cellHarmony: cell-level matching and holistic comparison of single-cell transcriptomes

Temporal dynamics and genomic programming of plasma cell fates

Affinity-matured plasma cells (PCs) of varying lifespans are generated through a germinal center (GC) response. The developmental dynamics and genomic programs of antigen-specific PC precursors remain to be elucidated. Here, using a model antigen in mice, we demonstrate biphasic generation of PC precursors, with those generating long-lived bone marrow PCs preferentially produced in the late phase of GC response. Clonal tracing using single-cell RNA sequencing and B cell antigen receptor sequencing in spleen and bone marrow compartments, coupled with adoptive transfer experiments, reveals a new PC transition state that gives rise to functionally competent PC precursors. The latter undergo clonal expansion, dependent on inducible expression of TIGIT. We propose a model for the proliferation and programming of precursors of long-lived PCs, based on extended antigen encounters in the GC.

An immunophenotype-coupled transcriptomic atlas of human hematopoietic progenitors

Analysis of the human hematopoietic progenitor compartment is being transformed by single-cell multimodal approaches. Cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) enables coupled surface protein and transcriptome profiling, thereby revealing genomic programs underlying progenitor states. To perform CITE-seq systematically on primary human bone marrow cells, we used titrations with 266 CITE-seq antibodies (antibody-derived tags) and machine learning to optimize a panel of 132 antibodies. Multimodal analysis resolved >80 stem, progenitor, immune, stromal and transitional cells defined by distinctive surface markers and transcriptomes. This dataset enables flow cytometry solutions for in silico-predicted cell states and identifies dozens of cell surface markers consistently detected across donors spanning race and sex. Finally, aligning annotations from this atlas, we nominate normal marrow equivalents for acute myeloid leukemia stem cell populations that differ in clinical response. This atlas serves as an advanced digital resource for hematopoietic progenitor analyses in human health and disease.

Single-cell RNA sequencing of a new transgenic t(8;21) preleukemia mouse model reveals regulatory networks promoting leukemic transformation

T(8;21)(q22;q22), which generates the AML1-ETO fusion oncoprotein, is a common chromosomal abnormality in acute myeloid leukemia (AML) patients. Despite having favorable prognosis, 40% of patients will relapse, highlighting the need for innovative models and application of the newest technologies to study t(8;21) leukemogenesis. Currently, available AML1-ETO mouse models have limited utility for studying the pre-leukemic stage because AML1-ETO produces mild hematopoietic phenotypes and no leukemic transformation. Conversely, overexpression of a truncated variant, AML1-ETO9a (AE9a), promotes fully penetrant leukemia and is too potent for studying pre-leukemic changes. To overcome these limitations, we devised a germline-transmitted Rosa26 locus AE9a knock-in mouse model that moderately overexpressed AE9a and developed leukemia with long latency and low penetrance. We observed pre-leukemic alterations in AE9a mice, including skewing of progenitors towards granulocyte/monocyte lineages and replating of stem and progenitor cells. Next, we performed single-cell RNA sequencing to identify specific cell populations that contribute to these pre-leukemic phenotypes. We discovered a subset of common myeloid progenitors that have heightened granulocyte/monocyte bias in AE9a mice. We also observed dysregulation of key hematopoietic transcription factor target gene networks, blocking cellular differentiation. Finally, we identified Sox4 activation as a potential contributor to stem cell self-renewal during the pre-leukemic stage.

Population-level single-cell genomics reveals conserved gene programs in systemic juvenile idiopathic arthritis

Systemic autoimmune and autoinflammatory diseases are characterized by genetic and cellular heterogeneity. While current single-cell genomics methods provide insights into known disease subtypes, these analysis methods do not readily reveal novel cell-type perturbation programs shared among distinct patient subsets. Here, we performed single-cell RNA-Seq of PBMCs of patients with systemic juvenile idiopathic arthritis (SJIA) with diverse clinical manifestations, including macrophage activation syndrome (MAS) and lung disease (LD). We introduced two new computational frameworks called UDON and SATAY-UDON, which define patient subtypes based on their underlying disrupted cellular programs as well as associated biomarkers or clinical features. Among twelve independently identified subtypes, this analysis uncovered what we believe to be a novel complement and interferon activation program identified in SJIA-LD monocytes. Extending these analyses to adult and pediatric lupus patients found new but also shared disease programs with SJIA, including interferon and complement activation. Finally, supervised comparison of these programs in a compiled single-cell pan-immune atlas of over 1,000 healthy donors found a handful of normal healthy donors with evidence of early inflammatory activation in subsets of monocytes and platelets, nominating possible biomarkers for early disease detection. Thus, integrative pan-immune single-cell analysis resolved what we believe to be new conserved gene programs underlying inflammatory disease pathogenesis and associated complications.

Sinus venosus adaptation models prolonged cardiovascular disease and reveals insights into evolutionary transitions of the vertebrate heart

How two-chambered hearts in basal vertebrates have evolved from single-chamber hearts found in ancestral chordates remains unclear. Here, we show that the teleost sinus venosus (SV) is a chamber-like vessel comprised of an outer layer of smooth muscle cells. We find that in adult zebrafish nr2f1a mutants, which lack atria, the SV comes to physically resemble the thicker bulbus arteriosus (BA) at the arterial pole of the heart through an adaptive, hypertensive response involving smooth muscle proliferation due to aberrant hemodynamic flow. Single cell transcriptomics show that smooth muscle and endothelial cell populations within the adapting SV also take on arterial signatures. Bulk transcriptomics of the blood sinuses flanking the tunicate heart reinforce a model of greater equivalency in ancestral chordate BA and SV precursors. Our data simultaneously reveal that secondary complications from congenital heart defects can develop in adult zebrafish similar to those in humans and that the foundation of equivalency between flanking auxiliary vessels may remain latent within basal vertebrate hearts.

Temporal dynamics and genomic programming of plasma cell fates

Affinity-matured plasma cells (PCs) of varying lifespans are generated through a germinal center (GC) response. The developmental dynamics and genomic programs of antigen-specific PC precursors remain to be elucidated. Using a model antigen, we demonstrate biphasic generation of PC precursors, with those generating long-lived bone marrow PCs preferentially produced in the late phase of GC response. Clonal tracing using scRNA-seq+BCR-seq in spleen and bone marrow compartments, coupled with adoptive transfer experiments, reveal a novel PC transition state that gives rise to functionally competent PC precursors. The latter undergo clonal expansion, dependent on inducible expression of TIGIT. We propose a model for the proliferation and programming of precursors of long-lived PCs, based on extended antigen encounters followed by reduced antigen availability.

ThPOK is a critical multifaceted regulator of myeloid lineage development

The transcription factor ThPOK (encoded by Zbtb7b) is well known for its role as a master regulator of CD4 lineage commitment in the thymus. Here, we report an unexpected and critical role of ThPOK as a multifaceted regulator of myeloid lineage commitment, differentiation and maturation. Using reporter and knockout mouse models combined with single-cell RNA-sequencing, progenitor transfer and colony assays, we show that ThPOK controls monocyte-dendritic cell versus granulocyte lineage production during homeostatic differentiation, and serves as a brake for neutrophil maturation in granulocyte lineage-specified cells through transcriptional regulation of lineage-specific transcription factors and RNA via altered messenger RNA splicing to reprogram intron retention.

Guided construction of single cell reference for human and mouse lung

Accurate cell type identification is a key and rate-limiting step in single-cell data analysis. Single-cell references with comprehensive cell types, reproducible and functionally validated cell identities, and common nomenclatures are much needed by the research community for automated cell type annotation, data integration, and data sharing. Here, we develop a computational pipeline utilizing the LungMAP CellCards as a dictionary to consolidate single-cell transcriptomic datasets of 104 human lungs and 17 mouse lung samples to construct LungMAP single-cell reference (CellRef) for both normal human and mouse lungs. CellRefs define 48 human and 40 mouse lung cell types catalogued from diverse anatomic locations and developmental time points. We demonstrate the accuracy and stability of LungMAP CellRefs and their utility for automated cell type annotation of both normal and diseased lungs using multiple independent methods and testing data. We develop user-friendly web interfaces for easy access and maximal utilization of the LungMAP CellRefs.

BAFF and APRIL counterregulate susceptibility to inflammation-induced preterm birth

Clinical evidence points to a function for B cell-activating factor (BAFF) in pregnancy. However, direct roles for BAFF-axis members in pregnancy have not been examined. Here, via utility of genetically modified mice, we report that BAFF promotes inflammatory responsiveness and increases susceptibility to inflammation-induced preterm birth (PTB). In contrast, we show that the closely related A proliferation-inducing ligand (APRIL) decreases inflammatory responsiveness and susceptibility to PTB. Known BAFF-axis receptors serve a redundant function in signaling BAFF/APRIL presence in pregnancy. Treatment with anti-BAFF/APRIL monoclonal antibodies or BAFF/APRIL recombinant proteins is sufficient to manipulate susceptibility to PTB. Notably, macrophages at the maternal-fetal interface produce BAFF, while BAFF and APRIL presence divergently shape macrophage gene expression and inflammatory function. Overall, our findings demonstrate that BAFF and APRIL play divergent inflammatory roles in pregnancy and provide therapeutic targets for mitigating risk of inflammation-induced PTB.

Multi-species meta-analysis identifies transcriptional signatures associated with cardiac endothelial responses in the ischaemic heart

AIM

Myocardial infarction remains the leading cause of heart failure. The adult human heart lacks the capacity to undergo endogenous regeneration. New blood vessel growth is integral to regenerative medicine necessitating a comprehensive understanding of the pathways that regulate vascular regeneration. We sought to define the transcriptomic dynamics of coronary endothelial cells following ischaemic injuries in the developing and adult mouse and human heart and to identify new mechanistic insights and targets for cardiovascular regeneration.

METHODS AND RESULTS

We carried out a comprehensive meta-analysis of integrated single-cell RNA-sequencing data of coronary vascular endothelial cells from the developing and adult mouse and human heart spanning healthy and acute and chronic ischaemic cardiac disease. We identified species-conserved gene regulatory pathways aligned to endogenous neovascularization. We annotated injury-associated temporal shifts of the endothelial transcriptome and validated four genes: VEGF-C, KLF4, EGR1, and ZFP36. Moreover, we showed that ZFP36 regulates human coronary endothelial cell proliferation and defined that VEGF-C administration in vivo enhances clonal expansion of the cardiac vasculature post-myocardial infarction. Finally, we constructed a coronary endothelial cell meta-atlas, CrescENDO, to empower future in-depth research to target pathways associated with coronary neovascularization.

CONCLUSION

We present a high-resolution single-cell meta-atlas of healthy and injured coronary endothelial cells in the mouse and human heart, revealing a suite of novel targets with great potential to promote vascular regeneration, and providing a rich resource for therapeutic development.

Expansion of interferon inducible gene pool via USP18 inhibition promotes cancer cell pyroptosis

While immunotherapy has emerged as a breakthrough cancer therapy, it is only effective in some patients, indicating the need of alternative therapeutic strategies. Induction of cancer immunogenic cell death (ICD) is one promising way to elicit potent adaptive immune responses against tumor-associated antigens. Type I interferon (IFN) is well known to play important roles in different aspects of immune responses, including modulating ICD in anti-tumor action. However, how to expand IFN effect in promoting ICD responses has not been addressed. Here we show that depletion of ubiquitin specific protease 18 (USP18), a negative regulator of IFN signaling, selectively induces cancer cell ICD. Lower USP18 expression correlates with better survival across human selected cancer types and delays cancer progression in mouse models. Mechanistically, nuclear USP18 controls the enhancer landscape of cancer cells and diminishes STAT2-mediated transcription complex binding to IFN-responsive elements. Consequently, USP18 suppression not only enhances expression of canonical IFN-stimulated genes (ISGs), but also activates the expression of a set of atypical ISGs and NF-κB target genes, including genes such as Polo like kinase 2 (PLK2), that induce cancer pyroptosis. These findings may support the use of targeting USP18 as a potential cancer immunotherapy.

Single-cell RNA-sequence analysis of human bone marrow reveals new targets for isolation of skeletal stem cells using spherical nucleic acids

There is a wealth of data indicating human bone marrow contains skeletal stem cells (SSC) with the capacity for osteogenic, chondrogenic and adipogenic differentiation. However, current methods to isolate SSCs are restricted by the lack of a defined marker, limiting understanding of SSC fate, immunophenotype, function and clinical application. The current study applied single-cell RNA-sequencing to profile human adult bone marrow populations from 11 donors and identified novel targets for SSC enrichment. Spherical nucleic acids were used to detect these mRNA targets in SSCs. This methodology was able to rapidly isolate potential SSCs found at a frequency of <1 in 1,000,000 in human bone marrow, with the capacity for tri-lineage differentiation in vitro and ectopic bone formation in vivo. The current studies detail the development of a platform to advance SSC enrichment from human bone marrow, offering an invaluable resource for further SSC characterisation, with significant therapeutic impact therein.

Automatic Cell Type Annotation Using Marker Genes for Single-Cell RNA Sequencing Data

Recent advancement in single-cell RNA sequencing (scRNA-seq) technology is gaining more and more attention. Cell type annotation plays an essential role in scRNA-seq data analysis. Several computational methods have been proposed for automatic annotation. Traditional cell type annotation is to first cluster the cells using unsupervised learning methods based on the gene expression profiles, then to label the clusters using the aggregated cluster-level expression profiles and the marker genes’ information. Such procedure relies heavily on the clustering results. As the purity of clusters cannot be guaranteed, false detection of cluster features may lead to wrong annotations. In this paper, we improve this procedure and propose an Automatic Cell type Annotation Method (ACAM). ACAM delineates a clear framework to conduct automatic cell annotation through representative cluster identification, representative cluster annotation using marker genes, and the remaining cells’ classification. Experiments on seven real datasets show the better performance of ACAM compared to six well-known cell type annotation methods.

Cellular and transcriptional dynamics of human neutrophils at steady state and upon stress

Traditionally viewed as poorly plastic, neutrophils are now recognized as functionally diverse; however, the extent and determinants of neutrophil heterogeneity in humans remain unclear. We performed a comprehensive immunophenotypic and transcriptome analysis, at a bulk and single-cell level, of neutrophils from healthy donors and patients undergoing stress myelopoiesis upon exposure to growth factors, transplantation of hematopoietic stem cells (HSC-T), development of pancreatic cancer and viral infection. We uncover an extreme diversity of human neutrophils in vivo, reflecting the rates of cell mobilization, differentiation and exposure to environmental signals. Integrated control of developmental and inducible transcriptional programs linked flexible granulopoietic outputs with elicitation of stimulus-specific functional responses. In this context, we detected an acute interferon (IFN) response in the blood of patients receiving HSC-T that was mirrored by marked upregulation of IFN-stimulated genes in neutrophils but not in monocytes. Systematic characterization of human neutrophil plasticity may uncover clinically relevant biomarkers and support the development of diagnostic and therapeutic tools.

Multiple Nf1 Schwann cell populations reprogram the plexiform neurofibroma tumor microenvironment

To define alterations early in tumor formation, we studied nerve tumors in neurofibromatosis 1 (NF1), a tumor predisposition syndrome. Affected individuals develop neurofibromas, benign tumors driven by NF1 loss in Schwann cells (SCs). By comparing normal nerve cells to plexiform neurofibroma (PN) cells using single-cell and bulk RNA sequencing, we identified changes in 5 SC populations, including a de novo SC progenitor-like (SCP-like) population. Long after Nf1 loss, SC populations developed PN-specific expression of Dcn, Postn, and Cd74, with sustained expression of the injury response gene Postn and showed dramatic expansion of immune and stromal cell populations; in corresponding human PNs, the immune and stromal cells comprised 90% of cells. Comparisons between injury-related and tumor monocytes/macrophages support early monocyte recruitment and aberrant macrophage differentiation. Cross-species analysis verified each SC population and unique conserved patterns of predicted cell-cell communication in each SC population. This analysis identified PROS1-AXL, FGF-FGFR, and MIF-CD74 and its effector pathway NF-κB as deregulated in NF1 SC populations, including SCP-like cells predicted to influence other types of SCs, stromal cells, and/or immune cells in mouse and human. These findings highlight remarkable changes in multiple types of SCs and identify therapeutic targets for PN.

scMTD: a statistical multidimensional imputation method for single-cell RNA-seq data leveraging transcriptome dynamic information

Background

Single-cell RNA sequencing (scRNA-seq) provides a powerful tool to capture transcriptomes at single-cell resolution. However, dropout events distort the gene expression levels and underlying biological signals, misleading the downstream analysis of scRNA-seq data.

Results

We develop a statistical model-based multidimensional imputation algorithm, scMTD, that identifies local cell neighbors and specific gene co-expression networks based on the pseudo-time of cells, leveraging information on cell-level, gene-level, and transcriptome dynamic to recover scRNA-seq data. Compared with the state-of-the-art imputation methods through several real-data-based analytical experiments, scMTD effectively recovers biological signals of transcriptomes and consistently outperforms the other algorithms in improving FISH validation, trajectory inference, differential expression analysis, clustering analysis, and identification of cell types.

Conclusions

scMTD maintains the gene expression characteristics, enhances the clustering of cell subpopulations, assists the study of gene expression dynamics, contributes to the discovery of rare cell types, and applies to both UMI-based and non-UMI-based data. Overall, scMTD’s reliability, applicability, and scalability make it a promising imputation approach for scRNA-seq data.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13578-022-00886-4.

The balance between protective and pathogenic immune responses to pneumonia in the neonatal lung is enforced by gut microbiota

Although modern clinical practices such as cesarean sections and perinatal antibiotics have improved infant survival, treatment with broad-spectrum antibiotics alters intestinal microbiota and causes dysbiosis. Infants exposed to perinatal antibiotics have an increased likelihood of life-threatening infections, including pneumonia. Here, we investigated how the gut microbiota sculpt pulmonary immune responses, promoting recovery and resolution of infection in newborn rhesus macaques. Early-life antibiotic exposure interrupted the maturation of intestinal commensal bacteria and disrupted the developmental trajectory of the pulmonary immune system, as assessed by single-cell proteomic and transcriptomic analyses. Early-life antibiotic exposure rendered newborn macaques more susceptible to bacterial pneumonia, concurrent with increases in neutrophil senescence and hyperinflammation, broad inflammatory cytokine signaling, and macrophage dysfunction. This pathogenic reprogramming of pulmonary immunity was further reflected by a hyperinflammatory signature in all pulmonary immune cell subsets coupled with a global loss of tissue-protective, homeostatic pathways in the lungs of dysbiotic newborns. Fecal microbiota transfer was associated with partial correction of the broad immune maladaptations and protection against severe pneumonia. These data demonstrate the importance of intestinal microbiota in programming pulmonary immunity and support the idea that gut microbiota promote the balance between pathways driving tissue repair and inflammatory responses associated with clinical recovery from infection in infants. Our results highlight a potential role for microbial transfer for immune support in these at-risk infants.

A potent myeloid response is rapidly activated in the lungs of premature Rhesus macaques exposed to intra-uterine inflammation

Up to 40% of preterm births are associated with histological chorioamnionitis (HCA), which leads to elevated levels of pro-inflammatory mediators and microbial products in the amniotic fluid, which come in contact with fetal lungs. Yet, fetal pulmonary immune responses to such exposure remain poorly characterized. To address this gap, we used our established HCA model, in which pregnant Rhesus macaques receive intraamniotic (IA) saline or LPS. IA LPS induced a potent and rapid myeloid cell response in fetal lungs, dominated by neutrophils and monocytes/macrophages. Infiltrating and resident myeloid cells exhibited transcriptional profiles consistent with exposure to TLR ligands, as well as cytokines, notably IL-1 and TNFα. Although simultaneous, in vivo blockade of IL-1 and TNFα signaling did not prevent the inflammatory cell recruitment, it blunted the lung overall inflammatory state reducing communication between, and activation of, infiltrating immune cells. Our data indicate that the fetal innate immune system can mount a rapid multi-faceted pulmonary immune response to in utero exposure to inflammation. These data provide mechanistic insights into the association between HCA and the postnatal lung morbidities of the premature infant and highlight therapeutic potential of inflammatory blockade in the fetus.

Inflammatory blockade prevents injury to the developing pulmonary gas exchange surface in preterm primates

Perinatal inflammatory stress is associated with early life morbidity and lifelong consequences for pulmonary health. Chorioamnionitis, an inflammatory condition affecting the placenta and fluid surrounding the developing fetus, affects 25 to 40% of preterm births. Severe chorioamnionitis with preterm birth is associated with significantly increased risk of pulmonary disease and secondary infections in childhood, suggesting that fetal inflammation may markedly alter the development of the lung. Here, we used intra-amniotic lipopolysaccharide (LPS) challenge to induce experimental chorioamnionitis in a prenatal rhesus macaque (Macaca mulatta) model that mirrors structural and temporal aspects of human lung development. Inflammatory injury directly disrupted the developing gas exchange surface of the primate lung, with extensive damage to alveolar structure, particularly the close association and coordinated differentiation of alveolar type 1 pneumocytes and specialized alveolar capillary endothelium. Single-cell RNA sequencing analysis defined a multicellular alveolar signaling niche driving alveologenesis that was extensively disrupted by perinatal inflammation, leading to a loss of gas exchange surface and alveolar simplification, with notable resemblance to chronic lung disease in newborns. Blockade of the inflammatory cytokines interleukin-1β and tumor necrosis factor-α ameliorated LPS-induced inflammatory lung injury by blunting stromal responses to inflammation and modulating innate immune activation in myeloid cells, restoring structural integrity and key signaling networks in the developing alveolus. These data provide new insight into the pathophysiology of developmental lung injury and suggest that modulating inflammation is a promising therapeutic approach to prevent fetal consequences of chorioamnionitis.

A Pathogenic Th17/CD38+ Macrophage Feedback Loop Drives Inflammatory Arthritis through TNF-α

The pathobiology of rheumatoid inflammatory diseases, including rheumatoid arthritis (RA) and psoriatic arthritis, involves the interplay between innate and adaptive immune components and resident synoviocytes. Single-cell analyses of patient samples and relevant mouse models have characterized many cellular subsets in RA. However, the impact of interactions between cell types is not fully understood. In this study, we temporally profiled murine arthritic synovial isolates at the single-cell level to identify perturbations similar to those found in human RA. Notably, murine macrophage subtypes like those found in RA patients were expanded in arthritis and linked to promoting the function of Th17 cells in the joint. In vitro experiments identified a capacity for murine macrophages to maintain the functionality and expansion of Th17 cells. Reciprocally, murine Th17 cell-derived TNF-α induced CD38⁺ macrophages that enhanced Th17 functionality. Murine synovial CD38⁺ macrophages were expanded during arthritis, and their depletion or blockade via TNF-α neutralization alleviated disease while reducing IL-17A-producing cells. These findings identify a cellular feedback loop that promotes Th17 cell pathogenicity through TNF-α to drive inflammatory arthritis.

Single-Cell Multiomics Reveals Clonal T-Cell Expansions and Exhaustion in Blastic Plasmacytoid Dendritic Cell Neoplasm

The immune system represents a major barrier to cancer progression, driving the evolution of immunoregulatory interactions between malignant cells and T-cells in the tumor environment. Blastic plasmacytoid dendritic cell neoplasm (BPDCN), a rare acute leukemia with plasmacytoid dendritic cell (pDC) differentiation, provides a unique opportunity to study these interactions. pDCs are key producers of interferon alpha (IFNA) that play an important role in T-cell activation at the interface between the innate and adaptive immune system. To assess how uncontrolled proliferation of malignant BPDCN cells affects the tumor environment, we catalog immune cell heterogeneity in the bone marrow (BM) of five healthy controls and five BPDCN patients by analyzing 52,803 single-cell transcriptomes, including 18,779 T-cells. We test computational techniques for robust cell type classification and find that T-cells in BPDCN patients consistently upregulate interferon alpha (IFNA) response and downregulate tumor necrosis factor alpha (TNFA) pathways. Integrating transcriptional data with T-cell receptor sequencing via shared barcodes reveals significant T-cell exhaustion in BPDCN that is positively correlated with T-cell clonotype expansion. By highlighting new mechanisms of T-cell exhaustion and immune evasion in BPDCN, our results demonstrate the value of single-cell multiomics to understand immune cell interactions in the tumor environment.

GM-CSF Programs Hematopoietic Stem and Progenitor Cells During Candida albicans Vaccination for Protection Against Reinfection

More mechanistic studies are needed to reveal the hidden details of in vivo-induced trained immunity. Here, using a Candida albicans live vaccine mouse model we show that vaccination protects mice against a secondary infection and increases the number of bone marrow, and especially, splenic trained monocytes. Moreover, vaccination expands and reprograms hematopoietic stem and progenitor cells (HSPCs) early during infection and mobilize them transiently to the spleen to produce trained macrophages. Trained HSPCs are not only primed for myeloid cell production but also reprogramed to produce a greater amount of proinflammatory cytokines in response to a second challenge. Additionally, their adoptive transfer is sufficient to protect mice against reinfection. Mechanistically, autocrine GM-CSF activation of HSPCs is responsible for the trained phenotype and essential for the vaccine-induced protection. Our findings reveal a fundamental role for HSPCs in the trained immune protective response, opening new avenues for disease prevention and treatment.

Progenitor translatome changes coordinated by Tsc1 increase perception of Wnt signals to end nephrogenesis

Mammalian nephron endowment is determined by the coordinated cessation of nephrogenesis in independent niches. Here we report that translatome analysis in Tsc1+/- nephron progenitor cells from mice with elevated nephron numbers reveals how differential translation of Wnt antagonists over agonists tips the balance between self-renewal and differentiation. Wnt agonists are poorly translated in young niches, resulting in an environment with low R-spondin and high Fgf20 promoting self-renewal. In older niches we find increased translation of Wnt agonists, including R-spondin and the signalosome-promoting Tmem59, and low Fgf20, promoting differentiation. This suggests that the tipping point for nephron progenitor exit from the niche is controlled by the gradual increase in stability and possibly clustering of Wnt/Fzd complexes in individual cells, enhancing the response to ureteric bud-derived Wnt9b inputs and driving synchronized differentiation. As predicted by these findings, removing one Rspo3 allele in nephron progenitors delays cessation and increases nephron numbers in vivo.

Integrating single-cell genomics pipelines to discover mechanisms of stem cell differentiation

Pluripotent stem cells underpin a growing sector that leverages their differentiation potential for research, industry, and clinical applications. This review evaluates the landscape of methods in single-cell transcriptomics that are enabling accelerated discovery in stem cell science. We focus on strategies for scaling stem cell differentiation through multiplexed single-cell analyses, for evaluating molecular regulation of cell differentiation using new analysis algorithms, and methods for integration and projection analysis to classify and benchmark stem cell derivatives against in vivo cell types. By discussing the available methods, comparing their strengths, and illustrating strategies for developing integrated analysis pipelines, we provide user considerations to inform their implementation and interpretation.

Cell fate dynamics and genomic programming of plasma cell precursors

This review is focused on the cellular dynamics and genomic programming of plasma cell (PC) precursors that arise during germinal center (GC) B cell responses in secondary lymphoid organs (SLOs) and give rise to PCs in the bone marrow. Considerable progress has been made in the phenotypic characterization of circulating and bone marrow PC precursors as well as their differentiated short-lived (SLPC) and long-lived (LLPC) counterparts, in the context of model antigen and vaccine responses. Importantly, it has been possible to infer the temporal dynamics of generation of PC precursors during a GC response. However, the nature of the PC precursors at their site of generation in SLOs, and their signaling and genomic states, remain to be elucidated. Our synthesis draws upon experimental studies conducted in murine models as well as in humans, the latter complemented with cell culture manipulations of PCs and their precursors. By integration of the studies in murine and human systems, which are being accelerated by new genomic methodologies, we highlight insights and hypotheses concerning the generation of PCs. This framework can be extended and explored from both fundamental and translational standpoints.

Insights into how environment shapes post-mortem RNA transcription in mouse brain

Most biological features that occur on the body after death were already deciphered by traditional medicine. However, the molecular mechanisms triggered in the cellular microenvironment are not fully comprehended yet. Previous studies reported gene expression alterations in the post-mortem condition, but little is known about how the environment could influence RNA degradation and transcriptional regulation. In this work, we analysed the transcriptome of mouse brain after death under three concealment simulations (air exposed, buried, and submerged). Our analyses identified 2,103 genes differentially expressed in all tested groups 48 h after death. Moreover, we identified 111 commonly upregulated and 497 commonly downregulated genes in mice from the concealment simulations. The gene functions shared by the individuals from the tested environments were associated with RNA homeostasis, inflammation, developmental processes, cell communication, cell proliferation, and lipid metabolism. Regarding the altered biological processes, we identified that the macroautophagy process was enriched in the upregulated genes and lipid metabolism was enriched in the downregulated genes. On the other hand, we also described a list of biomarkers associated with the submerged and buried groups, indicating that these environments can influence the post-mortem RNA abundance in its particular way.

PKM2-dependent metabolic skewing of hepatic Th17 cells regulates pathogenesis of non-alcoholic fatty liver disease

Emerging evidence suggests a key contribution to non-alcoholic fatty liver disease (NAFLD) pathogenesis by Th17 cells. The pathogenic characteristics and mechanisms of hepatic Th17 cells, however, remain unknown. Here, we uncover and characterize a distinct population of inflammatory hepatic CXCR3+Th17 (ihTh17) cells sufficient to exacerbate NAFLD pathogenesis. Hepatic ihTh17 cell accrual was dependent on the liver microenvironment and CXCR3 axis activation. Mechanistically, the pathogenic potential of ihTh17 cells correlated with increased chromatin accessibility, glycolytic output, and concomitant production of IL-17A, IFNγ, and TNFα. Modulation of glycolysis using 2-DG or cell-specific PKM2 deletion was sufficient to reverse ihTh17-centric inflammatory vigor and NAFLD severity. Importantly, ihTh17 cell characteristics, CXCR3 axis activation, and hepatic expression of glycolytic genes were conserved in human NAFLD. Together, our data show that the steatotic liver microenvironment regulates Th17 cell accrual, metabolism, and competence toward an ihTh17 fate. Modulation of these pathways holds potential for development of novel therapeutic strategies for NAFLD.

Machine Intelligence in Single-Cell Data Analysis: Advances and New Challenges

The rapid development of single-cell technologies allows for dissecting cellular heterogeneity at different omics layers with an unprecedented resolution. In-dep analysis of cellular heterogeneity will boost our understanding of complex biological systems or processes, including cancer, immune system and chronic diseases, thereby providing valuable insights for clinical and translational research. In this review, we will focus on the application of machine learning methods in single-cell multi-omics data analysis. We will start with the pre-processing of single-cell RNA sequencing (scRNA-seq) data, including data imputation, cross-platform batch effect removal, and cell cycle and cell-type identification. Next, we will introduce advanced data analysis tools and methods used for copy number variance estimate, single-cell pseudo-time trajectory analysis, phylogenetic tree inference, cell-cell interaction, regulatory network inference, and integrated analysis of scRNA-seq and spatial transcriptome data. Finally, we will present the latest analyzing challenges, such as multi-omics integration and integrated analysis of scRNA-seq data.

The Rhesus Macaque Serves As a Model for Human Lateral Branch Nephrogenesis

BACKGROUND

Most nephrons are added in late gestation. Truncated extrauterine nephrogenesis in premature infants results in fewer nephrons and significantly increased risk for CKD in adulthood. To overcome the ethical and technical difficulties associated with studies of late-gestation human fetal kidney development, third-trimester rhesus macaques served as a model to understand lateral branch nephrogenesis (LBN) at the molecular level.

METHODS

Immunostaining and 3D rendering assessed morphology. Single-cell (sc) and single-nucleus (sn) RNA-Seq were performed on four cortically enriched fetal rhesus kidneys of 129-131 days gestational age (GA). An integrative bioinformatics strategy was applied across single-cell modalities, species, and time. RNAScope validation studies were performed on human archival tissue.

RESULTS

Third-trimester rhesus kidney undergoes human-like LBN. scRNA-Seq of 23,608 cells revealed 37 transcriptionally distinct cell populations, including naïve nephron progenitor cells (NPCs), with the prior noted marker genes CITED1, MEOX1, and EYA1 (c25). These same populations and markers were reflected in snRNA-Seq of 5972 nuclei. Late-gestation rhesus NPC markers resembled late-gestation murine NPC, whereas early second-trimester human NPC markers aligned to midgestation murine NPCs. New, age-specific rhesus NPCs (SHISA8) and ureteric buds (POU3F4 and TWIST) predicted markers were verified in late-gestation human archival samples.

CONCLUSIONS

Rhesus macaque is the first model of bona fide LBN, enabling molecular studies of late gestation, human-like nephrogenesis. These molecular findings support the hypothesis that aging nephron progenitors have a distinct molecular signature and align to their earlier human counterparts, with unique markers highlighting LBN-specific progenitor maturation.

Improving Single-Cell RNA-seq Clustering by Integrating Pathways

Single-cell clustering is an important part of analyzing single-cell RNA-sequencing data. However, the accuracy and robustness of existing methods are disturbed by noise. One promising approach for addressing this challenge is integrating pathway information, which can alleviate noise and improve performance. In this work, we studied the impact on accuracy and robustness of existing single-cell clustering methods by integrating pathways. We collected 10 state-of-the-art single-cell clustering methods, 26 scRNA-seq datasets and four pathway databases, combined the AUCell method and the similarity network fusion to integrate pathway data and scRNA-seq data, and introduced three accuracy indicators, three noise generation strategies and robustness indicators. Experiments on this framework showed that integrating pathways can significantly improve the accuracy and robustness of most single-cell clustering methods.

Quantifying the effect of experimental perturbations at single-cell resolution

Current methods for comparing single-cell RNA sequencing datasets collected in multiple conditions focus on discrete regions of the transcriptional state space, such as clusters of cells. Here we quantify the effects of perturbations at the single-cell level using a continuous measure of the effect of a perturbation across the transcriptomic space. We describe this space as a manifold and develop a relative likelihood estimate of observing each cell in each of the experimental conditions using graph signal processing. This likelihood estimate can be used to identify cell populations specifically affected by a perturbation. We also develop vertex frequency clustering to extract populations of affected cells at the level of granularity that matches the perturbation response. The accuracy of our algorithm at identifying clusters of cells that are enriched or depleted in each condition is, on average, 57% higher than the next-best-performing algorithm tested. Gene signatures derived from these clusters are more accurate than those of six alternative algorithms in ground truth comparisons.

Integration and transfer learning of single-cell transcriptomes via cFIT

Large, comprehensive collections of single-cell RNA sequencing (scRNA-seq) datasets have been generated that allow for the full transcriptional characterization of cell types across a wide variety of biological and clinical conditions. As new methods arise to measure distinct cellular modalities, a key analytical challenge is to integrate these datasets or transfer knowledge from one to the other to better understand cellular identity and functions. Here, we present a simple yet surprisingly effective method named common factor integration and transfer learning (cFIT) for capturing various batch effects across experiments, technologies, subjects, and even species. The proposed method models the shared information between various datasets by a common factor space while allowing for unique distortions and shifts in genewise expression in each batch. The model parameters are learned under an iterative nonnegative matrix factorization (NMF) framework and then used for synchronized integration from across-domain assays. In addition, the model enables transferring via low-rank matrix from more informative data to allow for precise identification in data of lower quality. Compared with existing approaches, our method imposes weaker assumptions on the cell composition of each individual dataset; however, it is shown to be more reliable in preserving biological variations. We apply cFIT to multiple scRNA-seq datasets of developing brain from human and mouse, varying by technologies and developmental stages. The successful integration and transfer uncover the transcriptional resemblance across systems. The study helps establish a comprehensive landscape of brain cell-type diversity and provides insights into brain development.

Short-term exposure to intermittent hypoxia leads to changes in gene expression seen in chronic pulmonary disease

Obstructive sleep apnea (OSA) results from episodes of airway collapse and intermittent hypoxia (IH) and is associated with a host of health complications. Although the lung is the first organ to sense changes in oxygen levels, little is known about the consequences of IH to the lung hypoxia-inducible factor-responsive pathways. We hypothesized that exposure to IH would lead to cell-specific up- and downregulation of diverse expression pathways. We identified changes in circadian and immune pathways in lungs from mice exposed to IH. Among all cell types, endothelial cells showed the most prominent transcriptional changes. Upregulated genes in myofibroblast cells were enriched for genes associated with pulmonary hypertension and included targets of several drugs currently used to treat chronic pulmonary diseases. A better understanding of the pathophysiologic mechanisms underlying diseases associated with OSA could improve our therapeutic approaches, directing therapies to the most relevant cells and molecular pathways.

In situ mapping identifies distinct vascular niches for myelopoiesis

In contrast to nearly all other tissues, the anatomy of cell differentiation in the bone marrow remains unknown. This is owing to a lack of strategies for examining myelopoiesis-the differentiation of myeloid progenitors into a large variety of innate immune cells-in situ in the bone marrow. Such strategies are required to understand differentiation and lineage-commitment decisions, and to define how spatial organizing cues inform tissue function. Here we develop approaches for imaging myelopoiesis in mice, and generate atlases showing the differentiation of granulocytes, monocytes and dendritic cells. The generation of granulocytes and dendritic cells-monocytes localizes to different blood-vessel structures known as sinusoids, and displays lineage-specific spatial and clonal architectures. Acute systemic infection with Listeria monocytogenes induces lineage-specific progenitor clusters to undergo increased self-renewal of progenitors, but the different lineages remain spatially separated. Monocyte-dendritic cell progenitors (MDPs) map with nonclassical monocytes and conventional dendritic cells; these localize to a subset of blood vessels expressing a major regulator of myelopoiesis, colony-stimulating factor 1 (CSF1, also known as M-CSF)1. Specific deletion of Csf1 in endothelium disrupts the architecture around MDPs and their localization to sinusoids. Subsequently, there are fewer MDPs and their ability to differentiate is reduced, leading to a loss of nonclassical monocytes and dendritic cells during both homeostasis and infection. These data indicate that local cues produced by distinct blood vessels are responsible for the spatial organization of definitive blood cell differentiation.

Single-Cell RNA Sequencing to Disentangle the Blood System

The blood system is often represented as a tree-like structure with stem cells that give rise to mature blood cell types through a series of demarcated steps. Although this representation has served as a model of hierarchical tissue organization for decades, single-cell technologies are shedding new light on the abundance of cell type intermediates and the molecular mechanisms that ensure balanced replenishment of differentiated cells. In this Brief Review, we exemplify new insights into blood cell differentiation generated by single-cell RNA sequencing, summarize considerations for the application of this technology, and highlight innovations that are leading the way to understand hematopoiesis at the resolution of single cells. Graphic Abstract: A graphic abstract is available for this article.

Single-Cell Transcriptome Analysis in Plants: Advances and Challenges

The rapid and enthusiastic adoption of single-cell RNA sequencing (scRNA-seq) has demonstrated that this technology is far more than just another way to perform transcriptome analysis. It is not an exaggeration to say that the advent of scRNA-seq is revolutionizing the details of whole-transcriptome snapshots from a tissue to a cell. With this disruptive technology, it is now possible to mine heterogeneity between tissue types and within cells like never before. This enables more rapid identification of rare and novel cell types, simultaneous characterization of multiple different cell types and states, more accurate and integrated understanding of their roles in life processes, and more. However, we are only at the beginning of unlocking the full potential of scRNA-seq applications. This is particularly true for plant sciences, where single-cell transcriptome profiling is in its early stage and has many exciting challenges to overcome. In this review, we compare and evaluate recent pioneering studies using the Arabidopsis root model, which has established new paradigms for scRNA-seq studies in plants. We also explore several new and promising single-cell analysis tools that are available to those wishing to study plant development and physiology at unprecedented resolution and scale. In addition, we propose some future directions on the use of scRNA-seq technology to tackle some of the critical challenges in plant research and breeding.

Resolving single-cell heterogeneity from hundreds of thousands of cells through sequential hybrid clustering and NMF

Motivation

The rapid proliferation of single-cell RNA-sequencing (scRNA-Seq) technologies has spurred the development of diverse computational approaches to detect transcriptionally coherent populations. While the complexity of the algorithms for detecting heterogeneity has increased, most require significant user-tuning, are heavily reliant on dimension reduction techniques and are not scalable to ultra-large datasets. We previously described a multi-step algorithm, Iterative Clustering and Guide-gene Selection (ICGS), which applies intra-gene correlation and hybrid clustering to uniquely resolve novel transcriptionally coherent cell populations from an intuitive graphical user interface.

Results

We describe a new iteration of ICGS that outperforms state-of-the-art scRNA-Seq detection workflows when applied to well-established benchmarks. This approach combines multiple complementary subtype detection methods (HOPACH, sparse non-negative matrix factorization, cluster ‘fitness’, support vector machine) to resolve rare and common cell-states, while minimizing differences due to donor or batch effects. Using data from multiple cell atlases, we show that the PageRank algorithm effectively downsamples ultra-large scRNA-Seq datasets, without losing extremely rare or transcriptionally similar yet distinct cell types and while recovering novel transcriptionally distinct cell populations. We believe this new approach holds tremendous promise in reproducibly resolving hidden cell populations in complex datasets.

Availability and implementation

ICGS2 is implemented in Python. The source code and documentation are available at http://altanalyze.org.

Supplementary information

Supplementary data are available at Bioinformatics online.

Recapturing embryonic potential in the adult epicardium: Prospects for cardiac repair

Research into potential targets for cardiac repair encompasses recognition of tissue‐resident cells with intrinsic regenerative properties. The adult vertebrate heart is covered by mesothelium, named the epicardium, which becomes active in response to injury and contributes to repair, albeit suboptimally. Motivation to manipulate the epicardium for treatment of myocardial infarction is deeply rooted in its central role in cardiac formation and vasculogenesis during development. Moreover, the epicardium is vital to cardiac muscle regeneration in lower vertebrate and neonatal mammalian‐injured hearts. In this review, we discuss our current understanding of the biology of the mammalian epicardium in development and injury. Considering present challenges in the field, we further contemplate prospects for reinstating full embryonic potential in the adult epicardium to facilitate cardiac regeneration.

Single-Cell Analysis of Neonatal HSC Ontogeny Reveals Gradual and Uncoordinated Transcriptional Reprogramming that Begins before Birth

Fetal and adult hematopoietic stem cells (HSCs) have distinct proliferation rates, lineage biases, gene expression profiles, and gene dependencies. Although these differences are widely recognized, it is not clear how the transition from fetal to adult identity is coordinated. Here we show that murine HSCs and committed hematopoietic progenitor cells (HPCs) undergo a gradual, rather than precipitous, transition from fetal to adult transcriptional states. The transition begins prior to birth and is punctuated by a late prenatal spike in type I interferon signaling that promotes perinatal HPC expansion and sensitizes progenitors to the leukemogenic FLT3ITD mutation. Most other changes in gene expression and enhancer activation are imprecisely timed and poorly coordinated. Thus, heterochronic enhancer elements, and their associated transcripts, are activated independently of one another rather than as part of a robust network. This simplifies the regulatory programs that guide neonatal HSC/HPC ontogeny, but it creates heterogeneity within these populations.

Computational methods for the integrative analysis of single-cell data

Recent advances in single-cell technologies are providing exciting opportunities for dissecting tissue heterogeneity and investigating cell identity, fate and function. This is a pristine, exploding field that is flooding biologists with a new wave of data, each with its own specificities in terms of complexity and information content. The integrative analysis of genomic data, collected at different molecular layers from diverse cell populations, holds promise to address the full-scale complexity of biological systems. However, the combination of different single-cell genomic signals is computationally challenging, as these data are intrinsically heterogeneous for experimental, technical and biological reasons. Here, we describe the computational methods for the integrative analysis of single-cell genomic data, with a focus on the integration of single-cell RNA sequencing datasets and on the joint analysis of multimodal signals from individual cells.

Mouse models of neutropenia reveal progenitor-stage-specific defects

Advances in genetics and sequencing reveal a plethora of disease-associated and disease-causing genetic alterations. Resolving causality between genetics and disease requires generating accurate models for molecular dissection; however, the rapid expansion of single-cell landscapes presents a major challenge to accurate comparisons between mutants and their wild-type equivalents. Here, we generated mouse models of human severe congenital neutropenia (SCN) using patient-derived mutations in the Growth factor independent-1 (GFI1) transcription factor. To delineate the impact of SCN mutations, we generated single-cell references for granulopoietic genomic states with linked epitopes¹, aligned mutant cells to their wild-type equivalent and identified differentially expressed genes and epigenetic loci. We find that Gfi1-target genes are altered sequentially, as cells traverse successive states during differentiation. These cell-state-specific insights facilitated genetic rescue of granulocytic specification but not post-commitment defects in innate-immune effector function; underscoring the importance of evaluating the impact of mutations and therapy within each relevant cell state.

Pbx4 limits heart size and fosters arch artery formation by partitioning second heart field progenitors and restricting proliferation

Vertebrate heart development requires the integration of temporally distinct differentiating progenitors. However, few signals are understood that restrict the size of the later-differentiating outflow tract (OFT). We show that improper specification and proliferation of second heart field (SHF) progenitors in zebrafish lazarus (lzr) mutants, which lack the transcription factor Pbx4, produces enlarged hearts owing to an increase in ventricular and smooth muscle cells. Specifically, Pbx4 initially promotes the partitioning of the SHF into anterior progenitors, which contribute to the OFT, and adjacent endothelial cell progenitors, which contribute to posterior pharyngeal arches. Subsequently, Pbx4 limits SHF progenitor (SHFP) proliferation. Single cell RNA sequencing of nkx2.5+ cells revealed previously unappreciated distinct differentiation states and progenitor subpopulations that normally reside within the SHF and arterial pole of the heart. Specifically, the transcriptional profiles of Pbx4-deficient nkx2.5+ SHFPs are less distinct and display characteristics of normally discrete proliferative progenitor and anterior, differentiated cardiomyocyte populations. Therefore, our data indicate that the generation of proper OFT size and arch arteries requires Pbx-dependent stratification of unique differentiation states to facilitate both homeotic-like transformations and limit progenitor production within the SHF.

Eleven grand challenges in single-cell data science

The recent boom in microfluidics and combinatorial indexing strategies, combined with low sequencing costs, has empowered single-cell sequencing technology. Thousands-or even millions-of cells analyzed in a single experiment amount to a data revolution in single-cell biology and pose unique data science problems. Here, we outline eleven challenges that will be central to bringing this emerging field of single-cell data science forward. For each challenge, we highlight motivating research questions, review prior work, and formulate open problems. This compendium is for established researchers, newcomers, and students alike, highlighting interesting and rewarding problems for the coming years.

Genomic programming of IRF4-expressing human Langerhans cells

Langerhans cells (LC) can prime tolerogenic as well as immunogenic responses in skin, but the genomic states and transcription factors (TF) regulating these context-specific responses are unclear. Bulk and single-cell transcriptional profiling demonstrates that human migratory LCs are robustly programmed for MHC-I and MHC-II antigen presentation. Chromatin analysis reveals enrichment of ETS-IRF and AP1-IRF composite regulatory elements in antigen-presentation genes, coinciding with expression of the TFs, PU.1, IRF4 and BATF3 but not IRF8. Migration of LCs from the epidermis is accompanied by upregulation of IRF4, antigen processing components and co-stimulatory molecules. TNF stimulation augments LC cross-presentation while attenuating IRF4 expression. CRISPR-mediated editing reveals IRF4 to positively regulate the LC activation programme, but repress NF2EL2 and NF-kB pathway genes that promote responsiveness to oxidative stress and inflammatory cytokines. Thus, IRF4-dependent genomic programming of human migratory LCs appears to enable LC maturation while attenuating excessive inflammatory and immunogenic responses in the epidermis.

Langerhans cells (LC) can prime tolerogenic as well as immunogenic responses in the skin. Here the authors show, by transcriptomic, epigenetic and CRISPR editing analyses, that during LC migration and maturation the transcription factor IRF4 regulates expression of antigen presentation and co-stimulatory gene modules while attenuating inflammatory response genes.

DoubletDecon: Deconvoluting Doublets from Single-Cell RNA-Sequencing Data

Methods for single-cell RNA sequencing (scRNA-seq) have greatly advanced in recent years. While droplet- and well-based methods have increased the capture frequency of cells for scRNA-seq, these technologies readily produce technical artifacts, such as doublet cell captures. Doublets occurring between distinct cell types can appear as hybrid scRNA-seq profiles, but do not have distinct transcriptomes from individual cell states. We introduce DoubletDecon, an approach that detects doublets with a combination of deconvolution analyses and the identification of unique cell-state gene expression. We demonstrate the ability of DoubletDecon to identify synthetic, mixed-species, genetic, and cell-hashing cell doublets from scRNA-seq datasets of varying cellular complexity with a high sensitivity relative to alternative approaches. Importantly, this algorithm prevents the prediction of valid mixed-lineage and transitional cell states as doublets by considering their unique gene expression. DoubletDecon has an easy-to-use graphical user interface and is compatible with diverse species and unsupervised population detection algorithms.

LAmbDA: label ambiguous domain adaptation dataset integration reduces batch effects and improves subtype detection

MOTIVATION

Rapid advances in single cell RNA sequencing (scRNA-seq) have produced higher-resolution cellular subtypes in multiple tissues and species. Methods are increasingly needed across datasets and species to (i) remove systematic biases, (ii) model multiple datasets with ambiguous labels and (iii) classify cells and map cell type labels. However, most methods only address one of these problems on broad cell types or simulated data using a single model type. It is also important to address higher-resolution cellular subtypes, subtype labels from multiple datasets, models trained on multiple datasets simultaneously and generalizability beyond a single model type.

RESULTS

We developed a species- and dataset-independent transfer learning framework (LAmbDA) to train models on multiple datasets (even from different species) and applied our framework on simulated, pancreas and brain scRNA-seq experiments. These models mapped corresponding cell types between datasets with inconsistent cell subtype labels while simultaneously reducing batch effects. We achieved high accuracy in labeling cellular subtypes (weighted accuracy simulated 1 datasets: 90%; simulated 2 datasets: 94%; pancreas datasets: 88% and brain datasets: 66%) using LAmbDA Feedforward 1 Layer Neural Network with bagging. This method achieved higher weighted accuracy in labeling cellular subtypes than two other state-of-the-art methods, scmap and CaSTLe in brain (66% versus 60% and 32%). Furthermore, it achieved better performance in correctly predicting ambiguous cellular subtype labels across datasets in 88% of test cases compared with CaSTLe (63%), scmap (50%) and MetaNeighbor (50%). LAmbDA is model- and dataset-independent and generalizable to diverse data types representing an advance in biocomputing.

AVAILABILITY AND IMPLEMENTATION

github.com/tsteelejohnson91/LAmbDA.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Aging Human Hematopoietic Stem Cells Manifest Profound Epigenetic Reprogramming of Enhancers That May Predispose to Leukemia

Aging is associated with functional decline of hematopoietic stem cells (HSC) as well as an increased risk of myeloid malignancies. We performed an integrative characterization of epigenomic and transcriptomic changes, including single-cell RNA sequencing, during normal human aging. Lineage^-CD34⁺CD38^- cells [HSC-enriched (HSCe)] undergo age-associated epigenetic reprogramming consisting of redistribution of DNA methylation and reductions in H3K27ac, H3K4me1, and H3K4me3. This reprogramming of aged HSCe globally targets developmental and cancer pathways that are comparably altered in acute myeloid leukemia (AML) of all ages, encompassing loss of 4,646 active enhancers, 3,091 bivalent promoters, and deregulation of several epigenetic modifiers and key hematopoietic transcription factors, such as KLF6, BCL6, and RUNX3. Notably, in vitro downregulation of KLF6 results in impaired differentiation, increased colony-forming potential, and changes in expression that recapitulate aging and leukemia signatures. Thus, age-associated epigenetic reprogramming may form a predisposing condition for the development of age-related AML. SIGNIFICANCE: AML, which is more frequent in the elderly, is characterized by epigenetic deregulation. We demonstrate that epigenetic reprogramming of human HSCs occurs with age, affecting cancer and developmental pathways. Downregulation of genes epigenetically altered with age leads to impairment in differentiation and partially recapitulates aging phenotypes.This article is highlighted in the In This Issue feature, p. 983.