Integrated analysis of multimodal single-cell data.

The simultaneous measurement of multiple modalities represents an exciting frontier for single-cell genomics and necessitates computational methods that can define cellular states based on multimodal data. Here, we introduce "weighted-nearest neighbor" analysis, an unsupervised framework to learn the relative utility of each data type in each cell, enabling an integrative analysis of multiple modalities. We apply our procedure to a CITE-seq dataset of 211,000 human peripheral blood mononuclear cells (PBMCs) with panels extending to 228 antibodies to construct a multimodal reference atlas of the circulating immune system. Multimodal analysis substantially improves our ability to resolve cell states, allowing us to identify and validate previously unreported lymphoid subpopulations. Moreover, we demonstrate how to leverage this reference to rapidly map new datasets and to interpret immune responses to vaccination and coronavirus disease 2019 (COVID-19). Our approach represents a broadly applicable strategy to analyze single-cell multimodal datasets and to look beyond the transcriptome toward a unified and multimodal definition of cellular identity.

Comprehensive Integration of Single-Cell Data.

Single-cell transcriptomics has transformed our ability to characterize cell states, but deep biological understanding requires more than a taxonomic listing of clusters. As new methods arise to measure distinct cellular modalities, a key analytical challenge is to integrate these datasets to better understand cellular identity and function. Here, we develop a strategy to "anchor" diverse datasets together, enabling us to integrate single-cell measurements not only across scRNA-seq technologies, but also across different modalities. After demonstrating improvement over existing methods for integrating scRNA-seq data, we anchor scRNA-seq experiments with scATAC-seq to explore chromatin differences in closely related interneuron subsets and project protein expression measurements onto a bone marrow atlas to characterize lymphocyte populations. Lastly, we harmonize in situ gene expression and scRNA-seq datasets, allowing transcriptome-wide imputation of spatial gene expression patterns. Our work presents a strategy for the assembly of harmonized references and transfer of information across datasets.

A latent subset of human hematopoietic stem cells resists regenerative stress to preserve stemness.

Continuous supply of immune cells throughout life relies on the delicate balance in the hematopoietic stem cell (HSC) pool between long-term maintenance and meeting the demands of both normal blood production and unexpected stress conditions. Here we identified distinct subsets of human long-term (LT)-HSCs that responded differently to regeneration-mediated stress: an immune checkpoint ligand CD112lo subset that exhibited a transient engraftment restraint (termed latency) before contributing to hematopoietic reconstitution and a primed CD112hi subset that responded rapidly. This functional heterogeneity and CD112 expression are regulated by INKA1 through direct interaction with PAK4 and SIRT1, inducing epigenetic changes and defining an alternative state of LT-HSC quiescence that serves to preserve self-renewal and regenerative capacity upon regeneration-mediated stress. Collectively, our data uncovered the molecular intricacies underlying HSC heterogeneity and self-renewal regulation and point to latency as an orchestrated physiological response that balances blood cell demands with preserving a stem cell reservoir.

Efficient combinatorial targeting of RNA transcripts in single cells with Cas13 RNA Perturb-seq.

Pooled CRISPR screens coupled with single-cell RNA-sequencing have enabled systematic interrogation of gene function and regulatory networks. Here, we introduce Cas13 RNA Perturb-seq (CaRPool-seq), which leverages the RNA-targeting CRISPR-Cas13d system and enables efficient combinatorial perturbations alongside multimodal single-cell profiling. CaRPool-seq encodes multiple perturbations on a cleavable CRISPR array that is associated with a detectable barcode sequence, allowing for the simultaneous targeting of multiple genes. We compared CaRPool-seq to existing Cas9-based methods, highlighting its unique strength to efficiently profile combinatorially perturbed cells. Finally, we apply CaRPool-seq to perform multiplexed combinatorial perturbations of myeloid differentiation regulators in an acute myeloid leukemia (AML) model system and identify extensive interactions between different chromatin regulators that can enhance or suppress AML differentiation phenotypes.

Multiplexed detection of proteins, transcriptomes, clonotypes and CRISPR perturbations in single cells.

Multimodal single-cell assays provide high-resolution snapshots of complex cell populations, but are mostly limited to transcriptome plus an additional modality. Here, we describe expanded CRISPR-compatible cellular indexing of transcriptomes and epitopes by sequencing (ECCITE-seq) for the high-throughput characterization of at least five modalities of information from each single cell. We demonstrate application of ECCITE-seq to multimodal CRISPR screens with robust direct single-guide RNA capture and to clonotype-aware multimodal phenotyping of cancer samples.

Molecular transitions in early progenitors during human cord blood hematopoiesis.

Hematopoietic stem cells (HSCs) give rise to diverse cell types in the blood system, yet our molecular understanding of the early trajectories that generate this enormous diversity in humans remains incomplete. Here, we leverage Drop-seq, a massively parallel single-cell RNA sequencing (scRNA-seq) approach, to individually profile 20,000 progenitor cells from human cord blood, without prior enrichment or depletion for individual lineages based on surface markers. Our data reveal a transcriptional compendium of progenitor states in human cord blood, representing four committed lineages downstream from HSC, alongside the transcriptional dynamics underlying fate commitment. We identify intermediate stages that simultaneously co-express "primed" programs for multiple downstream lineages, and also observe striking heterogeneity in the early molecular transitions between myeloid subsets. Integrating our data with a recently published scRNA-seq dataset from human bone marrow, we illustrate the molecular similarity between these two commonly used systems and further explore the chromatin dynamics of "primed" transcriptional programs based on ATAC-seq. Finally, we demonstrate that Drop-seq data can be utilized to identify new heterogeneous surface markers of cell state that correlate with functional output.

Tumor-specific HSP90 inhibition as a therapeutic approach in JAK-mutant acute lymphoblastic leukemias.

The development of the dual Janus kinase 1/2 (JAK1/2) inhibitor ruxolitinib for the treatment of myeloproliferative neoplasms (MPNs) has led to studies of ruxolitinib in other clinical contexts, including JAK-mutated acute lymphoblastic leukemia (ALL). However, the limited ability of JAK inhibition to induce molecular or clinicopathological responses in MPNs suggests a need for development of better therapies for JAK kinase-dependent malignancies. Here, we demonstrate that heat shock protein 90 (HSP90) inhibition using a purine-scaffold HSP90 inhibitor in early clinical development is an effective therapeutic approach in JAK-dependent ALL and can overcome persistence to JAK-inhibitor therapy in ALL cells.

Characterizing the molecular regulation of inhibitory immune checkpoints with multimodal single-cell screens.

The expression of inhibitory immune checkpoint molecules, such as programmed death-ligand (PD-L)1, is frequently observed in human cancers and can lead to the suppression of T cell-mediated immune responses. Here, we apply expanded CRISPR-compatible (EC)CITE-seq, a technology that combines pooled CRISPR screens with single-cell mRNA and surface protein measurements, to explore the molecular networks that regulate PD-L1 expression. We also develop a computational framework, mixscape, that substantially improves the signal-to-noise ratio in single-cell perturbation screens by identifying and removing confounding sources of variation. Applying these tools, we identify and validate regulators of PD-L1 and leverage our multimodal data to identify both transcriptional and post-transcriptional modes of regulation. Specifically, we discover that the Kelch-like protein KEAP1 and the transcriptional activator NRF2 mediate the upregulation of PD-L1 after interferon (IFN)-γ stimulation. Our results identify a new mechanism for the regulation of immune checkpoints and present a powerful analytical framework for the analysis of multimodal single-cell perturbation screens.

Scalable, multimodal profiling of chromatin accessibility, gene expression and protein levels in single cells.

Recent technological advances have enabled massively parallel chromatin profiling with scATAC-seq (single-cell assay for transposase accessible chromatin by sequencing). Here we present ATAC with select antigen profiling by sequencing (ASAP-seq), a tool to simultaneously profile accessible chromatin and protein levels. Our approach pairs sparse scATAC-seq data with robust detection of hundreds of cell surface and intracellular protein markers and optional capture of mitochondrial DNA for clonal tracking, capturing three distinct modalities in single cells. ASAP-seq uses a bridging approach that repurposes antibody:oligonucleotide conjugates designed for existing technologies that pair protein measurements with single-cell RNA sequencing. Together with DOGMA-seq, an adaptation of CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) for measuring gene activity across the central dogma of gene regulation, we demonstrate the utility of systematic multi-omic profiling by revealing coordinated and distinct changes in chromatin, RNA and surface proteins during native hematopoietic differentiation and peripheral blood mononuclear cell stimulation and as a combinatorial decoder and reporter of multiplexed perturbations in primary T cells.

Single-cell RNA sequencing to explore immune cell heterogeneity.

Advances in single-cell RNA sequencing (scRNA-seq) have allowed for comprehensive analysis of the immune system. In this Review, we briefly describe the available scRNA-seq technologies together with their corresponding strengths and weaknesses. We discuss in depth how scRNA-seq can be used to deconvolve immune system heterogeneity by identifying novel distinct immune cell subsets in health and disease, characterizing stochastic heterogeneity within a cell population and building developmental 'trajectories' for immune cells. Finally, we discuss future directions of the field and present integrated approaches to complement molecular information from a single cell with studies of the environment, epigenetic state and cell lineage.

Integrating single-cell transcriptomic data across different conditions, technologies, and species.

Computational single-cell RNA-seq (scRNA-seq) methods have been successfully applied to experiments representing a single condition, technology, or species to discover and define cellular phenotypes. However, identifying subpopulations of cells that are present across multiple data sets remains challenging. Here, we introduce an analytical strategy for integrating scRNA-seq data sets based on common sources of variation, enabling the identification of shared populations across data sets and downstream comparative analysis. We apply this approach, implemented in our R toolkit Seurat (http://satijalab.org/seurat/), to align scRNA-seq data sets of peripheral blood mononuclear cells under resting and stimulated conditions, hematopoietic progenitors sequenced using two profiling technologies, and pancreatic cell 'atlases' generated from human and mouse islets. In each case, we learn distinct or transitional cell states jointly across data sets, while boosting statistical power through integrated analysis. Our approach facilitates general comparisons of scRNA-seq data sets, potentially deepening our understanding of how distinct cell states respond to perturbation, disease, and evolution.

Kinetics of adult hematopoietic stem cell differentiation in vivo.

Adult hematopoiesis has been studied in terms of progenitor differentiation potentials, whereas its kinetics in vivo is poorly understood. We combined inducible lineage tracing of endogenous adult hematopoietic stem cells (HSCs) with flow cytometry and single-cell RNA sequencing to characterize early steps of hematopoietic differentiation in the steady-state. Labeled cells, comprising primarily long-term HSCs and some short-term HSCs, produced megakaryocytic lineage progeny within 1 wk in a process that required only two to three cell divisions. Erythroid and myeloid progeny emerged simultaneously by 2 wk and included a progenitor population with expression features of both lineages. Myeloid progenitors at this stage showed diversification into granulocytic, monocytic, and dendritic cell types, and rare intermediate cell states could be detected. In contrast, lymphoid differentiation was virtually absent within the first 3 wk of tracing. These results show that continuous differentiation of HSCs rapidly produces major hematopoietic lineages and cell types and reveal fundamental kinetic differences between megakaryocytic, erythroid, myeloid, and lymphoid differentiation.

Dual Targeting of Oncogenic Activation and Inflammatory Signaling Increases Therapeutic Efficacy in Myeloproliferative Neoplasms.

Genetic and functional studies underscore the central role of JAK/STAT signaling in myeloproliferative neoplasms (MPNs). However, the mechanisms that mediate transformation in MPNs are not fully delineated, and clinically utilized JAK inhibitors have limited ability to reduce disease burden or reverse myelofibrosis. Here we show that MPN progenitor cells are characterized by marked alterations in gene regulation through differential enhancer utilization, and identify nuclear factor κB (NF-κB) signaling as a key pathway activated in malignant and non-malignant cells in MPN. Inhibition of BET bromodomain proteins attenuated NF-κB signaling and reduced cytokine production in vivo. Most importantly, combined JAK/BET inhibition resulted in a marked reduction in the serum levels of inflammatory cytokines, reduced disease burden, and reversed bone marrow fibrosis in vivo.

JAK2/IDH-mutant-driven myeloproliferative neoplasm is sensitive to combined targeted inhibition.

Patients with myeloproliferative neoplasms (MPNs) frequently progress to bone marrow failure or acute myeloid leukemia (AML), and mutations in epigenetic regulators such as the metabolic enzyme isocitrate dehydrogenase (IDH) are associated with poor outcomes. Here, we showed that combined expression of Jak2V617F and mutant IDH1R132H or Idh2R140Q induces MPN progression, alters stem/progenitor cell function, and impairs differentiation in mice. Jak2V617F Idh2R140Q-mutant MPNs were sensitive to small-molecule inhibition of IDH. Combined inhibition of JAK2 and IDH2 normalized the stem and progenitor cell compartments in the murine model and reduced disease burden to a greater extent than was seen with JAK inhibition alone. In addition, combined JAK2 and IDH2 inhibitor treatment also reversed aberrant gene expression in MPN stem cells and reversed the metabolite perturbations induced by concurrent JAK2 and IDH2 mutations. Combined JAK2 and IDH2 inhibitor therapy also showed cooperative efficacy in cells from MPN patients with both JAK2mut and IDH2mut mutations. Taken together, these data suggest that combined JAK and IDH inhibition may offer a therapeutic advantage in this high-risk MPN subtype.

Jak1 Integrates Cytokine Sensing to Regulate Hematopoietic Stem Cell Function and Stress Hematopoiesis.

JAK1 is a critical effector of pro-inflammatory cytokine signaling and plays important roles in immune function, while abnormal JAK1 activity has been linked to immunological and neoplastic diseases. Specific functions of JAK1 in the context of hematopoiesis, and specifically within hematopoietic stem cells (HSCs), have not clearly been delineated. Here, we show that conditional Jak1 loss in HSCs reduces their self-renewal and markedly alters lymphoid/myeloid differentiation in vivo. Jak1-deficient HSCs exhibit decreased competitiveness in vivo and are unable to rescue hematopoiesis in the setting of myelosuppression. They exhibit increased quiescence, an inability to enter the cell cycle in response to hematopoietic stress, and a marked reduction in cytokine sensing, including in response to type I interferons and IL-3. Moreover, Jak1 loss is not fully rescued by expression of a constitutively active Jak2 allele. Together, these data highlight an essential role for Jak1 in HSC homeostasis and stress responses.

Dose-dependent role of the cohesin complex in normal and malignant hematopoiesis.

Cohesin complex members have recently been identified as putative tumor suppressors in hematologic and epithelial malignancies. The cohesin complex guides chromosome segregation; however, cohesin mutant leukemias do not show genomic instability. We hypothesized that reduced cohesin function alters chromatin structure and disrupts cis-regulatory architecture of hematopoietic progenitors. We investigated the consequences of Smc3 deletion in normal and malignant hematopoiesis. Biallelic Smc3 loss induced bone marrow aplasia with premature sister chromatid separation and revealed an absolute requirement for cohesin in hematopoietic stem cell (HSC) function. In contrast, Smc3 haploinsufficiency increased self-renewal in vitro and in vivo, including competitive transplantation. Smc3 haploinsufficiency reduced coordinated transcriptional output, including reduced expression of transcription factors and other genes associated with lineage commitment. Smc3 haploinsufficiency cooperated with Flt3-ITD to induce acute leukemia in vivo, with potentiated Stat5 signaling and altered nucleolar topology. These data establish a dose dependency for cohesin in regulating chromatin structure and HSC function.

JAK-STAT pathway activation in malignant and nonmalignant cells contributes to MPN pathogenesis and therapeutic response.

UNLABELLED: The identification of JAK2/MPL mutations in patients with myeloproliferative neoplasms (MPN) has led to the clinical development of JAK kinase inhibitors, including ruxolitinib. Ruxolitinib reduces splenomegaly and systemic symptoms in myelofibrosis and improves overall survival; however, the mechanism by which JAK inhibitors achieve efficacy has not been delineated. Patients with MPN present with increased levels of circulating proinflammatory cytokines, which are mitigated by JAK inhibitor therapy. We sought to elucidate mechanisms by which JAK inhibitors attenuate cytokine-mediated pathophysiology. Single-cell profiling demonstrated that hematopoietic cells from myelofibrosis models and patient samples aberrantly secrete inflammatory cytokines. Pan-hematopoietic Stat3 deletion reduced disease severity and attenuated cytokine secretion, with similar efficacy as observed with ruxolitinib therapy. In contrast, Stat3 deletion restricted to MPN cells did not reduce disease severity or cytokine production. Consistent with these observations, we found that malignant and nonmalignant cells aberrantly secrete cytokines and JAK inhibition reduces cytokine production from both populations. SIGNIFICANCE: Our results demonstrate that JAK-STAT3-mediated cytokine production from malignant and nonmalignant cells contributes to MPN pathogenesis and that JAK inhibition in both populations is required for therapeutic efficacy. These findings provide novel insight into the mechanisms by which JAK kinase inhibition achieves therapeutic efficacy in MPNs.

Loss of BAP1 function leads to EZH2-dependent transformation.

The tumor suppressors BAP1 and ASXL1 interact to form a polycomb deubiquitinase complex that removes monoubiquitin from histone H2A lysine 119 (H2AK119Ub). However, BAP1 and ASXL1 are mutated in distinct cancer types, consistent with independent roles in regulating epigenetic state and malignant transformation. Here we demonstrate that Bap1 loss in mice results in increased trimethylated histone H3 lysine 27 (H3K27me3), elevated enhancer of zeste 2 polycomb repressive complex 2 subunit (Ezh2) expression, and enhanced repression of polycomb repressive complex 2 (PRC2) targets. These findings contrast with the reduction in H3K27me3 levels seen with Asxl1 loss. Conditional deletion of Bap1 and Ezh2 in vivo abrogates the myeloid progenitor expansion induced by Bap1 loss alone. Loss of BAP1 results in a marked decrease in H4K20 monomethylation (H4K20me1). Consistent with a role for H4K20me1 in the transcriptional regulation of EZH2, expression of SETD8-the H4K20me1 methyltransferase-reduces EZH2 expression and abrogates the proliferation of BAP1-mutant cells. Furthermore, mesothelioma cells that lack BAP1 are sensitive to EZH2 pharmacologic inhibition, suggesting a novel therapeutic approach for BAP1-mutant malignancies.