Measuring Signaling and RNA-Seq in the Same Cell Links Gene Expression to Dynamic Patterns of NF-κB Activation

Dynamical and combinatorial coding by MAPK p38 and NFκB in the inflammatory response of macrophages

Macrophages sense pathogens and orchestrate specific immune responses. Stimulus specificity is thought to be achieved through combinatorial and dynamical coding by signaling pathways. While NFκB dynamics are known to encode stimulus information, dynamical coding in other signaling pathways and their combinatorial coordination remain unclear. Here, we established live-cell microscopy to investigate how NFκB and p38 dynamics interface in stimulated macrophages. Information theory and machine learning revealed that p38 dynamics distinguish cytokine TNF from pathogen-associated molecular patterns and high doses from low, but contributed little to information-rich NFκB dynamics when both pathways are considered. This suggests that immune response genes benefit from decoding immune signaling dynamics or combinatorics, but not both. We found that the heterogeneity of the two pathways is surprisingly uncorrelated. Mathematical modeling revealed potential sources of uncorrelated heterogeneity in the branched pathway network topology and predicted it to drive gene expression variability. Indeed, genes dependent on both p38 and NFκB showed high scRNAseq variability and bimodality. These results identify combinatorial signaling as a mechanism to restrict NFκB-AND-p38-responsive inflammatory cytokine expression to few cells.

Co-imaging of RelA and c-Rel reveals features of NF-κB signaling for ligand discrimination

Individual cell sensing of external cues has evolved through the temporal patterns in signaling. Since nuclear factor κB (NF-κB) signaling dynamics have been examined using a single subunit, RelA, it remains unclear whether more information might be transmitted via other subunits. Using NF-κB double-knockin reporter mice, we monitored both canonical NF-κB subunits, RelA and c-Rel, simultaneously in single macrophages by quantitative live-cell imaging. We show that signaling features of RelA and c-Rel convey more information about the stimuli than those of either subunit alone. Machine learning is used to predict the ligand identity accurately based on RelA and c-Rel signaling features without considering the co-activated factors. Ligand discrimination is achieved through selective non-redundancy of RelA and c-Rel signaling dynamics, as well as their temporal coordination. These results suggest a potential role of c-Rel in fine-tuning immune responses and highlight the need for approaches that will elucidate the mechanisms regulating NF-κB subunit specificity.

Open questions in the NF-κB field

A variety of stress signals leads to activation of the inducible transcription factor NF-κB, one of the master regulators of the innate immune response. Despite a wealth of information available on the NF-κB core components and its control by different activation pathways and negative feedback loops, several levels of complexity hamper our understanding of the system. This has also contributed to the limited success of NF-κB inhibitors in the clinic and explains some of their unexpected effects. Here we consider the molecular and cellular events generating this complexity at all levels and point to a number of unresolved questions in the field. We also discuss potential future experimental and computational strategies to provide a deeper understanding of NF-κB and its coregulatory signaling networks.

Predicting gene-level sensitivity to JAK-STAT signaling perturbation using a mechanistic-to-machine learning framework

The Janus kinase (JAK)-signal transducer and activator of transcription (STAT) pathway integrates complex cytokine signals via a limited number of molecular components, inspiring numerous efforts to clarify the diversity and specificity of STAT transcription factor function. We developed a computational framework to make global cytokine-induced gene predictions from STAT phosphorylation dynamics, modeling macrophage responses to interleukin (IL)-6 and IL-10, which signal through common STATs, but with distinct temporal dynamics and contrasting functions. Our mechanistic-to-machine learning model identified cytokine-specific genes associated with late pSTAT3 time frames and a preferential pSTAT1 reduction upon JAK2 inhibition. We predicted and validated the impact of JAK2 inhibition on gene expression, identifying genes that were sensitive or insensitive to JAK2 variation. Thus, we successfully linked STAT signaling dynamics to gene expression to support future efforts targeting pathology-associated STAT-driven gene sets. This serves as a first step in developing multi-level prediction models to understand and perturb gene expression outputs from signaling systems. A record of this paper's transparent peer review process is included in the supplemental information.

Small transcriptional differences among cell clones lead to distinct NF-κB dynamics

Transcription factor dynamics is fundamental to determine the activation of accurate transcriptional programs and yet is heterogeneous at a single-cell level, even within homogeneous populations. We asked how such heterogeneity emerges for the nuclear factor κB (NF-κB). We found that clonal populations of immortalized fibroblasts derived from a single mouse embryo display robustly distinct NF-κB dynamics upon tumor necrosis factor ɑ (TNF-ɑ) stimulation including persistent, oscillatory, and weak activation, giving rise to differences in the transcription of its targets. By combining transcriptomics and simulations we show how less than two-fold differences in the expression levels of genes coding for key proteins of the signaling cascade and feedback system are predictive of the differences of the NF-κB dynamic response of the clones to TNF-ɑ and IL-1β. We propose that small transcriptional differences in the regulatory circuit of a transcription factor can lead to distinct signaling dynamics in cells within homogeneous cell populations and among different cell types.

Bridging live-cell imaging and next-generation cancer treatment

By providing spatial, molecular and morphological data over time, live-cell imaging can provide a deeper understanding of the cellular and signalling events that determine cancer response to treatment. Understanding this dynamic response has the potential to enhance clinical outcome by identifying biomarkers or actionable targets to improve therapeutic efficacy. Here, we review recent applications of live-cell imaging for uncovering both tumour heterogeneity in treatment response and the mode of action of cancer-targeting drugs. Given the increasing uses of T cell therapies, we discuss the unique opportunity of time-lapse imaging for capturing the interactivity and motility of immunotherapies. Although traditionally limited in the number of molecular features captured, novel developments in multidimensional imaging and multi-omics data integration offer strategies to connect single-cell dynamics to molecular phenotypes. We review the effect of these recent technological advances on our understanding of the cellular dynamics of tumour targeting and discuss their implication for next-generation precision medicine.

Quantifying dynamic pro-inflammatory gene expression and heterogeneity in single macrophage cells

Macrophages must respond appropriately to pathogens and other pro-inflammatory stimuli in order to perform their roles in fighting infection. One way in which inflammatory stimuli can vary is in their dynamics-that is, the amplitude and duration of stimulus experienced by the cell. In this study, we performed long-term live cell imaging in a microfluidic device to investigate how the pro-inflammatory genes IRF1, CXCL10, and CXCL9 respond to dynamic interferon-gamma (IFNγ) stimulation. We found that IRF1 responds to low concentration or short duration IFNγ stimulation, whereas CXCL10 and CXCL9 require longer or higherconcentration stimulation to be expressed. We also investigated the heterogeneity in the expression of each gene and found that CXCL10 and CXCL9 have substantial cell-to-cell variability. In particular, the expression of CXCL10 appears to be largely stochastic with a subpopulation of nonresponding cells across all the stimulation conditions tested. We developed both deterministic and stochastic models for the expression of each gene. Our modeling analysis revealed that the heterogeneity in CXCL10 can be attributed to a slow chromatin-opening step that is on a similar timescale to that of adaptation of the upstream signal. In this way, CXCL10 expression in individual cells can remain stochastic in response to each pulse of repeated stimulation, which we also validated by experiments. Together, we conclude that pro-inflammatory genes in the same signaling pathway can respond to dynamic IFNγ stimulus with very different response features and that upstream signal adaptation can contribute to shaping heterogeneous gene expression.

Time-dependent cell-state selection identifies transiently expressed genes regulating ILC2 activation

The decision of whether cells are activated or not is controlled through dynamic intracellular molecular networks. However, the low population of cells during the transition state of activation renders the analysis of the transcriptome of this state technically challenging. To address this issue, we have developed the Time-Dependent Cell-State Selection (TDCSS) technique, which employs live-cell imaging of secretion activity to detect an index of the transition state, followed by the simultaneous recovery of indexed cells for subsequent transcriptome analysis. In this study, we used the TDCSS technique to investigate the transition state of group 2 innate lymphoid cells (ILC2s) activation, which is indexed by the onset of interleukin (IL)-13 secretion. The TDCSS approach allowed us to identify time-dependent genes, including transiently induced genes (TIGs). Our findings of IL4 and MIR155HG as TIGs have shown a regulatory function in ILC2s activation.

Signaling is the pathway to macrophage function

Tissue and inflammatory contexts are well appreciated to shape macrophage function to promote health or disease. However, there has been minimal progress towards understanding how these contexts modify signaling-to-transcription networks. Integration of mechanistic modeling and data-driven approaches will be crucial for investigating how cell state impacts macrophage decision-making.

Predicting gene level sensitivity to JAK-STAT signaling perturbation using a mechanistic-to-machine learning framework

The JAK-STAT pathway integrates complex cytokine signals via a limited number of molecular components, inspiring numerous efforts to clarify the diversity and specificity of STAT transcription factor function. We developed a computational workflow to make global cytokine-induced gene predictions from STAT phosphorylation dynamics, modeling macrophage responses to IL-6 and IL-10, which signal through common STATs, but with distinct temporal dynamics and contrasting functions. Our mechanistic-to-machine learning model identified select cytokine-induced gene sets associated with late pSTAT3 timeframes and a preferential pSTAT1 reduction upon JAK2 inhibition. We predicted and validated the impact of JAK2 inhibition on gene expression, identifying dynamically regulated genes that were sensitive or insensitive to JAK2 variation. Thus, we successfully linked STAT signaling dynamics to gene expression to support future efforts targeting pathology-associated STAT-driven gene sets. This serves as a first step in developing multi-level prediction models to understand and perturb gene expression outputs from signaling systems.

Optogenetic manipulation identifies the roles of ERK and AKT dynamics in controlling mouse embryonic stem cell exit from pluripotency

ERK and AKT signaling control pluripotent cell self-renewal versus differentiation. ERK pathway activity over time (i.e., dynamics) is heterogeneous between individual pluripotent cells, even in response to the same stimuli. To analyze potential functions of ERK and AKT dynamics in controlling mouse embryonic stem cell (ESC) fates, we developed ESC lines and experimental pipelines for the simultaneous long-term manipulation and quantification of ERK or AKT dynamics and cell fates. We show that ERK activity duration or amplitude or the type of ERK dynamics (e.g., transient, sustained, or oscillatory) alone does not influence exit from pluripotency, but the sum of activity over time does. Interestingly, cells retain memory of previous ERK pulses, with duration of memory retention dependent on duration of previous pulse length. FGF receptor/AKT dynamics counteract ERK-induced pluripotency exit. These findings improve our understanding of how cells integrate dynamics from multiple signaling pathways and translate them into cell fate cues.

Embryonic stem cell ERK, AKT, plus STAT3 response dynamics combinatorics are heterogeneous but NANOG state independent

Signaling is central in cell fate regulation, and relevant information is encoded in its activity over time (i.e., dynamics). However, simultaneous dynamics quantification of several pathways in single mammalian stem cells has not yet been accomplished. Here we generate mouse embryonic stem cell (ESC) lines simultaneously expressing fluorescent reporters for ERK, AKT, and STAT3 signaling activity, which all control pluripotency. We quantify their single-cell dynamics combinations in response to different self-renewal stimuli and find striking heterogeneity for all pathways, some dependent on cell cycle but not pluripotency states, even in ESC populations currently assumed to be highly homogeneous. Pathways are mostly independently regulated, but some context-dependent correlations exist. These quantifications reveal surprising single-cell heterogeneity in the important cell fate control layer of signaling dynamics combinations and raise fundamental questions about the role of signaling in (stem) cell fate control.

Benchmarking differential abundance methods for finding condition-specific prototypical cells in multi-sample single-cell datasets

Modern single-cell data analysis relies on statistical testing (e.g. differential expression testing) to identify genes or proteins that are up-or down-regulated in relation to cell-types or clinical outcomes. However, existing algorithms for such statistical testing are often limited by technical noise and cellular heterogeneity, which lead to false-positive results. To constrain the analysis to a compact and phenotype-related cell population, differential abundance (DA) testing methods were employed to identify subgroups of cells whose abundance changed significantly in response to disease progression, or experimental perturbation. Despite the effectiveness of DA testing algorithms of identifying critical cell-states, there are no systematic benchmarking or comparative studies to compare their usages in practice. Herein, we performed the first comprehensive benchmarking study to objectively evaluate and compare the benefits and potential downsides of current state-of-the-art DA testing methods. We benchmarked six DA testing methods on several practical tasks, using both synthetic and real single-cell datasets. The task evaluated include, recognizing true DA subpopulations, appropriate handing of batch effects, runtime efficiency, and hyperparameter usability and robustness. Based on various evaluation results, this paper gives dataset-specific suggestions for the usage of DA testing methods.

Robotic data acquisition with deep learning enables cell image-based prediction of transcriptomic phenotypes

Single-cell whole-transcriptome analysis is the gold standard approach to identifying molecularly defined cell phenotypes. However, this approach cannot be used for dynamics measurements such as live-cell imaging. Here, we developed a multifunctional robot, the automated live imaging and cell picking system (ALPS) and used it to perform single-cell RNA sequencing for microscopically observed cells with multiple imaging modes. Using robotically obtained data that linked cell images and the whole transcriptome, we successfully predicted transcriptome-defined cell phenotypes in a noninvasive manner using cell image-based deep learning. This noninvasive approach opens a window to determine the live-cell whole transcriptome in real time. Moreover, this work, which is based on a data-driven approach, is a proof of concept for determining the transcriptome-defined phenotypes (i.e., not relying on specific genes) of any cell from cell images using a model trained on linked datasets.

The role of mitosis in generating fitness heterogeneity

Cancer cells have heterogeneous fitness, and this heterogeneity stems from genetic and epigenetic sources. Here, we sought to assess the contribution of asymmetric mitosis (AM) and time on the variability of fitness in sister cells. Around one quarter of sisters had differences in fitness, assessed as the intermitotic time (IMT), from 330 to 510 min. Phenotypes related to fitness, such as ERK activity (herein referring to ERK1 and ERK2, also known as MAPK3 and MAPK1, respectively), DNA damage and nuclear morphological phenotypes were also asymmetric at mitosis or turned asymmetric over the course of the cell cycle. The ERK activity of mother cell was found to influence the ERK activity and the IMT of the daughter cells, and cells with ERK asymmetry at mitosis produced more offspring with AMs, suggesting heritability of the AM phenotype for ERK activity. Our findings demonstrate how variabilities in sister cells can be generated, contributing to the phenotype heterogeneities in tumor cells.

NF-κB dynamics in the language of immune cells

Biological discovery has been driven by advances in throughput and resolution of analysis technologies. They have also created an indelible bias for snapshot-based knowledge. Even though recent methods such as multi-omics single-cell assays have empowered immunological investigations, they still provide snapshots of cellular behaviors and thus, have inherent limitations in reconstructing unsynchronized dynamic events across individual cells. Here, we present a rationale for how NF-κB may convey specificity of contextual information through subtle quantitative features of its signaling dynamics. The next frontier of predictive understanding should involve functional characterization of NF-κB signaling dynamics and their immunological implications. This may help solve the apparent paradox that a ubiquitously activated transcription factor can shape accurate responses to different immune challenges.

Combining single-cell tracking and omics improves blood stem cell fate regulator identification

Molecular programs initiating cell fate divergence (CFD) are difficult to identify. Current approaches usually compare cells long after CFD initiation, therefore missing molecular changes at its start. Ideally, single cells that differ in their CFD molecular program but are otherwise identical are compared early in CFD. This is possible in diverging sister cells, which were identical until their mother's division and thus differ mainly in CFD properties. In asymmetrically dividing cells, divergent daughter fates are prospectively committed during division, and diverging sisters can thus be identified at the start of CFD. Using asymmetrically dividing blood stem cells, we developed a pipeline (ie, trackSeq) for imaging, tracking, isolating, and transcriptome sequencing of single cells. Their identities, kinship, and histories are maintained throughout, massively improving molecular noise filtering and candidate identification. In addition to many identified blood stem CFD regulators, we offer here this pipeline for use in CFDs other than asymmetric division.

Integrating temporal single-cell gene expression modalities for trajectory inference and disease prediction

BACKGROUND

Current methods for analyzing single-cell datasets have relied primarily on static gene expression measurements to characterize the molecular state of individual cells. However, capturing temporal changes in cell state is crucial for the interpretation of dynamic phenotypes such as the cell cycle, development, or disease progression. RNA velocity infers the direction and speed of transcriptional changes in individual cells, yet it is unclear how these temporal gene expression modalities may be leveraged for predictive modeling of cellular dynamics.

RESULTS

Here, we present the first task-oriented benchmarking study that investigates integration of temporal sequencing modalities for dynamic cell state prediction. We benchmark ten integration approaches on ten datasets spanning different biological contexts, sequencing technologies, and species. We find that integrated data more accurately infers biological trajectories and achieves increased performance on classifying cells according to perturbation and disease states. Furthermore, we show that simple concatenation of spliced and unspliced molecules performs consistently well on classification tasks and can be used over more memory intensive and computationally expensive methods.

CONCLUSIONS

This work illustrates how integrated temporal gene expression modalities may be leveraged for predicting cellular trajectories and sample-associated perturbation and disease phenotypes. Additionally, this study provides users with practical recommendations for task-specific integration of single-cell gene expression modalities.

Quantifying the phenotypic information in mRNA abundance

Quantifying the dependency between mRNA abundance and downstream cellular phenotypes is a fundamental open problem in biology. Advances in multimodal single‐cell measurement technologies provide an opportunity to apply new computational frameworks to dissect the contribution of individual genes and gene combinations to a given phenotype. Using an information theory approach, we analyzed multimodal data of the expression of 83 genes in the Ca²⁺ signaling network and the dynamic Ca²⁺ response in the same cell. We found that the overall expression levels of these 83 genes explain approximately 60% of Ca²⁺ signal entropy. The average contribution of each single gene was 17%, revealing a large degree of redundancy between genes. Using different heuristics, we estimated the dependency between the size of a gene set and its information content, revealing that on average, a set of 53 genes contains 54% of the information about Ca²⁺ signaling. Our results provide the first direct quantification of information content about complex cellular phenotype that exists in mRNA abundance measurements.

New Insights in the Interplay Between African Swine Fever Virus and Innate Immunity and Its Impact on Viral Pathogenicity

The continuous spread of African swine fever virus (ASFV) in Europe and Asia represents a major threat to livestock health, with billions of dollars of income losses and major perturbations of the global pig industry. One striking feature of African swine fever (ASF) is the existence of different forms of the disease, ranging from acute with mortality rates approaching 100% to chronic, with mild clinical manifestations. These differences in pathogenicity have been linked to genomic alterations present in attenuated ASFV strains (and absent in virulent ones) and differences in the immune response of infected animals. In this mini-review, we summarized current knowledge on the connection between ASFV pathogenicity and the innate immune response induced in infected hosts, with a particular focus on the pathways involved in ASFV detection. Indeed, recent studies have highlighted the key role of the DNA sensor cGAS in ASFV sensing. We discussed what other pathways may be involved in ASFV sensing and inflammasome activation and summarized recent findings on the viral ASFV genes involved in the modulation of the interferon (IFN) and nuclear factor kappa B (NF-κB) pathways.

Insights on the NF-κB System Using Live Cell Imaging: Recent Developments and Future Perspectives

The transcription factor family of nuclear factor kappa B (NF-κB) proteins is widely recognized as a key player in inflammation and the immune responses, where it plays a fundamental role in translating external inflammatory cues into precise transcriptional programs, including the timely expression of a wide variety of cytokines/chemokines. Live cell imaging in single cells showed approximately 15 years ago that the canonical activation of NF-κB upon stimulus is very dynamic, including oscillations of its nuclear localization with a period close to 1.5 hours. This observation has triggered a fruitful interdisciplinary research line that has provided novel insights on the NF-κB system: how its heterogeneous response differs between cell types but also within homogeneous populations; how NF-κB dynamics translate external cues into intracellular signals and how NF-κB dynamics affects gene expression. Here we review the main features of this live cell imaging approach to the study of NF-κB, highlighting the key findings, the existing gaps of knowledge and hinting towards some of the potential future steps of this thriving research field.

Frontiers in single cell analysis: multimodal technologies and their clinical perspectives

Single cell multimodal analysis is at the frontier of single cell research: it defines the roles and functions of distinct cell types through simultaneous analysis to provide unprecedented insight into cellular processes. Current single cell approaches are rapidly moving toward multimodal characterizations. It replaces one-dimensional single cell analysis, for example by allowing for simultaneous measurement of transcription and post-transcriptional regulation, epigenetic modifications and/or surface protein expression. By providing deeper insights into single cell processes, multimodal single cell analyses paves the way to new understandings in various cellular processes such as cell fate decisions, physiological heterogeneity or genotype-phenotype linkages. At the forefront of this, microfluidics is key for high-throughput single cell analysis. Here, we present an overview of the recent multimodal microfluidic platforms having a potential in biomedical research, with a specific focus on their potential clinical applications.

Signaling oscillations in embryonic development

Tight spatiotemporal control of cellular behavior and cell fate decisions is paramount to the formation of multicellular organisms during embryonic development. Intercellular communication via signaling pathways mediates this control. Interestingly, these signaling pathways are not static, but dynamic and change in activity over time. Signaling oscillations as a specific type of dynamics are found in various signaling pathways and model systems. Functions of oscillations include the regulation of periodic events or the transmission of information by encoding signals in the dynamic properties of a signaling pathway. For instance, signaling oscillations in neural or pancreatic progenitor cells modulate their proliferation and differentiation. Oscillations between neighboring cells can also be synchronized, leading to the emergence of waves traveling through the tissue. Such population-wide signaling oscillations regulate for example the consecutive segmentation of vertebrate embryos, a process called somitogenesis. Here, we outline our current understanding of signaling oscillations in embryonic development, how signaling oscillations are generated, how they are studied and how they contribute to the regulation of embryonic development.

Functional Hallmarks of Healthy Macrophage Responses: Their Regulatory Basis and Disease Relevance

Macrophages are first responders for the immune system. In this role, they have both effector functions for neutralizing pathogens and sentinel functions for alerting other immune cells of diverse pathologic threats, thereby initiating and coordinating a multipronged immune response. Macrophages are distributed throughout the body-they circulate in the blood, line the mucosal membranes, reside within organs, and survey the connective tissue. Several reviews have summarized their diverse roles in different physiological scenarios and in the initiation or amplification of different pathologies. In this review, we propose that both the effector and the sentinel functions of healthy macrophages rely on three hallmark properties: response specificity, context dependence, and stimulus memory. When these hallmark properties are diminished, the macrophage's biological functions are impaired, which in turn results in increased risk for immune dysregulation, manifested by immune deficiency or autoimmunity. We review the evidence and the molecular mechanisms supporting these functional hallmarks.

NfκB signaling dynamics and their target genes differ between mouse blood cell types and induce distinct cell behavior

Cells can use signaling pathway activity over time (i.e., dynamics) to control cell fates. However, little is known about the potential existence and function of signaling dynamics in primary hematopoietic stem and progenitor cells (HSPCs). Here, we use time-lapse imaging and tracking of single murine HSPCs from GFP-p65/H2BmCherry reporter mice to quantify their nuclear factor κB (NfκB) activity dynamics in response to TNFα and IL1β. We find response dynamics to be heterogeneous between individual cells, with cell type specific dynamics distributions. Transcriptome sequencing of single cells physically isolated after live dynamics quantification shows activation of different target gene programs in cells with different dynamics. Finally, artificial induction of oscillatory NfκB activity causes changes in GMP behavior. Thus, HSPC behavior can be influenced by signaling dynamics, which are tightly regulated during hematopoietic differentiation and enable cell type specific responses to the same signaling inputs.

Dynamics of Drosophila endoderm specification

During early Drosophila embryogenesis, a network of gene regulatory interactions orchestrates terminal patterning, playing a critical role in the subsequent formation of the gut. We utilized CRISPR gene editing at endogenous loci to create live reporters of transcription and light-sheet microscopy to monitor the individual components of the posterior gut patterning network across 90 min prior to gastrulation. We developed a computational approach for fusing imaging datasets of the individual components into a common multivariable trajectory. Data fusion revealed low intrinsic dimensionality of posterior patterning and cell fate specification in wild-type embryos. The simple structure that we uncovered allowed us to construct a model of interactions within the posterior patterning regulatory network and make testable predictions about its dynamics at the protein level. The presented data fusion strategy is a step toward establishing a unified framework that would explore how stochastic spatiotemporal signals give rise to highly reproducible morphogenetic outcomes.

DynaMorph: self-supervised learning of morphodynamic states of live cells

A cell’s shape and motion represent fundamental aspects of cell identity and can be highly predictive of function and pathology. However, automated analysis of the morphodynamic states remains challenging for most cell types, especially primary human cells where genetic labeling may not be feasible. To enable automated and quantitative analysis of morphodynamic states, we developed DynaMorph—a computational framework that combines quantitative live cell imaging with self-supervised learning. To demonstrate the robustness and utility of this approach, we used DynaMorph to annotate morphodynamic states observed with label-free measurements of optical density and anisotropy of live microglia isolated from human brain tissue. These cells show complex behavior and have varied responses to disease-relevant perturbations. DynaMorph generates quantitative morphodynamic representations that can be used to compare the effects of the perturbations. Using DynaMorph, we identify distinct morphodynamic states of microglia polarization and detect rare transition events between states. The concepts and the methods presented here can facilitate automated discovery of functional states of diverse cellular systems.

Systems approaches to investigate the role of NF-κB signaling in aging

The nuclear factor-κB (NF-κB) signaling pathway is one of the most well-studied pathways related to inflammation, and its involvement in aging has attracted considerable attention. As aging is a complex phenomenon and is the result of a multi-step process, the involvement of the NF-κB pathway in aging remains unclear. To elucidate the role of NF-κB in the regulation of aging, different systems biology approaches have been employed. A multi-omics data-driven approach can be used to interpret and clarify unknown mechanisms but cannot generate mechanistic regulatory structures alone. In contrast, combining this approach with a mathematical modeling approach can identify the mechanistics of the phenomena of interest. The development of single-cell technologies has also helped clarify the heterogeneity of the NF-κB response and underlying mechanisms. Here, we review advances in the understanding of the regulation of aging by NF-κB by focusing on omics approaches, single-cell analysis, and mathematical modeling of the NF-κB network.

Profiling transcription factor activity dynamics using intronic reads in time-series transcriptome data

Activities of transcription factors (TFs) are temporally modulated to regulate dynamic cellular processes, including development, homeostasis, and disease. Recent developments of bioinformatic tools have enabled the analysis of TF activities using transcriptome data. However, because these methods typically use exon-based target expression levels, the estimated TF activities have limited temporal accuracy. To address this, we proposed a TF activity measure based on intron-level information in time-series RNA-seq data, and implemented it to decode the temporal control of TF activities during dynamic processes. We showed that TF activities inferred from intronic reads can better recapitulate instantaneous TF activities compared to the exon-based measure. By analyzing public and our own time-series transcriptome data, we found that intron-based TF activities improve the characterization of temporal phasing of cycling TFs during circadian rhythm, and facilitate the discovery of two temporally opposing TF modules during T cell activation. Collectively, we anticipate that the proposed approach would be broadly applicable for decoding global transcriptional architecture during dynamic processes.

The Systems and Synthetic Biology of Auxin

Auxin biology as a field has been at the forefront of advances in delineating the structures, dynamics, and control of plant growth networks. Advances have been enabled by combining the complementary fields of top-down, holistic systems biology and bottom-up, build-to-understand synthetic biology. Continued collaboration between these approaches will facilitate our understanding of and ability to engineer auxin's control of plant growth, development, and physiology. There is a need for the application of similar complementary approaches to improving equity and justice through analysis and redesign of the human systems in which this research is undertaken.

CellVGAE: an unsupervised scRNA-seq analysis workflow with graph attention networks

MOTIVATION

Single-cell RNA sequencing allows high-resolution views of individual cells for libraries of up to millions of samples, thus motivating the use of deep learning for analysis. In this study, we introduce the use of graph neural networks for the unsupervised exploration of scRNA-seq data by developing a variational graph autoencoder architecture with graph attention layers that operates directly on the connectivity between cells, focusing on dimensionality reduction and clustering. With the help of several case studies, we show that our model, named CellVGAE, can be effectively used for exploratory analysis even on challenging datasets, by extracting meaningful features from the data and providing the means to visualize and interpret different aspects of the model.

RESULTS

We show that CellVGAE is more interpretable than existing scRNA-seq variational architectures by analysing the graph attention coefficients. By drawing parallels with other scRNA-seq studies on interpretability, we assess the validity of the relationships modelled by attention, and furthermore, we show that CellVGAE can intrinsically capture information such as pseudotime and NF-ĸB activation dynamics, the latter being a property that is not generally shared by existing neural alternatives. We then evaluate the dimensionality reduction and clustering performance on 9 difficult and well-annotated datasets by comparing with three leading neural and non-neural techniques, concluding that CellVGAE outperforms competing methods. Finally, we report a decrease in training times of up to × 20 on a dataset of 1.3 million cells compared to existing deep learning architectures.

AVAILABILITYAND IMPLEMENTATION

The CellVGAE code is available at https://github.com/davidbuterez/CellVGAE.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

The Origins of Phenotypic Heterogeneity in Cancer

Heterogeneity is a pervasive feature of cancer, and understanding the sources and regulatory mechanisms underlying heterogeneity could provide key insights to help improve the diagnosis and treatment of cancer. In this review, we discuss the origin of heterogeneity in the phenotype of individual cancer cells. Genotype-phenotype (G-P) maps are widely used in evolutionary biology to represent the complex interactions of genes and the environment that lead to phenotypes that impact fitness. Here, we present the rationale of an extended G-P (eG-P) map with a cone structure in cancer. The eG-P cone is formed by cells that are similar at the genome layer but gradually increase variability in the epigenome, transcriptome, proteome, metabolome and signalome layers to produce large variability at the phenome layer. Experimental evidence from single-cell -omics analyses supporting the cancer eG-P cone concept is presented, and the impact of epimutations and the interaction of cancer and tumor microenvironmental eG-P cones are integrated with the current understanding of cancer biology. The eG-P cone concept uncovers potential therapeutic strategies to reduce cancer evolution and improve cancer treatment. More methods to study phenotypes in single cells will be key to better understand cancer cell fitness in tumor biology and therapeutics.

High-throughput and high-efficiency sample preparation for single-cell proteomics using a nested nanowell chip

Global quantification of protein abundances in single cells could provide direct information on cellular phenotypes and complement transcriptomics measurements. However, single-cell proteomics is still immature and confronts many technical challenges. Herein we describe a nested nanoPOTS (N2) chip to improve protein recovery, operation robustness, and processing throughput for isobaric-labeling-based scProteomics workflow. The N2 chip reduces reaction volume to <30 nL and increases capacity to >240 single cells on a single microchip. The tandem mass tag (TMT) pooling step is simplified by adding a microliter droplet on the nested nanowells to combine labeled single-cell samples. In the analysis of ~100 individual cells from three different cell lines, we demonstrate that the N2 chip-based scProteomics platform can robustly quantify ~1500 proteins and reveal membrane protein markers. Our analyses also reveal low protein abundance variations, suggesting the single-cell proteome profiles are highly stable for the cells cultured under identical conditions.

HES1 protein oscillations are necessary for neural stem cells to exit from quiescence

Quiescence is a dynamic process of reversible cell cycle arrest. High-level persistent expression of the HES1 transcriptional repressor, which oscillates with an ultradian periodicity in proliferative neural stem cells (NSCs), is thought to mediate quiescence. However, it is not known whether this is due to a change in levels or dynamics. Here, we induce quiescence in embryonic NSCs with BMP4, which does not increase HES1 level, and we find that HES1 continues to oscillate. To assess the role of HES1 dynamics, we express persistent HES1 under a moderate strength promoter, which overrides the endogenous oscillations while maintaining the total HES1 level within physiological range. We find that persistent HES1 does not affect proliferation or entry into quiescence; however, exit from quiescence is impeded. Thus, oscillatory expression of HES1 is specifically required for NSCs to exit quiescence, a finding of potential importance for controlling reactivation of stem cells in tissue regeneration and cancer.

Two-level modeling approach to identify the regulatory dynamics capturing drug response heterogeneity in single-cells

Single-cell multimodal technologies reveal the scales of cellular heterogeneity impairing cancer treatment, yet cell response dynamics remain largely underused to decipher the mechanisms of drug resistance they take part in. As the phenotypic heterogeneity of a clonal cell population informs on the capacity of each single-cell to recapitulate the whole range of observed behaviors, we developed a modeling approach utilizing single-cell response data to identify regulatory reactions driving population heterogeneity in drug response. Dynamic data of hundreds of HeLa cells treated with TNF-related apoptosis-inducing ligand (TRAIL) were used to characterize the fate-determining kinetic parameters of an apoptosis receptor reaction model. Selected reactions sets were augmented to incorporate a mechanism that leads to the separation of the opposing response phenotypes. Using a positive feedback loop motif to identify the reaction set, we show that caspase-8 is able to encapsulate high levels of heterogeneity by introducing a response delay and amplifying the initial differences arising from natural protein expression variability. Our approach enables the identification of fate-determining reactions that drive the population response heterogeneity, providing regulatory targets to curb the cell dynamics of drug resistance.

Stimulus-specific responses in innate immunity: Multilayered regulatory circuits

Immune sentinel cells initiate immune responses to pathogens and tissue injury and are capable of producing highly stimulus-specific responses. Insight into the mechanisms underlying such specificity has come from the identification of regulatory factors and biochemical pathways, as well as the definition of signaling circuits that enable combinatorial and temporal coding of information. Here, we review the multi-layered molecular mechanisms that underlie stimulus-specific gene expression in macrophages. We categorize components of inflammatory and anti-pathogenic signaling pathways into five layers of regulatory control and discuss unifying mechanisms determining signaling characteristics at each layer. In this context, we review mechanisms that enable combinatorial and temporal encoding of information, identify recurring regulatory motifs and principles, and present strategies for integrating experimental and computational approaches toward the understanding of signaling specificity in innate immunity.

DeepLINK: Deep learning inference using knockoffs with applications to genomics

We propose a deep learning-based knockoffs inference framework, DeepLINK, that guarantees the false discovery rate (FDR) control in high-dimensional settings. DeepLINK is applicable to a broad class of covariate distributions described by the possibly nonlinear latent factor models. It consists of two major parts: an autoencoder network for the knockoff variable construction and a multilayer perceptron network for feature selection with the FDR control. The empirical performance of DeepLINK is investigated through extensive simulation studies, where it is shown to achieve FDR control in feature selection with both high selection power and high prediction accuracy. We also apply DeepLINK to three real data applications to demonstrate its practical utility.

Analyzing signaling activity and function in hematopoietic cells

Cells constantly sense their environment, allowing the adaption of cell behavior to changing needs. Fine-tuned responses to complex inputs are computed by signaling pathways, which are wired in complex connected networks. Their activity is highly context-dependent, dynamic, and heterogeneous even between closely related individual cells. Despite lots of progress, our understanding of the precise implementation, relevance, and possible manipulation of cellular signaling in health and disease therefore remains limited. Here, we discuss the requirements, potential, and limitations of the different current technologies for the analysis of hematopoietic stem and progenitor cell signaling and its effect on cell fates.

Computational methods for characterizing and learning from heterogeneous cell signaling data

Heterogeneity in cell signaling pathways is increasingly appreciated as a fundamental feature of cell biology and a driver of clinically relevant disease phenotypes. Understanding the causes of heterogeneity, the cellular mechanisms used to control heterogeneity, and the downstream effects of heterogeneity in single cells are all key obstacles for manipulating cellular populations and treating disease. Recent advances in genetic engineering, including multiplexed fluorescent reporters, have provided unprecedented measurements of signaling heterogeneity, but these vast data sets are often difficult to interpret, necessitating the use of computational techniques to extract meaning from the data. Here, we review recent advances in computational methods for extracting meaning from these novel data streams. In particular, we evaluate how machine learning methods related to dimensionality reduction and classification can identify structure in complex, dynamic datasets, simplifying interpretation. We also discuss how mechanistic models can be merged with heterogeneous data to understand the underlying differences between cells in a population. These methods are still being developed, but the work reviewed here offers useful applications of specific analysis techniques that could enable the translation of single-cell signaling data to actionable biological understanding.

Six distinct NFκB signaling codons convey discrete information to distinguish stimuli and enable appropriate macrophage responses

Macrophages initiate inflammatory responses via the transcription factor NFκB. The temporal pattern of NFκB activity determines which genes are expressed and thus, the type of response that ensues. Here, we examined how information about the stimulus is encoded in the dynamics of NFκB activity. We generated an mVenus-RelA reporter mouse line to enable high-throughput live-cell analysis of primary macrophages responding to host- and pathogen-derived stimuli. An information-theoretic workflow identified six dynamical features-termed signaling codons-that convey stimulus information to the nucleus. In particular, oscillatory trajectories were a hallmark of responses to cytokine but not pathogen-derived stimuli. Single-cell imaging and RNA sequencing of macrophages from a mouse model of Sjögren's syndrome revealed inappropriate responses to stimuli, suggestive of confusion of two NFκB signaling codons. Thus, the dynamics of NFκB signaling classify immune threats through six signaling codons, and signal confusion based on defective codon deployment may underlie the etiology of some inflammatory diseases.

Time will tell: The temporal code of immune threats

Activation of NF-κB is a common downstream consequence of inflammatory stimulation, yet it achieves stimulus-specific transcriptional responses. In this issue of Immunity, Adelaja et al. use single-cell imaging and computational approaches to understand temporal features of NF-κB dynamics that transmit information about immune threats.

Epigenetically regulated digital signaling defines epithelial innate immunity at the tissue level

To prevent damage to the host or its commensal microbiota, epithelial tissues must match the intensity of the immune response to the severity of a biological threat. Toll-like receptors allow epithelial cells to identify microbe associated molecular patterns. However, the mechanisms that mitigate biological noise in single cells to ensure quantitatively appropriate responses remain unclear. Here we address this question using single cell and single molecule approaches in mammary epithelial cells and primary organoids. We find that epithelial tissues respond to bacterial microbe associated molecular patterns by activating a subset of cells in an all-or-nothing (i.e. digital) manner. The maximum fraction of responsive cells is regulated by a bimodal epigenetic switch that licenses the TLR2 promoter for transcription across multiple generations. This mechanism confers a flexible memory of inflammatory events as well as unique spatio-temporal control of epithelial tissue-level immune responses. We propose that epigenetic licensing in individual cells allows for long-term, quantitative fine-tuning of population-level responses.

NF-κB responds to absolute differences in cytokine concentrations

Cells receive a wide range of dynamic signaling inputs during immune regulation, but how gene regulatory networks measure such dynamic inputs is not well understood. Here, we used microfluidic single-cell analysis and mathematical modeling to study how the NF-κB pathway responds to immune inputs that vary over time such as increasing, decreasing, or fluctuating cytokine signals. We found that NF-κB activity responded to the absolute difference in cytokine concentration and not to the concentration itself. Our analyses revealed that negative feedback by the regulatory proteins A20 and IκBα enabled differential responses to changes in cytokine dose by providing a short-term memory of previous cytokine concentrations and by continuously resetting kinase cycling and receptor abundance. Investigation of NF-κB target gene expression showed that cells exhibited distinct transcriptional responses under different dynamic cytokine profiles. Our results demonstrate how cells use simple network motifs and transcription factor dynamics to efficiently extract information from complex signaling environments.