Causal Gene Regulatory Network Modeling and Genomics: Second-Generation Challenges

Identifying reproducible transcription regulator coexpression patterns with single cell transcriptomics

The proliferation of single cell transcriptomics has potentiated our ability to unveil patterns that reflect dynamic cellular processes such as the regulation of gene transcription. In this study, we leverage a broad collection of single cell RNA-seq data to identify the gene partners whose expression is most coordinated with each human and mouse transcription regulator (TR). We assembled 120 human and 103 mouse scRNA-seq datasets from the literature (>28 million cells), constructing a single cell coexpression network for each. We aimed to understand the consistency of TR coexpression profiles across a broad sampling of biological contexts, rather than examine the preservation of context-specific signals. Our workflow therefore explicitly prioritizes the patterns that are most reproducible across cell types. Towards this goal, we characterize the similarity of each TR's coexpression within and across species. We create single cell coexpression rankings for each TR, demonstrating that this aggregated information recovers literature curated targets on par with ChIP-seq data. We then combine the coexpression and ChIP-seq information to identify candidate regulatory interactions supported across methods and species. Finally, we highlight interactions for the important neural TR ASCL1 to demonstrate how our compiled information can be adopted for community use.

Lineage-Selective Dependencies in Pediatric Cancers

The quest for effective cancer therapeutics has traditionally centered on targeting mutated or overexpressed oncogenic proteins. However, challenges arise in cancers with low mutational burden or when the mutated oncogene is not conventionally targetable, which are common situations in childhood cancers. This obstacle has sparked large-scale unbiased screens to identify collateral genetic dependencies crucial for cancer cell growth. These screens have revealed promising targets for therapeutic intervention in the form of lineage-selective dependency genes, which may have an expanded therapeutic window compared to pan-lethal dependencies. Many lineage-selective dependencies regulate gene expression and are closely tied to the developmental origins of pediatric tumors. Placing lineage-selective dependencies in a transcriptional network model is helpful for understanding their roles in driving malignant cell behaviors. Here, we discuss the identification of lineage-selective dependencies and how two transcriptional models, core regulatory circuits and gene regulatory networks, can serve as frameworks for understanding their individual and collective actions, particularly in cancers affecting children and young adults.

Identifying Reproducible Transcription Regulator Coexpression Patterns with Single Cell Transcriptomics

The proliferation of single cell transcriptomics has potentiated our ability to unveil patterns that reflect dynamic cellular processes such as the regulation of gene transcription. In this study, we leverage a broad collection of single cell RNA-seq data to identify the gene partners whose expression is most coordinated with each human and mouse transcription regulator (TR). We assembled 120 human and 103 mouse scRNA-seq datasets from the literature (>28 million cells), constructing a single cell coexpression network for each. We aimed to understand the consistency of TR coexpression profiles across a broad sampling of biological contexts, rather than examine the preservation of context-specific signals. Our workflow therefore explicitly prioritizes the patterns that are most reproducible across cell types. Towards this goal, we characterize the similarity of each TR's coexpression within and across species. We create single cell coexpression rankings for each TR, demonstrating that this aggregated information recovers literature curated targets on par with ChIP-seq data. We then combine the coexpression and ChIP-seq information to identify candidate regulatory interactions supported across methods and species. Finally, we highlight interactions for the important neural TR ASCL1 to demonstrate how our compiled information can be adopted for community use.

LM-Merger: A workflow for merging logical models with an application to gene regulation

Motivation

Gene regulatory network (GRN) models provide mechanistic understanding of genetic interactions that regulate gene expression and, consequently, influence cellular behavior. Dysregulated gene expression plays a critical role in disease progression and treatment response, making GRN models a promising tool for precision medicine. While researchers have built many models to describe specific subsets of gene interactions, more comprehensive models that cover a broader range of genes are challenging to build. This necessitates the development of automated approaches for merging existing models.

Results

We present LM-Merger, a workflow for semi-automatically merging logical GRN models. The workflow consists of five main steps: (a) model identification, (b) model standardization and annotation, (c) model verification, (d) model merging, and (d) model evaluation. We demonstrate the feasibility and benefit of this workflow with two pairs of published models pertaining to acute myeloid leukemia (AML). The integrated models were able to retain the predictive accuracy of the original models, while expanding coverage of the biological system. Notably, when applied to a new dataset, the integrated models outperformed the individual models in predicting patient response. This study highlights the potential of logical model merging to advance systems biology research and our understanding of complex diseases.

Availability and implementation

The workflow and accompanying tools, including modules for model standardization, automated logical model merging, and evaluation, are available at https://github.com/IlyaLab/LogicModelMerger/.

Characterizing the targets of transcription regulators by aggregating ChIP-seq and perturbation expression data sets

Mapping the gene targets of chromatin-associated transcription regulators (TRs) is a major goal of genomics research. ChIP-seq of TRs and experiments that perturb a TR and measure the differential abundance of gene transcripts are a primary means by which direct relationships are tested on a genomic scale. It has been reported that there is a poor overlap in the evidence across gene regulation strategies, emphasizing the need for integrating results from multiple experiments. Although research consortia interested in gene regulation have produced a valuable trove of high-quality data, there is an even greater volume of TR-specific data throughout the literature. In this study, we show a workflow for the identification, uniform processing, and aggregation of ChIP-seq and TR perturbation experiments for the ultimate purpose of ranking human and mouse TR-target interactions. Focusing on an initial set of eight regulators (ASCL1, HES1, MECP2, MEF2C, NEUROD1, PAX6, RUNX1, and TCF4), we identified 497 experiments suitable for analysis. We used this corpus to examine data concordance, to identify systematic patterns of the two data types, and to identify putative orthologous interactions between human and mouse. We build upon commonly used strategies to forward a procedure for aggregating and combining these two genomic methodologies, assessing these rankings against independent literature-curated evidence. Beyond a framework extensible to other TRs, our work also provides empirically ranked TR-target listings, as well as transparent experiment-level gene summaries for community use.

Multi-modular structure of the gene regulatory network for specification and commitment of murine T cells

T cells develop from multipotent progenitors by a gradual process dependent on intrathymic Notch signaling and coupled with extensive proliferation. The stages leading them to T-cell lineage commitment are well characterized by single-cell and bulk RNA analyses of sorted populations and by direct measurements of precursor-product relationships. This process depends not only on Notch signaling but also on multiple transcription factors, some associated with stemness and multipotency, some with alternative lineages, and others associated with T-cell fate. These factors interact in opposing or semi-independent T cell gene regulatory network (GRN) subcircuits that are increasingly well defined. A newly comprehensive picture of this network has emerged. Importantly, because key factors in the GRN can bind to markedly different genomic sites at one stage than they do at other stages, the genes they significantly regulate are also stage-specific. Global transcriptome analyses of perturbations have revealed an underlying modular structure to the T-cell commitment GRN, separating decisions to lose "stem-ness" from decisions to block alternative fates. Finally, the updated network sheds light on the intimate relationship between the T-cell program, which depends on the thymus, and the innate lymphoid cell (ILC) program, which does not.

Promoter methylation in a mixed feedback loop circadian clock model

We investigate how epigenetic modifications to clock gene promoters affect transcriptomic activity in the circadian clock. Motivated by experimental observations that link DNA methylation with the behavior of the clock, we introduce and analyze an extension of the mixed feedback loop (MFL) model of François and Hakim. We extend the original model to include an additional methylated promoter state and allow for reversible protein sequestration, an important feature for circadian applications. First, working with the general form of the MFL model, we find that the qualitative behavior of the model is dictated by the promoter state with the highest transcription rate. We then build on the work of Kim and Forger, who analyzed the stability of the mammalian circadian clock by using a reduced form of the MFL model. We present a rigorous procedure for translating between the MFL model and the reduction of Kim and Forger. We then propose a model reduction more appropriate for the study of oscillatory promoter states, making use of a fully coupled quasi-steady-state approximation rather than the standard partially uncoupled quasi-steady-state approach. Working with the novel reduced form of the model, we find substantial differences in the transcription function and show that, although methylation contributes to period control, excessive methylation can abolish rhythmicity. Altogether our results show that even in a minimal clock model, DNA methylation has a nontrivial influence on the system's ability to oscillate.

Langley M, Fisher CL, Aguilar-Hidalgo D, Shukla S, Yachie-Kinoshita A, Hughes M, M. McNagny K, Zandstra PW. IQCELL: A platform for predicting the effect of gene perturbations on developmental trajectories using single-cell RNA-seq data

The increasing availability of single-cell RNA-sequencing (scRNA-seq) data from various developmental systems provides the opportunity to infer gene regulatory networks (GRNs) directly from data. Herein we describe IQCELL, a platform to infer, simulate, and study executable logical GRNs directly from scRNA-seq data. Such executable GRNs allow simulation of fundamental hypotheses governing developmental programs and help accelerate the design of strategies to control stem cell fate. We first describe the architecture of IQCELL. Next, we apply IQCELL to scRNA-seq datasets from early mouse T-cell and red blood cell development, and show that the platform can infer overall over 74% of causal gene interactions previously reported from decades of research. We will also show that dynamic simulations of the generated GRN qualitatively recapitulate the effects of known gene perturbations. Finally, we implement an IQCELL gene selection pipeline that allows us to identify candidate genes, without prior knowledge. We demonstrate that GRN simulations based on the inferred set yield results similar to the original curated lists. In summary, the IQCELL platform offers a versatile tool to infer, simulate, and study executable GRNs in dynamic biological systems.

Proteomic/transcriptomic analysis of erythropoiesis

PURPOSE OF REVIEW

Erythropoiesis is a hierarchical process by which hematopoietic stem cells give rise to red blood cells through gradual cell fate restriction and maturation. Deciphering this process requires the establishment of dynamic gene regulatory networks (GRNs) that predict the response of hematopoietic cells to signals from the environment. Although GRNs have historically been derived from transcriptomic data, recent proteomic studies have revealed a major role for posttranscriptional mechanisms in regulating gene expression during erythropoiesis. These new findings highlight the need to integrate proteomic data into GRNs for a refined understanding of erythropoiesis.

RECENT FINDINGS

Here, we review recent proteomic studies that have furthered our understanding of erythropoiesis with a focus on quantitative mass spectrometry approaches to measure the abundance of transcription factors and cofactors during differentiation. Furthermore, we highlight challenges that remain in integrating transcriptomic, proteomic, and other omics data into a predictive model of erythropoiesis, and discuss the future prospect of single-cell proteomics.

SUMMARY

Recent proteomic studies have considerably expanded our knowledge of erythropoiesis beyond the traditional transcriptomic-centric perspective. These findings have both opened up new avenues of research to increase our understanding of erythroid differentiation, while at the same time presenting new challenges in integrating multiple layers of information into a comprehensive gene regulatory model.

How haematopoiesis research became a fertile ground for regulatory network biology as pioneered by Eric Davidson

PURPOSE OF REVIEW

This historical perspective reviews how work of Eric H. Davidson was a catalyst and exemplar for explaining haematopoietic cell fate determination through gene regulation.

RECENT FINDINGS

Researchers studying blood and immune cells pioneered many of the early mechanistic investigations of mammalian gene regulatory processes. These efforts included the characterization of complex gene regulatory sequences exemplified by the globin and T-cell/B-cell receptor gene loci, as well as the identification of many key regulatory transcription factors through the fine mapping of chromosome translocation breakpoints in leukaemia patients. As the repertoire of known regulators expanded, assembly into gene regulatory network models became increasingly important, not only to account for the truism that regulatory genes do not function in isolation but also to devise new ways of extracting biologically meaningful insights from even more complex information. Here we explore how Eric H. Davidson's pioneering studies of gene regulatory network control in nonvertebrate model organisms have had an important and lasting impact on research into blood and immune cell development.

SUMMARY

The intellectual framework developed by Davidson continues to contribute to haematopoietic research, and his insistence on demonstrating logic and causality still challenges the frontier of research today.

A dichotomy of gene regulatory associations during the activated B-cell to plasmablast transition

The activated B-cell (ABC) to plasmablast transition encompasses the cusp of antibody-secreting cell (ASC) differentiation. We explore this transition with integrated analysis in human cells, focusing on changes that follow removal from CD40-mediated signals. Within hours of input signal loss, cell growth programs shift toward enhanced proliferation, accompanied by ER-stress response, and up-regulation of ASC features. Clustering of genomic occupancy for IRF4, BLIMP1, XBP1, and CTCF with histone marks identifies a dichotomy: XBP1 and IRF4 link to induced but not repressed gene modules in plasmablasts, whereas BLIMP1 links to modules of ABC genes that are repressed, but not to activated genes. Between ABC and plasmablast states, IRF4 shifts away from AP1/IRF composite elements while maintaining occupancy at IRF and ETS/IRF elements. This parallels the loss of BATF expression, which is identified as a potential BLIMP1 target. In plasmablasts, IRF4 acquires an association with CTCF, a feature maintained in plasma cell myeloma lines. Thus, shifting occupancy links IRF4 to both ABC and ASC gene expression, whereas BLIMP1 occupancy links to repression of the activation state.

Absolute Quantification of Transcription Factors Reveals Principles of Gene Regulation in Erythropoiesis

Dynamic cellular processes such as differentiation are driven by changes in the abundances of transcription factors (TFs). However, despite years of studies, our knowledge about the protein copy number of TFs in the nucleus is limited. Here, by determining the absolute abundances of 103 TFs and co-factors during the course of human erythropoiesis, we provide a dynamic and quantitative scale for TFs in the nucleus. Furthermore, we establish the first gene regulatory network of cell fate commitment that integrates temporal protein stoichiometry data with mRNA measurements. The model revealed quantitative imbalances in TFs' cross-antagonistic relationships that underlie lineage determination. Finally, we made the surprising discovery that, in the nucleus, co-repressors are dramatically more abundant than co-activators at the protein level, but not at the RNA level, with profound implications for understanding transcriptional regulation. These analyses provide a unique quantitative framework to understand transcriptional regulation of cell differentiation in a dynamic context.

Modeling binary and graded cone cell fate patterning in the mouse retina

Nervous systems are incredibly diverse, with myriad neuronal subtypes defined by gene expression. How binary and graded fate characteristics are patterned across tissues is poorly understood. Expression of opsin photopigments in the cone photoreceptors of the mouse retina provides an excellent model to address this question. Individual cones express S-opsin only, M-opsin only, or both S-opsin and M-opsin. These cell populations are patterned along the dorsal-ventral axis, with greater M-opsin expression in the dorsal region and greater S-opsin expression in the ventral region. Thyroid hormone signaling plays a critical role in activating M-opsin and repressing S-opsin. Here, we developed an image analysis approach to identify individual cone cells and evaluate their opsin expression from immunofluorescence imaging tiles spanning roughly 6 mm along the D-V axis of the mouse retina. From analyzing the opsin expression of ~250,000 cells, we found that cones make a binary decision between S-opsin only and co-expression competent fates. Co-expression competent cells express graded levels of S- and M-opsins, depending nonlinearly on their position in the dorsal-ventral axis. M- and S-opsin expression display differential, inverse patterns. Using these single-cell data, we developed a quantitative, probabilistic model of cone cell decisions in the retinal tissue based on thyroid hormone signaling activity. The model recovers the probability distribution for cone fate patterning in the mouse retina and describes a minimal set of interactions that are necessary to reproduce the observed cell fates. Our study provides a paradigm describing how differential responses to regulatory inputs generate complex patterns of binary and graded cell fates.

The development of a cell in a mammalian tissue is governed by a complex regulatory network that responds to many input signals to give the cell a distinct identity, a process referred to as cell-fate specification. Some of these cell fates have binary on-or-off gene expression patterns, while others have graded gene expression that changes across the tissue. Differentiation of the photoreceptor cells that sense light in the mouse retina provides a good example of this process. Here, we explore how complex patterns of cell fates are specified in the mouse retina by building a computational model based on analysis of a large number of photoreceptor cells from microscopy images of whole retinas. We use the data and the model to study what exactly it means for a cell to have a binary or graded cell fate and how these cell fates can be distinguished from each other. Our study shows how tens-of-thousands of individual photoreceptor cells can be patterned across a complex tissue by a regulatory network, creating a different outcome depending upon the received inputs.

Dynamic control of the T-cell specification gene regulatory network

Specification of multipotent blood precursor cells in postnatal mice to become committed T-cell precursors involves a gene regulatory network of several interacting but functionally distinct modules. Many links of this network have been defined by perturbation tests and by functional genomics. However, using the network model to predict real-life kinetics of the commitment process is still difficult, partly due to the tenacity of repressive chromatin states, and to the ability of transcription factors to affect each other's binding site choices through competitive recruitment to alternative sites ("coregulator theft"). To predict kinetics, future models will need to incorporate mechanistic information about chromatin state change dynamics and more sophisticated understanding of the proteomics and cooperative DNA site choices of transcription factor complexes.