Systematic comparison of sequencing-based spatial transcriptomic methods.

Recent developments of sequencing-based spatial transcriptomics (sST) have catalyzed important advancements by facilitating transcriptome-scale spatial gene expression measurement. Despite this progress, efforts to comprehensively benchmark different platforms are currently lacking. The extant variability across technologies and datasets poses challenges in formulating standardized evaluation metrics. In this study, we established a collection of reference tissues and regions characterized by well-defined histological architectures, and used them to generate data to compare 11 sST methods. We highlighted molecular diffusion as a variable parameter across different methods and tissues, significantly affecting the effective resolutions. Furthermore, we observed that spatial transcriptomic data demonstrate unique attributes beyond merely adding a spatial axis to single-cell data, including an enhanced ability to capture patterned rare cell states along with specific markers, albeit being influenced by multiple factors including sequencing depth and resolution. Our study assists biologists in sST platform selection, and helps foster a consensus on evaluation standards and establish a framework for future benchmarking efforts that can be used as a gold standard for the development and benchmarking of computational tools for spatial transcriptomic analysis.

Slide-tags enables single-nucleus barcoding for multimodal spatial genomics.

Recent technological innovations have enabled the high-throughput quantification of gene expression and epigenetic regulation within individual cells, transforming our understanding of how complex tissues are constructed1-6. However, missing from these measurements is the ability to routinely and easily spatially localize these profiled cells. We developed a strategy, Slide-tags, in which single nuclei within an intact tissue section are tagged with spatial barcode oligonucleotides derived from DNA-barcoded beads with known positions. These tagged nuclei can then be used as an input into a wide variety of single-nucleus profiling assays. Application of Slide-tags to the mouse hippocampus positioned nuclei at less than 10 µm spatial resolution and delivered whole-transcriptome data that are indistinguishable in quality from ordinary single-nucleus RNA-sequencing data. To demonstrate that Slide-tags can be applied to a wide variety of human tissues, we performed the assay on brain, tonsil and melanoma. We revealed cell-type-specific spatially varying gene expression across cortical layers and spatially contextualized receptor-ligand interactions driving B cell maturation in lymphoid tissue. A major benefit of Slide-tags is that it is easily adaptable to almost any single-cell measurement technology. As a proof of principle, we performed multiomic measurements of open chromatin, RNA and T cell receptor (TCR) sequences in the same cells from metastatic melanoma, identifying transcription factor motifs driving cancer cell state transitions in spatially distinct microenvironments. Slide-tags offers a universal platform for importing the compendium of established single-cell measurements into the spatial genomics repertoire.

Benchmarking long-read RNA-sequencing analysis tools using in silico mixtures.

The lack of benchmark data sets with inbuilt ground-truth makes it challenging to compare the performance of existing long-read isoform detection and differential expression analysis workflows. Here, we present a benchmark experiment using two human lung adenocarcinoma cell lines that were each profiled in triplicate together with synthetic, spliced, spike-in RNAs (sequins). Samples were deeply sequenced on both Illumina short-read and Oxford Nanopore Technologies long-read platforms. Alongside the ground-truth available via the sequins, we created in silico mixture samples to allow performance assessment in the absence of true positives or true negatives. Our results show that StringTie2 and bambu outperformed other tools from the six isoform detection tools tested, DESeq2, edgeR and limma-voom were best among the five differential transcript expression tools tested and there was no clear front-runner for performing differential transcript usage analysis between the five tools compared, which suggests further methods development is needed for this application.

Spatiotemporal transcriptomic maps of whole mouse embryos at the onset of organogenesis.

Spatiotemporal orchestration of gene expression is required for proper embryonic development. The use of single-cell technologies has begun to provide improved resolution of early regulatory dynamics, including detailed molecular definitions of most cell states during mouse embryogenesis. Here we used Slide-seq to build spatial transcriptomic maps of complete embryonic day (E) 8.5 and E9.0, and partial E9.5 embryos. To support their utility, we developed sc3D, a tool for reconstructing and exploring three-dimensional 'virtual embryos', which enables the quantitative investigation of regionalized gene expression patterns. Our measurements along the main embryonic axes of the developing neural tube revealed several previously unannotated genes with distinct spatial patterns. We also characterized the conflicting transcriptional identity of 'ectopic' neural tubes that emerge in Tbx6 mutant embryos. Taken together, we present an experimental and computational framework for the spatiotemporal investigation of whole embryonic structures and mutant phenotypes.

A posterior probability based Bayesian method for single-cell RNA-seq data imputation.

Single-cell RNA-sequencing (scRNA-seq) data suffer from a lot of zeros. Such dropout events impede the downstream data analyses. We propose BayesImpute to infer and impute dropouts from the scRNA-seq data. Using the expression rate and coefficient of variation of the genes within the cell subpopulation, BayesImpute first determines likely dropouts, and then constructs the posterior distribution for each gene and uses the posterior mean to impute dropout values. Some simulated and real experiments show that BayesImpute can effectively identify dropout events and reduce the introduction of false positive signals. Additionally, BayesImpute successfully recovers the true expression levels of missing values, restores the gene-to-gene and cell-to-cell correlation coefficient, and maintains the biological information in bulk RNA-seq data. Furthermore, BayesImpute boosts the clustering and visualization of cell subpopulations and improves the identification of differentially expressed genes. We further demonstrate that, in comparison to other statistical-based imputation methods, BayesImpute is scalable and fast with minimal memory usage.

Slide-tags: scalable, single-nucleus barcoding for multi-modal spatial genomics.

Recent technological innovations have enabled the high-throughput quantification of gene expression and epigenetic regulation within individual cells, transforming our understanding of how complex tissues are constructed. Missing from these measurements, however, is the ability to routinely and easily spatially localise these profiled cells. We developed a strategy, Slide-tags, in which single nuclei within an intact tissue section are 'tagged' with spatial barcode oligonucleotides derived from DNA-barcoded beads with known positions. These tagged nuclei can then be used as input into a wide variety of single-nucleus profiling assays. Application of Slide-tags to the mouse hippocampus positioned nuclei at less than 10 micron spatial resolution, and delivered whole-transcriptome data that was indistinguishable in quality from ordinary snRNA-seq. To demonstrate that Slide-tags can be applied to a wide variety of human tissues, we performed the assay on brain, tonsil, and melanoma. We revealed cell-type-specific spatially varying gene expression across cortical layers and spatially contextualised receptor-ligand interactions driving B-cell maturation in lymphoid tissue. A major benefit of Slide-tags is that it is easily adaptable to virtually any single-cell measurement technology. As proof of principle, we performed multiomic measurements of open chromatin, RNA, and T-cell receptor sequences in the same cells from metastatic melanoma. We identified spatially distinct tumour subpopulations to be differentially infiltrated by an expanded T-cell clone and undergoing cell state transition driven by spatially clustered accessible transcription factor motifs. Slide-tags offers a universal platform for importing the compendium of established single-cell measurements into the spatial genomics repertoire.

The expanding vistas of spatial transcriptomics.

The formation and maintenance of tissue integrity requires complex, coordinated activities by thousands of genes and their encoded products. Until recently, transcript levels could only be quantified for a few genes in tissues, but advances in DNA sequencing, oligonucleotide synthesis and fluorescence microscopy have enabled the invention of a suite of spatial transcriptomics technologies capable of measuring the expression of many, or all, genes in situ. These technologies have evolved rapidly in sensitivity, multiplexing and throughput. As such, they have enabled the determination of the cell-type architecture of tissues, the querying of cell-cell interactions and the monitoring of molecular interactions between tissue components. The rapidly evolving spatial genomics landscape will enable generalized high-throughput genomic measurements and perturbations to be performed in the context of tissues. These advances will empower hypothesis generation and biological discovery and bridge the worlds of tissue biology and genomics.

Single-cell multiomics reveal the scale of multilayered adaptations enabling CLL relapse during venetoclax therapy.

Venetoclax (VEN) inhibits the prosurvival protein BCL2 to induce apoptosis and is a standard therapy for chronic lymphocytic leukemia (CLL), delivering high complete remission rates and prolonged progression-free survival in relapsed CLL but with eventual loss of efficacy. A spectrum of subclonal genetic changes associated with VEN resistance has now been described. To fully understand clinical resistance to VEN, we combined single-cell short- and long-read RNA-sequencing to reveal the previously unappreciated scale of genetic and epigenetic changes underpinning acquired VEN resistance. These appear to be multilayered. One layer comprises changes in the BCL2 family of apoptosis regulators, especially the prosurvival family members. This includes previously described mutations in BCL2 and amplification of the MCL1 gene but is heterogeneous across and within individual patient leukemias. Changes in the proapoptotic genes are notably uncommon, except for single cases with subclonal losses of BAX or NOXA. Much more prominent was universal MCL1 gene upregulation. This was driven by an overlying layer of emergent NF-κB (nuclear factor kappa B) activation, which persisted in circulating cells during VEN therapy. We discovered that MCL1 could be a direct transcriptional target of NF-κB. Both the switch to alternative prosurvival factors and NF-κB activation largely dissipate following VEN discontinuation. Our studies reveal the extent of plasticity of CLL cells in their ability to evade VEN-induced apoptosis. Importantly, these findings pinpoint new approaches to circumvent VEN resistance and provide a specific biological justification for the strategy of VEN discontinuation once a maximal response is achieved rather than maintaining long-term selective pressure with the drug.

Benchmarking UMI-based single-cell RNA-seq preprocessing workflows.

BACKGROUND: Single-cell RNA-sequencing (scRNA-seq) technologies and associated analysis methods have rapidly developed in recent years. This includes preprocessing methods, which assign sequencing reads to genes to create count matrices for downstream analysis. While several packaged preprocessing workflows have been developed to provide users with convenient tools for handling this process, how they compare to one another and how they influence downstream analysis have not been well studied. RESULTS: Here, we systematically benchmark the performance of 10 end-to-end preprocessing workflows (Cell Ranger, Optimus, salmon alevin, alevin-fry, kallisto bustools, dropSeqPipe, scPipe, zUMIs, celseq2, and scruff) using datasets yielding different biological complexity levels generated by CEL-Seq2 and 10x Chromium platforms. We compare these workflows in terms of their quantification properties directly and their impact on normalization and clustering by evaluating the performance of different method combinations. While the scRNA-seq preprocessing workflows compared vary in their detection and quantification of genes across datasets, after downstream analysis with performant normalization and clustering methods, almost all combinations produce clustering results that agree well with the known cell type labels that provided the ground truth in our analysis. CONCLUSIONS: In summary, the choice of preprocessing method was found to be less important than other steps in the scRNA-seq analysis process. Our study comprehensively compares common scRNA-seq preprocessing workflows and summarizes their characteristics to guide workflow users.

Comprehensive characterization of single-cell full-length isoforms in human and mouse with long-read sequencing.

A modified Chromium 10x droplet-based protocol that subsamples cells for both short-read and long-read (nanopore) sequencing together with a new computational pipeline (FLAMES) is developed to enable isoform discovery, splicing analysis, and mutation detection in single cells. We identify thousands of unannotated isoforms and find conserved functional modules that are enriched for alternative transcript usage in different cell types and species, including ribosome biogenesis and mRNA splicing. Analysis at the transcript level allows data integration with scATAC-seq on individual promoters, improved correlation with protein expression data, and linked mutations known to confer drug resistance to transcriptome heterogeneity.

Modulation of Vagal Sensory Neurons via High Mobility Group Box-1 and Receptor for Advanced Glycation End Products: Implications for Respiratory Viral Infections.

Vagal sensory neurons contribute to the symptoms and pathogenesis of inflammatory pulmonary diseases through processes that involve changes to their morphological and functional characteristics. The alarmin high mobility group box-1 (HMGB1) is an early mediator of pulmonary inflammation and can have actions on neurons in a range of inflammatory settings. We hypothesized that HMGB1 can regulate the growth and function of vagal sensory neurons and we set out to investigate this and the mechanisms involved. Culturing primary vagal sensory neurons from wildtype mice in the presence of HMGB1 significantly increased neurite outgrowth, while acute application of HMGB1 to isolated neurons under patch clamp electrophysiological investigation produced inward currents and enhanced action potential firing. Transcriptional analyses revealed the expression of the cognate HMGB1 receptors, Receptor for Advanced Glycation End products (RAGE) and Toll-like Receptor 4 (TLR4), in subsets of vagal sensory neurons. HMGB1-evoked growth and electrophysiological responses were significantly reduced in primary vagal sensory neurons harvested from RAGE deficient mice and completely absent in neurons from RAGE/TLR4 double deficient mice. Immunohistochemical analysis of vagal sensory neurons collected from mice after intranasal infection with murine pneumovirus or influenza A virus (IAV), or after intratracheal administration with the viral mimetic PolyI:C, revealed a significant increase in nuclear-to-cytoplasm translocation of HMGB1 compared to mock-inoculated mice. Neurons cultured from virus infected wildtype mice displayed a significant increase in neurite outgrowth, which was not observed for neurons from virus infected RAGE or RAGE/TLR4 deficient mice. These data suggest that HMGB1 can enhance vagal sensory neuron growth and excitability, acting primarily via sensory neuron RAGE. Activation of the HMGB1-RAGE axis in vagal sensory neurons could be an important mechanism leading to vagal hyperinnervation and hypersensitivity in chronic pulmonary disease.

Single-cell immune profiling reveals distinct immune response in asymptomatic COVID-19 patients.

While some individuals infected by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) present mild-to-severe disease, many SARS-CoV-2-infected individuals are asymptomatic. We sought to identify the distinction of immune response between asymptomatic and moderate patients. We performed single-cell transcriptome and T-cell/B-cell receptor (TCR/BCR) sequencing in 37 longitudinal collected peripheral blood mononuclear cell samples from asymptomatic, moderate, and severe patients with healthy controls. Asymptomatic patients displayed increased CD56briCD16- natural killer (NK) cells and upregulation of interferon-gamma in effector CD4+ and CD8+ T cells and NK cells. They showed more robust TCR clonal expansion, especially in effector CD4+ T cells, but lack strong BCR clonal expansion compared to moderate patients. Moreover, asymptomatic patients have lower interferon-stimulated genes (ISGs) expression in general but large interpatient variability, whereas moderate patients showed various magnitude and temporal dynamics of the ISGs expression across multiple cell populations but lower than a patient with severe disease. Our data provide evidence of different immune signatures to SARS-CoV-2 in asymptomatic infections.

The long and the short of it: unlocking nanopore long-read RNA sequencing data with short-read differential expression analysis tools.

Application of Oxford Nanopore Technologies' long-read sequencing platform to transcriptomic analysis is increasing in popularity. However, such analysis can be challenging due to the high sequence error and small library sizes, which decreases quantification accuracy and reduces power for statistical testing. Here, we report the analysis of two nanopore RNA-seq datasets with the goal of obtaining gene- and isoform-level differential expression information. A dataset of synthetic, spliced, spike-in RNAs ('sequins') as well as a mouse neural stem cell dataset from samples with a null mutation of the epigenetic regulator Smchd1 was analysed using a mix of long-read specific tools for preprocessing together with established short-read RNA-seq methods for downstream analysis. We used limma-voom to perform differential gene expression analysis, and the novel FLAMES pipeline to perform isoform identification and quantification, followed by DRIMSeq and limma-diffSplice (with stageR) to perform differential transcript usage analysis. We compared results from the sequins dataset to the ground truth, and results of the mouse dataset to a previous short-read study on equivalent samples. Overall, our work shows that transcriptomic analysis of long-read nanopore data using long-read specific preprocessing methods together with short-read differential expression methods and software that are already in wide use can yield meaningful results.

Clonal multi-omics reveals Bcor as a negative regulator of emergency dendritic cell development.

Despite advances in single-cell multi-omics, a single stem or progenitor cell can only be tested once. We developed clonal multi-omics, in which daughters of a clone act as surrogates of the founder, thereby allowing multiple independent assays per clone. With SIS-seq, clonal siblings in parallel "sister" assays are examined either for gene expression by RNA sequencing (RNA-seq) or for fate in culture. We identified, and then validated using CRISPR, genes that controlled fate bias for different dendritic cell (DC) subtypes. This included Bcor as a suppressor of plasmacytoid DC (pDC) and conventional DC type 2 (cDC2) numbers during Flt3 ligand-mediated emergency DC development. We then developed SIS-skew to examine development of wild-type and Bcor-deficient siblings of the same clone in parallel. We found Bcor restricted clonal expansion, especially for cDC2s, and suppressed clonal fate potential, especially for pDCs. Therefore, SIS-seq and SIS-skew can reveal the molecular and cellular mechanisms governing clonal fate.

Single-cell analyses reveal the clonal and molecular aetiology of Flt3L-induced emergency dendritic cell development.

Regulation of haematopoietic stem and progenitor cell (HSPC) fate is crucial during homeostasis and under stress conditions. Here we examine the aetiology of the Flt3 ligand (Flt3L)-mediated increase of type 1 conventional dendritic cells (cDC1s). Using cellular barcoding we demonstrate this occurs through selective clonal expansion of HSPCs that are primed to produce cDC1s and not through activation of cDC1 fate by other HSPCs. In particular, multi/oligo-potent clones selectively amplify their cDC1 output, without compromising the production of other lineages, via a process we term tuning. We then develop Divi-Seq to simultaneously profile the division history, surface phenotype and transcriptome of individual HSPCs. We discover that Flt3L-responsive HSPCs maintain a proliferative 'early progenitor'-like state, leading to the selective expansion of multiple transitional cDC1-primed progenitor stages that are marked by Irf8 expression. These findings define the mechanistic action of Flt3L through clonal tuning, which has important implications for other models of 'emergency' haematopoiesis.

A new lymphoid-primed progenitor marked by Dach1 downregulation identified with single cell multi-omics.

A classical view of blood cell development is that multipotent hematopoietic stem and progenitor cells (HSPCs) become lineage-restricted at defined stages. Lin-c-Kit+Sca-1+Flt3+ cells, termed lymphoid-primed multipotent progenitors (LMPPs), have lost megakaryocyte and erythroid potential but are heterogeneous in their fate. Here, through single-cell RNA sequencing, we identify the expression of Dach1 and associated genes in this fraction as being coexpressed with myeloid/stem genes but inversely correlated with lymphoid genes. Through generation of Dach1-GFP reporter mice, we identify a transcriptionally and functionally unique Dach1-GFP- subpopulation within LMPPs with lymphoid potential with low to negligible classic myeloid potential. We term these 'lymphoid-primed progenitors' (LPPs). These findings define an early definitive branch point of lymphoid development in hematopoiesis and a means for prospective isolation of LPPs.

Transcriptional Profiling of Individual Airway Projecting Vagal Sensory Neurons.

Bronchopulmonary sensory neurons are derived from the vagal sensory ganglia and are essential for monitoring the physical and chemical environment of the airways and lungs. Subtypes are heterogenous in their responsiveness to stimuli, phenotype, and developmental origin, but they collectively serve to regulate normal respiratory and pulmonary processes and elicit a diverse range of defensive physiological responses that protect against noxious stimuli. In this study, we aimed to investigate the transcriptional features of vagal bronchopulmonary sensory neurons using single-cell RNA sequencing (scRNA-seq) to provide a deeper insight into their molecular profiles. Retrogradely labeled vagal sensory neurons projecting to the airways and lungs were hierarchically clustered into five types reflecting their developmental lineage (neural crest versus placodal) and putative function (nociceptors versus mechanoreceptors). The purinergic receptor subunit P2rx2 is known to display restricted expression in placodal-derived nodose neurons, and we demonstrate that the gene profiles defining cells high and low in expression of P2rx2 include G protein coupled receptors and ion channels, indicative of preferential expression in nodose or jugular neurons. Our results provide valuable insight into the transcriptional characteristics of bronchopulmonary sensory neurons and provide rational targets for future physiological investigations.

CellBench: R/Bioconductor software for comparing single-cell RNA-seq analysis methods.

MOTIVATION: Bioinformatic analysis of single-cell gene expression data is a rapidly evolving field. Hundreds of bespoke methods have been developed in the past few years to deal with various aspects of single-cell analysis and consensus on the most appropriate methods to use under different settings is still emerging. Benchmarking the many methods is therefore of critical importance and since analysis of single-cell data usually involves multi-step pipelines, effective evaluation of pipelines involving different combinations of methods is required. Current benchmarks of single-cell methods are mostly implemented with ad-hoc code that is often difficult to reproduce or extend, and exhaustive manual coding of many combinations is infeasible in most instances. Therefore, new software is needed to manage pipeline benchmarking. RESULTS: The CellBench R software facilitates method comparisons in either a task-centric or combinatorial way to allow pipelines of methods to be evaluated in an effective manner. CellBench automatically runs combinations of methods, provides facilities for measuring running time and delivers output in tabular form which is highly compatible with tidyverse R packages for summary and visualization. Our software has enabled comprehensive benchmarking of single-cell RNA-seq normalization, imputation, clustering, trajectory analysis and data integration methods using various performance metrics obtained from data with available ground truth. CellBench is also amenable to benchmarking other bioinformatics analysis tasks. AVAILABILITY AND IMPLEMENTATION: Available from https://bioconductor.org/packages/CellBench.

Interconversion between Tumorigenic and Differentiated States in Acute Myeloid Leukemia.

Tumors are composed of phenotypically heterogeneous cancer cells that often resemble various differentiation states of their lineage of origin. Within this hierarchy, it is thought that an immature subpopulation of tumor-propagating cancer stem cells (CSCs) differentiates into non-tumorigenic progeny, providing a rationale for therapeutic strategies that specifically eradicate CSCs or induce their differentiation. The clinical success of these approaches depends on CSC differentiation being unidirectional rather than reversible, yet this question remains unresolved even in prototypically hierarchical malignancies, such as acute myeloid leukemia (AML). Here, we show in murine and human models of AML that, upon perturbation of endogenous expression of the lineage-determining transcription factor PU.1 or withdrawal of established differentiation therapies, some mature leukemia cells can de-differentiate and reacquire clonogenic and leukemogenic properties. Our results reveal plasticity of CSC maturation in AML, highlighting the need to therapeutically eradicate cancer cells across a range of differentiation states.