploration of complex regulatory dynamics in a primary human tissue at single-cell resolution.

Clonal inactivation of TERT impairs stem cell competition.

Telomerase is intimately associated with stem cells and cancer, because it catalytically elongates telomeres-nucleoprotein caps that protect chromosome ends1. Overexpression of telomerase reverse transcriptase (TERT) enhances the proliferation of cells in a telomere-independent manner2-8, but so far, loss-of-function studies have provided no evidence that TERT has a direct role in stem cell function. In many tissues, homeostasis is shaped by stem cell competition, a process in which stem cells compete on the basis of inherent fitness. Here we show that conditional deletion of Tert in the spermatogonial stem cell (SSC)-containing population in mice markedly impairs competitive clone formation. Using lineage tracing from the Tert locus, we find that TERT-expressing SSCs yield long-lived clones, but that clonal inactivation of TERT promotes stem cell differentiation and a genome-wide reduction in open chromatin. This role for TERT in competitive clone formation occurs independently of both its reverse transcriptase activity and the canonical telomerase complex. Inactivation of TERT causes reduced activity of the MYC oncogene, and transgenic expression of MYC in the TERT-deleted pool of SSCs efficiently rescues clone formation. Together, these data reveal a catalytic-activity-independent requirement for TERT in enhancing stem cell competition, uncover a genetic connection between TERT and MYC and suggest that a selective advantage for stem cells with high levels of TERT contributes to telomere elongation in the male germline during homeostasis and ageing.

ecDNA hubs drive cooperative intermolecular oncogene expression.

Extrachromosomal DNA (ecDNA) is prevalent in human cancers and mediates high expression of oncogenes through gene amplification and altered gene regulation1. Gene induction typically involves cis-regulatory elements that contact and activate genes on the same chromosome2,3. Here we show that ecDNA hubs-clusters of around 10-100 ecDNAs within the nucleus-enable intermolecular enhancer-gene interactions to promote oncogene overexpression. ecDNAs that encode multiple distinct oncogenes form hubs in diverse cancer cell types and primary tumours. Each ecDNA is more likely to transcribe the oncogene when spatially clustered with additional ecDNAs. ecDNA hubs are tethered by the bromodomain and extraterminal domain (BET) protein BRD4 in a MYC-amplified colorectal cancer cell line. The BET inhibitor JQ1 disperses ecDNA hubs and preferentially inhibits ecDNA-derived-oncogene transcription. The BRD4-bound PVT1 promoter is ectopically fused to MYC and duplicated in ecDNA, receiving promiscuous enhancer input to drive potent expression of MYC. Furthermore, the PVT1 promoter on an exogenous episome suffices to mediate gene activation in trans by ecDNA hubs in a JQ1-sensitive manner. Systematic silencing of ecDNA enhancers by CRISPR interference reveals intermolecular enhancer-gene activation among multiple oncogene loci that are amplified on distinct ecDNAs. Thus, protein-tethered ecDNA hubs enable intermolecular transcriptional regulation and may serve as units of oncogene function and cooperative evolution and as potential targets for cancer therapy.

Circular ecDNA promotes accessible chromatin and high oncogene expression.

Oncogenes are commonly amplified on particles of extrachromosomal DNA (ecDNA) in cancer1,2, but our understanding of the structure of ecDNA and its effect on gene regulation is limited. Here, by integrating ultrastructural imaging, long-range optical mapping and computational analysis of whole-genome sequencing, we demonstrate the structure of circular ecDNA. Pan-cancer analyses reveal that oncogenes encoded on ecDNA are among the most highly expressed genes in the transcriptome of the tumours, linking increased copy number with high transcription levels. Quantitative assessment of the chromatin state reveals that although ecDNA is packaged into chromatin with intact domain structure, it lacks higher-order compaction that is typical of chromosomes and displays significantly enhanced chromatin accessibility. Furthermore, ecDNA is shown to have a significantly greater number of ultra-long-range interactions with active chromatin, which provides insight into how the structure of circular ecDNA affects oncogene function, and connects ecDNA biology with modern cancer genomics and epigenetics.

Polycomb-mediated genome architecture enables long-range spreading of H3K27 methylation.

Polycomb-group proteins play critical roles in gene silencing through the deposition of histone H3 lysine 27 trimethylation (H3K27me3) and chromatin compaction. This process is essential for embryonic stem cell (ESC) pluripotency, differentiation, and development. Polycomb repressive complex 2 (PRC2) can both read and write H3K27me3, enabling progressive spreading of H3K27me3 on the linear genome. Long-range Polycomb-associated DNA contacts have also been described, but their regulation and role in gene silencing remain unclear. Here, we apply H3K27me3 HiChIP, a protein-directed chromosome conformation method, and optical reconstruction of chromatin architecture to profile long-range Polycomb-associated DNA loops that span tens to hundreds of megabases across multiple topological associated domains in mouse ESCs and human induced pluripotent stem cells. We find that H3K27me3 loop anchors are enriched for Polycomb nucleation points and coincide with key developmental genes. Genetic deletion of H3K27me3 loop anchors results in disruption of spatial contact between distant loci and altered H3K27me3 in cis, both locally and megabases away on the same chromosome. In mouse embryos, loop anchor deletion leads to ectopic activation of the partner gene, suggesting that Polycomb-associated loops control gene silencing during development. Further, we find that alterations in PRC2 occupancy resulting from an RNA binding­deficient EZH2 mutant are accompanied by loss of Polycomb-associated DNA looping. Together, these results suggest PRC2 uses RNA binding to enhance long-range chromosome folding and H3K27me3 spreading. Developmental gene loci have unique roles in Polycomb spreading, emerging as important architectural elements of the epigenome.

HiChIRP reveals RNA-associated chromosome conformation.

Modular domains of long non-coding RNAs can serve as scaffolds to bring distant regions of the linear genome into spatial proximity. Here, we present HiChIRP, a method leveraging bio-orthogonal chemistry and optimized chromosome conformation capture conditions, which enables interrogation of chromatin architecture focused around a specific RNA of interest down to approximately ten copies per cell. HiChIRP of three nuclear RNAs reveals insights into promoter interactions (7SK), telomere biology (telomerase RNA component) and inflammatory gene regulation (lincRNA-EPS).

An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues.

We present Omni-ATAC, an improved ATAC-seq protocol for chromatin accessibility profiling that works across multiple applications with substantial improvement of signal-to-background ratio and information content. The Omni-ATAC protocol generates chromatin accessibility profiles from archival frozen tissue samples and 50-µm sections, revealing the activities of disease-associated DNA elements in distinct human brain structures. The Omni-ATAC protocol enables the interrogation of personal regulomes in tissue context and translational studies.

Decoding polygenic diseases: advances in noncoding variant prioritization and validation.

Genome-wide association studies (GWASs) provide a key foundation for elucidating the genetic underpinnings of common polygenic diseases. However, these studies have limitations in their ability to assign causality to particular genetic variants, especially those residing in the noncoding genome. Over the past decade, technological and methodological advances in both analytical and empirical prioritization of noncoding variants have enabled the identification of causative variants by leveraging orthogonal functional evidence at increasing scale. In this review, we present an overview of these approaches and describe how this workflow provides the groundwork necessary to move beyond associations toward genetically informed studies on the molecular and cellular mechanisms of polygenic disease.

CHOIR improves significance-based detection of cell types and states from single-cell data.

Clustering is a critical step in the analysis of single-cell data, as it enables the discovery and characterization of putative cell types and states. However, most popular clustering tools do not subject clustering results to statistical inference testing, leading to risks of overclustering or underclustering data and often resulting in ineffective identification of cell types with widely differing prevalence. To address these challenges, we present CHOIR (clustering hierarchy optimization by iterative random forests), which applies a framework of random forest classifiers and permutation tests across a hierarchical clustering tree to statistically determine which clusters represent distinct populations. We demonstrate the enhanced performance of CHOIR through extensive benchmarking against 14 existing clustering methods across 100 simulated and 4 real single-cell RNA-seq, ATAC-seq, spatial transcriptomic, and multi-omic datasets. CHOIR can be applied to any single-cell data type and provides a flexible, scalable, and robust solution to the important challenge of identifying biologically relevant cell groupings within heterogeneous single-cell data.

Convergent epigenetic evolution drives relapse in acute myeloid leukemia.

Relapse of acute myeloid leukemia (AML) is highly aggressive and often treatment refractory. We analyzed previously published AML relapse cohorts and found that 40% of relapses occur without changes in driver mutations, suggesting that non-genetic mechanisms drive relapse in a large proportion of cases. We therefore characterized epigenetic patterns of AML relapse using 26 matched diagnosis-relapse samples with ATAC-seq. This analysis identified a relapse-specific chromatin accessibility signature for mutationally stable AML, suggesting that AML undergoes epigenetic evolution at relapse independent of mutational changes. Analysis of leukemia stem cell (LSC) chromatin changes at relapse indicated that this leukemic compartment underwent significantly less epigenetic evolution than non-LSCs, while epigenetic changes in non-LSCs reflected overall evolution of the bulk leukemia. Finally, we used single-cell ATAC-seq paired with mitochondrial sequencing (mtscATAC) to map clones from diagnosis into relapse along with their epigenetic features. We found that distinct mitochondrially-defined clones exhibit more similar chromatin accessibility at relapse relative to diagnosis, demonstrating convergent epigenetic evolution in relapsed AML. These results demonstrate that epigenetic evolution is a feature of relapsed AML and that convergent epigenetic evolution can occur following treatment with induction chemotherapy.

Acute myeloid leukemia (or AML for short) is a type of blood cancer characterized by abnormally high production of immature white blood cells. Despite advances in AML treatment, many patients relapse after an initially successful first round of treatment. As a result, understanding the factors contributing to relapse is essential for developing effective treatments for the disease. Like most cancers, AML can evolve because of changes to the DNA sequence in cells that cause them to grow uncontrollably or resist treatment. Alongside these genetic mutations, AML cells also undergo 'epigenetic' changes, where regions of the DNA are modified and genes can be switched on or off without altering the DNA sequence. Previous research has demonstrated that epigenetic changes contribute to the development of AML, however, it was not clear if these changes could also make cells resistant to treatment without acquiring new DNA mutations. Nuno, Azizi et al. addressed this question by analyzing the epigenetic states of AML cells from 26 patients at the time of their diagnosis and after treatment when the disease had relapsed. Analysis revealed that almost half of the patients with AML experienced a relapse without acquiring new DNA mutations. Instead, these AML cells developed specific epigenetic changes that helped them to resist cancer treatment. Moreover, studying individual AML cells from different patients showed that the cells became more epigenetically similar at relapse, suggesting that they converge towards a more treatment-resistant disease. Future experiments will determine exactly how these epigenetic changes lead to treatment resistance. Currently, most of the drugs used to treat AML are either chemotherapies or ones that target specific DNA mutations. The findings of Nuno, Azizi et al. suggest that drugs targeting specific epigenetic changes may be more effective for some patients. Further studies will be needed to determine which patients may benefit and which epigenetic drugs could be useful.

Systematic identification of post-transcriptional regulatory modules.

In our cells, a limited number of RNA binding proteins (RBPs) are responsible for all aspects of RNA metabolism across the entire transcriptome. To accomplish this, RBPs form regulatory units that act on specific target regulons. However, the landscape of RBP combinatorial interactions remains poorly explored. Here, we perform a systematic annotation of RBP combinatorial interactions via multimodal data integration. We build a large-scale map of RBP protein neighborhoods by generating in vivo proximity-dependent biotinylation datasets of 50 human RBPs. In parallel, we use CRISPR interference with single-cell readout to capture transcriptomic changes upon RBP knockdowns. By combining these physical and functional interaction readouts, along with the atlas of RBP mRNA targets from eCLIP assays, we generate an integrated map of functional RBP interactions. We then use this map to match RBPs to their context-specific functions and validate the predicted functions biochemically for four RBPs. This study provides a detailed map of RBP interactions and deconvolves them into distinct regulatory modules with annotated functions and target regulons. This multimodal and integrative framework provides a principled approach for studying post-transcriptional regulatory processes and enriches our understanding of their underlying mechanisms.

Single-cell chromatin accessibility reveals malignant regulatory programs in primary human cancers.

To identify cancer-associated gene regulatory changes, we generated single-cell chromatin accessibility landscapes across eight tumor types as part of The Cancer Genome Atlas. Tumor chromatin accessibility is strongly influenced by copy number alterations that can be used to identify subclones, yet underlying cis-regulatory landscapes retain cancer type-specific features. Using organ-matched healthy tissues, we identified the "nearest healthy" cell types in diverse cancers, demonstrating that the chromatin signature of basal-like-subtype breast cancer is most similar to secretory-type luminal epithelial cells. Neural network models trained to learn regulatory programs in cancer revealed enrichment of model-prioritized somatic noncoding mutations near cancer-associated genes, suggesting that dispersed, nonrecurrent, noncoding mutations in cancer are functional. Overall, these data and interpretable gene regulatory models for cancer and healthy tissue provide a framework for understanding cancer-specific gene regulation.

Whole genome deconvolution unveils Alzheimer's resilient epigenetic signature.

Assay for Transposase Accessible Chromatin by sequencing (ATAC-seq) accurately depicts the chromatin regulatory state and altered mechanisms guiding gene expression in disease. However, bulk sequencing entangles information from different cell types and obscures cellular heterogeneity. To address this, we developed Cellformer, a deep learning method that deconvolutes bulk ATAC-seq into cell type-specific expression across the whole genome. Cellformer enables cost-effective cell type-specific open chromatin profiling in large cohorts. Applied to 191 bulk samples from 3 brain regions, Cellformer identifies cell type-specific gene regulatory mechanisms involved in resilience to Alzheimer's disease, an uncommon group of cognitively healthy individuals that harbor a high pathological load of Alzheimer's disease. Cell type-resolved chromatin profiling unveils cell type-specific pathways and nominates potential epigenetic mediators underlying resilience that may illuminate therapeutic opportunities to limit the cognitive impact of the disease. Cellformer is freely available to facilitate future investigations using high-throughput bulk ATAC-seq data.

Systematic benchmarking of single-cell ATAC-sequencing protocols.

Single-cell assay for transposase-accessible chromatin by sequencing (scATAC-seq) has emerged as a powerful tool for dissecting regulatory landscapes and cellular heterogeneity. However, an exploration of systemic biases among scATAC-seq technologies has remained absent. In this study, we benchmark the performance of eight scATAC-seq methods across 47 experiments using human peripheral blood mononuclear cells (PBMCs) as a reference sample and develop PUMATAC, a universal preprocessing pipeline, to handle the various sequencing data formats. Our analyses reveal significant differences in sequencing library complexity and tagmentation specificity, which impact cell-type annotation, genotype demultiplexing, peak calling, differential region accessibility and transcription factor motif enrichment. Our findings underscore the importance of sample extraction, method selection, data processing and total cost of experiments, offering valuable guidance for future research. Finally, our data and analysis pipeline encompasses 169,000 PBMC scATAC-seq profiles and a best practices code repository for scATAC-seq data analysis, which are freely available to extend this benchmarking effort to future protocols.

Deep learning approaches for noncoding variant prioritization in neurodegenerative diseases.

Determining how noncoding genetic variants contribute to neurodegenerative dementias is fundamental to understanding disease pathogenesis, improving patient prognostication, and developing new clinical treatments. Next generation sequencing technologies have produced vast amounts of genomic data on cell type-specific transcription factor binding, gene expression, and three-dimensional chromatin interactions, with the promise of providing key insights into the biological mechanisms underlying disease. However, this data is highly complex, making it challenging for researchers to interpret, assimilate, and dissect. To this end, deep learning has emerged as a powerful tool for genome analysis that can capture the intricate patterns and dependencies within these large datasets. In this review, we organize and discuss the many unique model architectures, development philosophies, and interpretation methods that have emerged in the last few years with a focus on using deep learning to predict the impact of genetic variants on disease pathogenesis. We highlight both broadly-applicable genomic deep learning methods that can be fine-tuned to disease-specific contexts as well as existing neurodegenerative disease research, with an emphasis on Alzheimer's-specific literature. We conclude with an overview of the future of the field at the intersection of neurodegeneration, genomics, and deep learning.

Chromatin accessibility profiling by ATAC-seq.

The assay for transposase-accessible chromatin using sequencing (ATAC-seq) provides a simple and scalable way to detect the unique chromatin landscape associated with a cell type and how it may be altered by perturbation or disease. ATAC-seq requires a relatively small number of input cells and does not require a priori knowledge of the epigenetic marks or transcription factors governing the dynamics of the system. Here we describe an updated and optimized protocol for ATAC-seq, called Omni-ATAC, that is applicable across a broad range of cell and tissue types. The ATAC-seq workflow has five main steps: sample preparation, transposition, library preparation, sequencing and data analysis. This protocol details the steps to generate and sequence ATAC-seq libraries, with recommendations for sample preparation and downstream bioinformatic analysis. ATAC-seq libraries for roughly 12 samples can be generated in 10 h by someone familiar with basic molecular biology, and downstream sequencing analysis can be implemented using benchmarked pipelines by someone with basic bioinformatics skills and with access to a high-performance computing environment.

Epigenomic priming of immune genes implicates oligodendroglia in multiple sclerosis susceptibility.

Multiple sclerosis (MS) is characterized by a targeted attack on oligodendroglia (OLG) and myelin by immune cells, which are thought to be the main drivers of MS susceptibility. We found that immune genes exhibit a primed chromatin state in single mouse and human OLG in a non-disease context, compatible with transitions to immune-competent states in MS. We identified BACH1 and STAT1 as transcription factors involved in immune gene regulation in oligodendrocyte precursor cells (OPCs). A subset of immune genes presents bivalency of H3K4me3/H3K27me3 in OPCs, with Polycomb inhibition leading to their increased activation upon interferon gamma (IFN-γ) treatment. Some MS susceptibility single-nucleotide polymorphisms (SNPs) overlap with these regulatory regions in mouse and human OLG. Treatment of mouse OPCs with IFN-γ leads to chromatin architecture remodeling at these loci and altered expression of interacting genes. Thus, the susceptibility for MS may involve OLG, which therefore constitutes novel targets for immunological-based therapies for MS.

Single-cell spatial proteomic imaging for human neuropathology.

Neurodegenerative disorders are characterized by phenotypic changes and hallmark proteopathies. Quantifying these in archival human brain tissues remains indispensable for validating animal models and understanding disease mechanisms. We present a framework for nanometer-scale, spatial proteomics with multiplex ion beam imaging (MIBI) for capturing neuropathological features. MIBI facilitated simultaneous, quantitative imaging of 36 proteins on archival human hippocampus from individuals spanning cognitively normal to dementia. Customized analysis strategies identified cell types and proteopathies in the hippocampus across stages of Alzheimer's disease (AD) neuropathologic change. We show microglia-pathologic tau interactions in hippocampal CA1 subfield in AD dementia. Data driven, sample independent creation of spatial proteomic regions identified persistent neurons in pathologic tau neighborhoods expressing mitochondrial protein MFN2, regardless of cognitive status, suggesting a survival advantage. Our study revealed unique insights from multiplexed imaging and data-driven approaches for neuropathologic analysis and serves broadly as a methodology for spatial proteomic analysis of archival human neuropathology. TEASER: Multiplex Ion beam Imaging enables deep spatial phenotyping of human neuropathology-associated cellular and disease features.

The Cell Type-Specific 5hmC Landscape and Dynamics of Healthy Human Hematopoiesis and TET2-Mutant Preleukemia.

The conversion of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) is a key step in DNA demethylation that is mediated by ten-eleven translocation (TET) enzymes, which require ascorbate/vitamin C. Here, we report the 5hmC landscape of normal hematopoiesis and identify cell type-specific 5hmC profiles associated with active transcription and chromatin accessibility of key hematopoietic regulators. We utilized CRISPR/Cas9 to model TET2 loss-of-function mutations in primary human hematopoietic stem and progenitor cells (HSPC). Disrupted cells exhibited increased colonies in serial replating, defective erythroid/megakaryocytic differentiation, and in vivo competitive advantage and myeloid skewing coupled with reduction of 5hmC at erythroid-associated gene loci. Azacitidine and ascorbate restored 5hmC abundance and slowed or reverted the expansion of TET2-mutant clones in vivo. These results demonstrate the key role of 5hmC in normal hematopoiesis and TET2-mutant phenotypes and raise the possibility of utilizing these agents to further our understanding of preleukemia and clonal hematopoiesis. SIGNIFICANCE: We show that 5-hydroxymethylation profiles are cell type-specific and associated with transcriptional abundance and chromatin accessibility across human hematopoiesis. TET2 loss caused aberrant growth and differentiation phenotypes and disrupted 5hmC and transcriptional landscapes. Treatment of TET2 KO HSPCs with ascorbate or azacitidine reverted 5hmC profiles and restored aberrant phenotypes. This article is highlighted in the In This Issue feature, p. 265.

GATM and GAMT synthesize creatine locally throughout the mammalian body and within oligodendrocytes of the brain.

The enzymes glycine amidinotransferase, mitochondrial (GATM also known as AGAT) and guanidinoacetate N-methyltransferase (GAMT) function together to synthesize creatine from arginine, glycine, and S-Adenosyl methionine. Deficiency in either enzyme or the creatine transporter, CT1, results in a devastating neurological disorder, Cerebral Creatine Deficiency Syndrome (CCDS). To better understand the pathophysiology of CCDS, we mapped the distribution of GATM and GAMT at single cell resolution, leveraging RNA sequencing analysis combined with in vivo immunofluorescence (IF). Using the mouse as a model system, we find that GATM and GAMT are coexpressed in several tissues with distinct and overlapping cellular sources, implicating local synthesis as an important mechanism of creatine metabolism in numerous organs. Extending previous findings at the RNA level, our analysis demonstrates that oligodendrocytes express the highest level of Gatm and Gamt of any cell type in the body. We confirm this finding in the mouse brain by IF, where GATM localizes to the mitochondria of oligodendrocytes, whereas both oligodendrocytes and cerebral cortical neurons express GAMT. Interestingly, the latter is devoid of GATM. Single nucleus assay for transposase-accessible chromatin sequencing (snATAC-seq) analysis of 4 brain regions highlights a similar primacy of oligodendrocytes in the expression of GATM and GAMT in the human central nervous system. Importantly, an active putative regulatory element within intron 2 of human GATM is detected in oligodendrocytes but not neurons.