LineageVAE: Reconstructing Historical Cell States and Transcriptomes toward Unobserved Progenitors.

MOTIVATION: Single-cell RNA sequencing (scRNA-seq) enables comprehensive characterization of the cell state. However, its destructive nature prohibits measuring gene expression changes during dynamic processes such as embryogenesis. Although recent studies integrating scRNA-seq with lineage tracing have provided clonal insights between progenitor and mature cells, challenges remain. Because of their experimental nature, observations are sparse, and cells observed in the early state are not the exact progenitors of cells observed at later time points. To overcome these limitations, we developed LineageVAE, a novel computational methodology that utilizes deep learning based on the property that cells sharing barcodes have identical progenitors. RESULTS: LineageVAE is a deep generative model that transforms scRNA-seq observations with identical lineage barcodes into sequential trajectories toward a common progenitor in a latent cell state space. This method enables the reconstruction of unobservable cell state transitions, historical transcriptomes, and regulatory dynamics at a single-cell resolution. Applied to hematopoiesis and reprogrammed fibroblast datasets, LineageVAE demonstrated its ability to restore backward cell state transitions and infer progenitor heterogeneity and transcription factor activity along differentiation trajectories. AVAILABILITY AND IMPLEMENTATION: The LineageVAE model was implemented in Python using the PyTorch deep learning library. The code is available on GitHub at https://github.com/LzrRacer/LineageVAE/. SUPPLEMENTARY INFORMATION: Available at Bioinformatics online.

Identifying genetic regulatory variants that affect transcription factor activity.

Genetic variants affecting gene expression levels in humans have been mapped in the Genotype-Tissue Expression (GTEx) project. Trans-acting variants impacting many genes simultaneously through a shared transcription factor (TF) are of particular interest. Here, we developed a generalized linear model (GLM) to estimate protein-level TF activity levels in an individual-specific manner from GTEx RNA sequencing (RNA-seq) profiles. It uses observed differential gene expression after TF perturbation as a predictor and, by analyzing differential expression within pairs of neighboring genes, controls for the confounding effect of variation in chromatin state along the genome. We inferred genotype-specific activities for 55 TFs across 49 tissues. Subsequently performing genome-wide association analysis on this virtual trait revealed TF activity quantitative trait loci (aQTLs) that, as a set, are enriched for functional features. Altogether, the set of tools we introduce here highlights the potential of genetic association studies for cellular endophenotypes based on a network-based multi-omics approach. The transparent peer review record is available.

Computational tools for inferring transcription factor activity.

Transcription factors (TFs) are essential players in orchestrating the regulatory landscape in cells. Still, their exact modes of action and dependencies on other regulatory aspects remain elusive. Since TFs act cell type-specific and each TF has its own characteristics, untangling their regulatory interactions from an experimental point of view is laborious and convoluted. Thus, there is an ongoing development of computational tools that estimate transcription factor activity (TFA) from a variety of data modalities, either based on a mapping of TFs to their putative target genes or in a genome-wide, gene-unspecific fashion. These tools can help to gain insights into TF regulation and to prioritize candidates for experimental validation. We want to give an overview of available computational tools that estimate TFA, illustrate examples of their application, debate common result validation strategies, and discuss assumptions and concomitant limitations.

Linking Expression of Cell-Surface Receptors with Transcription Factors by Computational Analysis of Paired Single-Cell Proteomes and Transcriptomes.

Complex signaling and transcriptional programs control the development and physiology of specialized cell types. Genetic perturbations in these programs cause human cancers to arise from a diverse set of specialized cell types and developmental states. Understanding these complex systems and their potential to drive cancer is critical for the development of immunotherapies and druggable targets. Pioneering single-cell multi-omics technologies that analyze transcriptional states have been coupled with the expression of cell-surface receptors. This chapter describes SPaRTAN (Single-cell Proteomic and RNA-based Transcription factor Activity Network), a computational framework, to link transcription factors with cell-surface protein expression. SPaRTAN uses CITE-seq (cellular indexing of transcriptomes and epitopes by sequencing) data and cis-regulatory sites to model the effect of interactions between transcription factors and cell-surface receptors on gene expression. We demonstrate the pipeline for SPaRTAN using CITE-seq data from peripheral blood mononuclear cells.

Dynamic activity in cis-regulatory elements of leukocytes identifies transcription factor activation and stratifies COVID-19 severity in ICU patients.

Transcription factor programs mediating the immune response to coronavirus disease 2019 (COVID-19) are not fully understood. Capturing active transcription initiation from cis-regulatory elements such as enhancers and promoters by capped small RNA sequencing (csRNA-seq), in contrast to capturing steady-state transcripts by conventional RNA-seq, allows unbiased identification of the underlying transcription factor activity and regulatory pathways. Here, we profile transcription initiation in critically ill COVID-19 patients, identifying transcription factor motifs that correlate with clinical lung injury and disease severity. Unbiased clustering reveals distinct subsets of cis-regulatory elements that delineate the cell type, pathway-specific, and combinatorial transcription factor activity. We find evidence of critical roles of regulatory networks, showing that STAT/BCL6 and E2F/MYB regulatory programs from myeloid cell populations are activated in patients with poor disease outcomes and associated with COVID-19 susceptibility genetic variants. More broadly, we demonstrate how capturing acute, disease-mediated changes in transcription initiation can provide insight into the underlying molecular mechanisms and stratify patient disease severity.

scATACpipe: A nextflow pipeline for comprehensive and reproducible analyses of single cell ATAC-seq data.

Single cell ATAC-seq (scATAC-seq) has become the most widely used method for profiling open chromatin landscape of heterogeneous cell populations at a single-cell resolution. Although numerous software tools and pipelines have been developed, an easy-to-use, scalable, reproducible, and comprehensive pipeline for scATAC-seq data analyses is still lacking. To fill this gap, we developed scATACpipe, a Nextflow pipeline, for performing comprehensive analyses of scATAC-seq data including extensive quality assessment, preprocessing, dimension reduction, clustering, peak calling, differential accessibility inference, integration with scRNA-seq data, transcription factor activity and footprinting analysis, co-accessibility inference, and cell trajectory prediction. scATACpipe enables users to perform the end-to-end analysis of scATAC-seq data with three sub-workflow options for preprocessing that leverage 10x Genomics Cell Ranger ATAC software, the ultra-fast Chromap procedures, and a set of custom scripts implementing current best practices for scATAC-seq data preprocessing. The pipeline extends the R package ArchR for downstream analysis with added support to any eukaryotic species with an annotated reference genome. Importantly, scATACpipe generates an all-in-one HTML report for the entire analysis and outputs cluster-specific BAM, BED, and BigWig files for visualization in a genome browser. scATACpipe eliminates the need for users to chain different tools together and facilitates reproducible and comprehensive analyses of scATAC-seq data from raw reads to various biological insights with minimal changes of configuration settings for different computing environments or species. By applying it to public datasets, we illustrated the utility, flexibility, versatility, and reliability of our pipeline, and demonstrated that our scATACpipe outperforms other workflows.

Protoplast for Gene Functional Analysis in Arabidopsis.

Protoplast, a plant cell without cell wall, can be readily transfected by exogenous macromolecules (DNA, RNA, protein) and therefore offer a versatile single cell-based functional analysis system to rapidly assess these exogenous macromolecules' functions. Properly prepared Arabidopsis leaf mesophyll protoplasts exhibit similar responses as intact plants to diverse abiotic and biotic stress signals as well as different hormones and nutrients, based on well-established reporter and marker gene assays. The protoplast transient expression system has been proven to be a vital and reliable tool for elucidation of the activities of transcription factors and protein kinases, protein subcellular localization and trafficking, protein-protein interaction, and protein stabilities in various signal transduction pathways. Moreover, protoplast also offers a platform for single cell-based plant regeneration, gene silencing, and genome editing. Healthy protoplasts isolated from plant tissues and the high transfection efficiency are key steps for successful use of the protoplast transient expression system. In this chapter, we describe the detailed methods of the protoplast transient expression system in Arabidopsis, including plant material preparation, high-quality maxi-plasmid DNA extraction, non-stressed protoplast isolation, highly efficient PEG-calcium transfection of plasmid DNA, and protoplast culture and harvest. We also provide several examples of gene functional analysis using this protoplast transient expression system.

Transcription Factor Activity Inference in Systemic Lupus Erythematosus.

BACKGROUND: Systemic Lupus Erythematosus (SLE) is a systemic autoimmune disease with diverse clinical manifestations. Although most of the SLE-associated loci are located in regulatory regions, there is a lack of global information about transcription factor (TFs) activities, the mode of regulation of the TFs, or the cell or sample-specific regulatory circuits. The aim of this work is to decipher TFs implicated in SLE. METHODS: In order to decipher regulatory mechanisms in SLE, we have inferred TF activities from transcriptomic data for almost all human TFs, defined clusters of SLE patients based on the estimated TF activities and analyzed the differential activity patterns among SLE and healthy samples in two different cohorts. The Transcription Factor activity matrix was used to stratify SLE patients and define sets of TFs with statistically significant differential activity among the disease and control samples. RESULTS: TF activities were able to identify two main subgroups of patients characterized by distinct neutrophil-to-lymphocyte ratio (NLR), with consistent patterns in two independent datasets-one from pediatric patients and other from adults. Furthermore, after contrasting all subgroups of patients and controls, we obtained a significant and robust list of 14 TFs implicated in the dysregulation of SLE by different mechanisms and pathways. Among them, well-known regulators of SLE, such as STAT or IRF, were found, but others suggest new pathways that might have important roles in SLE. CONCLUSIONS: These results provide a foundation to comprehend the regulatory mechanism underlying SLE and the established regulatory factors behind SLE heterogeneity that could be potential therapeutic targets.

Using FRAP to Quantify Changes in Transcription Factor Dynamics After Cell Stimulation: Cell Culture, FRAP, Data Analysis, and Visualization.

Here we show how to measure the mobility of transcription factors using fluorescence recovery after photobleaching (FRAP). Transcription factors are DNA-binding proteins that, upon binding to specific DNA motifs, regulate transcription of their target genes. FRAP is a simple, fast, and cost-effective method, and is a widely used quantitative method to measure the dynamics of fluorescently labeled molecules in solution, membranes, and inside living cells. Dynamics, specified by the immobile fraction, recovery half-time, diffusion constant, and ratio of molecules contributing to different phases of FRAP recovery, can be quantified by FRAP. This can be useful to understand their function in gene regulation. This tutorial is intended to familiarize the reader with the FRAP procedure to quantify transcription factor dynamics using a standard confocal microscope and analysis using MATLAB (MathWorks®). This article will guide the reader through the preconditions of FRAP, and a detailed and step-by-step procedure of preparing cells, bleaching protocol, data analysis in MATLAB, and visualization of the FRAP data.

Transcriptional activity of FIGLA, NEUROG2, and EGR1 transcription factors associated with polymorphisms in the proximal regulatory region of GPR54 gene in cattle.

Activity of transcription factors affect synthesis of G-protein coupled receptor 54 (GPR54), an important factor in regulation of initiation of puberty. Expression of the GPR54 gene in cattle is associated with polymorphisms in the proximal regulatory region (PRR) of the GPR54 gene. Transcription resulting in production of GPR54 mRNA transcript occurs as a result of transcription factor (TF) interactions in the PRR. Polymorphisms in the PRR may be associated with extent of activity of these TFs. Folliculogenesis-specific BHLH TF (FIGLA), neurogenin 2 (NEUROG2), and early growth response 1 (EGR1) are important in modulation of ovarian follicle development and neurons synthesizing GnRH, thus, regulating biosynthesis of luteinizing hormone. The aim of this study, therefore, was to assess the transcription-activating potential of binding sites for FIGLA, NEUROG2, and EGR1 TFs in the GPR54 promoter of cattle. Two luciferase-based promoters, ATC and CCT, which contain three single nucleotide polymorphisms (SNPs), A/C-794, T/C-663, and C/T-601, in the GPR54 PRR, were analyzed to evaluate gene expression and activation of different promoters by FIGLA, NEUROG2, and EGR1. The FIGLA induced GPR54 transcription through the CCT, whereas NEUROG2 and EGR1 induced GPR54 transcription through the ATC promoter-binding site. The CCT-activating effects of FIGLA were greater (2.56-fold) than the ATC-activating effects (P < 0.05). The ATC-activating effects of NEUROG2 and EGR1 were markedly greater (12.91- and 8.41-fold; P < 0.01) than CCT-activating effects. The polymorphisms, CCT and ATC, of the cattle GPR54 affect the activity of transcription factors, therefore, have an important effect on production of GPR54 mRNA transcript.

Transfer of regulatory knowledge from human to mouse for functional genomics analysis.

Transcriptome profiling followed by differential gene expression analysis often leads to lists of genes that are hard to analyze and interpret. Functional genomics tools are powerful approaches for downstream analysis, as they summarize the large and noisy gene expression space into a smaller number of biological meaningful features. In particular, methods that estimate the activity of processes by mapping transcripts level to process members are popular. However, footprints of either a pathway or transcription factor (TF) on gene expression show superior performance over mapping-based gene sets. These footprints are largely developed for humans and their usability in the broadly-used model organism Mus musculus is uncertain. Evolutionary conservation of the gene regulatory system suggests that footprints of human pathways and TFs can functionally characterize mice data. In this paper we analyze this hypothesis. We perform a comprehensive benchmark study exploiting two state-of-the-art footprint methods, DoRothEA and an extended version of PROGENy. These methods infer TF and pathway activity, respectively. Our results show that both can recover mouse perturbations, confirming our hypothesis that footprints are conserved between mice and humans. Subsequently, we illustrate the usability of PROGENy and DoRothEA by recovering pathway/TF-disease associations from newly generated disease sets. Additionally, we provide pathway and TF activity scores for a large collection of human and mouse perturbation and disease experiments (2374). We believe that this resource, available for interactive exploration and download (https://saezlab.shinyapps.io/footprint_scores/), can have broad applications including the study of diseases and therapeutics.

Chromatin Immunoprecipitation Assay for Analyzing Transcription Factor Activity at the Level of Peripheral Myelin Gene Promoters.

Disruption of epigenetic regulators of transcription is a central mechanism of oncogenesis. Differential gene expression is facilitated by transcriptional regulatory mechanisms and chromatin modifications through DNA-protein interactions. One of the widely used assays to study this is chromatin immunoprecipitation (ChIP) assay, which enables the analysis of association between regulatory molecules, specific promoters, and histone modifications within the context of the cell. This is of immense value as ChIP assays can provide a glimpse of the regulatory mechanisms involved in gene expression in vivo. It is also a powerful technique for analyzing histone modifications as well as binding sites for proteins that bind either directly or indirectly to DNA. The basic steps in this protocol are fixation, sonication, immunoprecipitation, and analysis of the immunoprecipitated DNA. Although ChIP is a versatile tool, this procedure requires the optimization of the various reaction conditions. Here, we present a detailed application of the ChIP assay to study the differential recruitment of transcriptional factors at the level of peripheral myelin gene promoters.

Network-based approaches that exploit inferred transcription factor activity to analyze the impact of genetic variation on gene expression.

Over the past decade, a number of methods have emerged for inferring protein-level transcription factor activities in individual samples based on prior information about the structure of the gene regulatory network. We discuss how this has enabled new methods for dissecting trans-acting mechanisms that underpin genetic variation in gene expression.

Environment Tunes Propagation of Cell-to-Cell Variation in the Human Macrophage Gene Network.

Cell-to-cell variation in gene expression and the propagation of such variation (PoV or "noise propagation") from one gene to another in the gene network, as reflected by gene-gene correlation across single cells, are commonly observed in single-cell transcriptomic studies and can shape the phenotypic diversity of cell populations. While gene network "rewiring" is known to accompany cellular adaptation to different environments, how PoV changes between environments and its underlying regulatory mechanisms are less understood. Here, we systematically explored context-dependent PoV among genes in human macrophages, utilizing different cytokines as natural perturbations of multiple molecular parameters that may influence PoV. Our single-cell, epigenomic, computational, and stochastic simulation analyses reveal that environmental adaptation can tune PoV to potentially shape cellular heterogeneity by changing parameters such as the degree of phosphorylation and transcription factor-chromatin interactions. This quantitative tuning of PoV may be a widespread, yet underexplored, property of cellular adaptation to distinct environments.

Past Roadblocks and New Opportunities in Transcription Factor Network Mapping.

One of the principal mechanisms by which cells differentiate and respond to changes in external signals or conditions is by changing the activity levels of transcription factors (TFs). This changes the transcription rates of target genes via the cell's TF network, which ultimately contributes to reconfiguring cellular state. Since microarrays provided our first window into global cellular state, computational biologists have eagerly attacked the problem of mapping TF networks, a key part of the cell's control circuitry. In retrospect, however, steady-state mRNA abundance levels were a poor substitute for TF activity levels and gene transcription rates. Likewise, mapping TF binding through chromatin immunoprecipitation proved less predictive of functional regulation and less amenable to systematic elucidation of complete networks than originally hoped. This review explains these roadblocks and the current, unprecedented blossoming of new experimental techniques built on second-generation sequencing, which hold out the promise of rapid progress in TF network mapping.

Dynamic transcription factor activity profiles reveal key regulatory interactions during megakaryocytic and erythroid differentiation.

The directed differentiation toward erythroid (E) or megakaryocytic (MK) lineages by the MK-E progenitor (MEP) could enhance the ex vivo generation of red blood cells and platelets for therapeutic transfusions. The lineage choice at the MEP bifurcation is controlled in large part by activity within the intracellular signal transduction network, the output of which determines the activity of transcription factors (TFs) and ultimately gene expression. Although many TFs have been implicated, E or MK differentiation is a complex process requiring multiple days, and the dynamics of TF activities during commitment and terminal maturation are relatively unexplored. Herein, we applied a living cell array for the large-scale, dynamic quantification of TF activities during MEP bifurcation. A panel of hematopoietic TFs (GATA-1, GATA-2, SCL/TAL1, FLI-1, NF-E2, PU.1, c-Myb) was characterized during E and MK differentiation of bipotent K562 cells. Dynamic TF activity profiles associated with differentiation towards each lineage were identified, and validated with previous reports. From these activity profiles, we show that GATA-1 is an important hub during early hemin- and PMA-induced differentiation, and reveal several characteristic TF interactions for E and MK differentiation that confirm regulatory mechanisms documented in the literature. Additionally, we highlight several novel TF interactions at various stages of E and MK differentiation. Furthermore, we investigated the mechanism by which nicotinamide (NIC) promoted terminal MK maturation using an MK-committed cell line, CHRF-288-11 (CHRF). Concomitant with its enhancement of ploidy, NIC strongly enhanced the activity of three TFs with known involvement in terminal MK maturation: FLI-1, NF-E2, and p53. Dynamic profiling of TF activity represents a novel tool to complement traditional assays focused on mRNA and protein expression levels to understand progenitor cell differentiation.

Identifying regulatory mechanisms underlying tumorigenesis using locus expression signature analysis.

Retroviral insertional mutagenesis is a powerful tool for identifying putative cancer genes in mice. To uncover the regulatory mechanisms by which common insertion loci affect downstream processes, we supplemented genotyping data with genome-wide mRNA expression profiling data for 97 tumors induced by retroviral insertional mutagenesis. We developed locus expression signature analysis, an algorithm to construct and interpret the differential gene expression signature associated with each common insertion locus. Comparing locus expression signatures to promoter affinity profiles allowed us to build a detailed map of transcription factors whose protein-level regulatory activity is modulated by a particular locus. We also predicted a large set of drugs that might mitigate the effect of the insertion on tumorigenesis. Taken together, our results demonstrate the potential of a locus-specific signature approach for identifying mammalian regulatory mechanisms in a cancer context.

Retinoic Acid Induced 1, RAI1: A Dosage Sensitive Gene Related to Neurobehavioral Alterations Including Autistic Behavior.

Genomic structural changes, such as gene Copy Number Variations (CNVs) are extremely abundant in the human genome. An enormous effort is currently ongoing to recognize and catalogue human CNVs and their associations with abnormal phenotypic outcomes. Recently, several reports related neuropsychiatric diseases (i.e. autism spectrum disorders, schizophrenia, mental retardation, behavioral problems, epilepsy) with specific CNV. Moreover, for some conditions, both the deletion and duplication of the same genomic segment are related to the phenotype. Syndromes associated with CNVs (microdeletion and microduplication) have long been known to display specific neurobehavioral traits. It is important to note that not every gene is susceptible to gene dosage changes and there are only a few dosage sensitive genes. Smith-Magenis (SMS) and Potocki-Lupski (PTLS) syndromes are associated with a reciprocal microdeletion and microduplication within chromosome 17p11.2. in humans. The dosage sensitive gene responsible for most phenotypes in SMS has been identified: the Retinoic Acid Induced 1 (RAI1). Studies on mouse models and humans suggest that RAI1 is likely the dosage sensitive gene responsible for clinical features in PTLS. In addition, the human RAI1 gene has been implicated in several neurobehavioral traits as spinocerebellar ataxia (SCA2), schizophrenia and non syndromic autism. In this review we discuss the evidence of RAI1 as a dosage sensitive gene, its relationship with different neurobehavioral traits, gene structure and mutations, and what is known about its molecular and cellular function, as a first step in the elucidation of the mechanisms that relate dosage sensitive genes with abnormal neurobehavioral outcomes.