Evolution Of Regulatory Networks Research Articles

Transcriptome and Metabolome Reveal Key Genes from the Plant Hormone Signal Transduction Pathway Regulating Plant Height and Leaf Size in Capsicum baccatum.

Plant structure-related agronomic traits like plant height and leaf size are critical for growth, development, and crop yield. Defining the types of genes involved in regulating plant structure size is essential for the molecular-assisted breeding of peppers. This research conducted comparative transcriptome analyses using Capsicum baccatum germplasm HNUCB0112 and HNUCB0222 and their F2 generation as materials. A total of 6574 differentially expressed genes (DEGs) were detected, which contain 379 differentially expressed transcription factors, mainly including transcription factor families such as TCP, WRKY, AUX/IAA, and MYB. Seven classes of DEGs were annotated in the plant hormone signal transduction pathway, including indole acetic acid (IAA), gibberellin (GA), cytokinin (CK), abscisic acid (ABA), jasmonic acid (JA), ethylene (ET), and salicylic acid (SA). The 26 modules were obtained by WGCNA analysis, and the MEpink module was positively correlated with plant height and leaf size, and hub genes associated with plant height and leaf size were anticipated. Differential genes were verified by qRT-PCR, which was consistent with the RNA-Seq results, demonstrating the accuracy of the sequencing results. These results enhance our understanding of the developmental regulatory networks governing pepper key traits like plant height and leaf size and offer new information for future research on the pepper plant architecture system.

Chromatin sensing: Integration of environmental signals to reprogram plant development through chromatin regulators.

Chromatin regulation in eukaryotes plays pivotal roles in controlling developmental regulatory gene network. This review explores the intricate interplay between chromatin regulators and environmental signals, elucidating their roles in shaping plant development. As sessile organisms, plants have evolved sophisticated mechanisms to perceive and respond to environmental cues, orchestrating developmental programs that ensure adaptability and survival. A central aspect of this dynamic response lies in the modulation of versatile gene regulatory networks, mediated in part by various chromatin regulators. Here, we summarized the current understanding of the molecular mechanisms through which chromatin regulators integrate environmental signals, influencing key aspects of plant development.

Dissecting the molecular basis of spike traits by integrating gene regulatory networks and genetic variation in wheat

Spike architecture influences both grain weight and grain number per spike, which are the two major components of grain yield in bread wheat (Triticum aestivum L.). However, the complex wheat genome and the influence of various environmental factors pose challenges in mapping the causal genes that affect spike traits. Here, we systematically identified genes involved in spike trait formation by integrating information on genomic variation and gene regulatory networks controlling young spike development in wheat. We identified 170 loci that are responsible for variations in spike length, spikelet number per spike, and grain number per spike through genome-wide association study and meta-QTL analyses. We constructed gene regulatory networks for young inflorescences at the double ridge stage and the floret primordium stage, in which the spikelet meristem and the floret meristem are predominant, respectively, by integrating transcriptome, histone modification, chromatin accessibility, eQTL, and protein–protein interactome data. From these networks, we identified 169 hub genes located in 76 of the 170 QTL regions whose polymorphisms are significantly associated with variation in spike traits. The functions of TaZF-B1, VRT-B2, and TaSPL15-A/D in establishment of wheat spike architecture were verified. This study provides valuable molecular resources for understanding spike traits and demonstrates that combining genetic analysis and developmental regulatory networks is a robust approach for dissection of complex traits.

An intronic LINE-1 regulates IFNAR1 expression in human immune cells

BackgroundDespite their origins as selfish parasitic sequences, some transposons in the human genome have been co-opted to serve as regulatory elements, contributing to the evolution of transcriptional networks. Most well-characterized examples of transposon-derived regulatory elements derive from endogenous retroviruses (ERVs), due to the intrinsic regulatory activity of proviral long terminal repeat regions. However, one subclass of transposable elements, the Long Interspersed Nuclear Elements (LINEs), have been largely overlooked in the search for functional regulatory transposons, and considered to be broadly epigenetically repressed.ResultsWe examined the chromatin state of LINEs by analyzing epigenomic data from human immune cells. Many LINEs are marked by the repressive H3K9me3 modification, but a subset exhibits evidence of enhancer activity in human immune cells despite also showing evidence of epigenetic repression. We hypothesized that these competing forces of repressive and activating epigenetic marks might lead to inducible enhancer activity. We investigated a specific L1M2a element located within the first intron of Interferon Alpha/Beta Receptor 1 (IFNAR1). This element shows epigenetic signatures of B cell-specific enhancer activity, despite being repressed by the Human Silencing Hub (HUSH) complex. CRISPR deletion of the element in B lymphoblastoid cells revealed that the element acts as an enhancer that regulates both steady state and interferon-inducible expression of IFNAR1.ConclusionsOur study experimentally demonstrates that an L1M2a element was co-opted to function as an interferon-inducible enhancer of IFNAR1, creating a feedback loop wherein IFNAR1 is transcriptionally upregulated by interferon signaling. This finding suggests that other LINEs may exhibit cryptic cell type-specific or context-dependent enhancer activity. LINEs have received less attention than ERVs in the effort to understand the contribution of transposons to the regulatory landscape of cellular genomes, but these are likely important, lineage-specific players in the rapid evolution of immune system regulatory networks and deserve further study.

Transcription factor expression levels and environmental signals constrain transcription factor innovation.

Evolutionary innovation of transcription factors frequently drives phenotypic diversification and adaptation to environmental change. Transcription factors can gain or lose connections to target genes, resulting in novel regulatory responses and phenotypes. However the frequency of functional adaptation varies between different regulators, even when they are closely related. To identify factors influencing propensity for innovation, we utilise a Pseudomonas fluorescens SBW25 strain rendered incapable of flagellar mediated motility in soft-agar plates via deletion of the flagellar master regulator (fleQ). This bacterium can evolve to rescue flagellar motility via gene regulatory network rewiring of an alternative transcription factor to rescue activity of FleQ. Previously, we have identified two members (out of 22) of the RpoN-dependent enhancer binding protein (RpoN-EBP) family of transcription factors (NtrC and PFLU1132) that are capable of innovating in this way. These two transcription factors rescue motility repeatably and reliably in a strict hierarchy – with NtrC the only route in a ∆fleQ background, and PFLU1132 the only route in a ∆fleQ∆ntrC background. However, why other members in the same transcription factor family have not been observed to rescue flagellar activity is unclear. Previous work shows that protein homology cannot explain this pattern within the protein family (RpoN-EBPs), and mutations in strains that rescued motility suggested high levels of transcription factor expression and activation drive innovation. We predict that mutations that increase expression of the transcription factor are vital to unlock evolutionary potential for innovation. Here, we construct titratable expression mutant lines for 11 of the RpoN-EBPs in P. fluorescens . We show that in five additional RpoN-EBPs (FleR, HbcR, GcsR, DctD, AauR and PFLU2209), high expression levels result in different mutations conferring motility rescue, suggesting alternative rewiring pathways. Our results indicate that expression levels (and not protein homology) of RpoN-EBPs are a key constraining factor in determining evolutionary potential for innovation. This suggests that transcription factors that can achieve high expression through few mutational changes, or transcription factors that are active in the selective environment, are more likely to innovate and contribute to adaptive gene regulatory network evolution.

Immune landscape in rejection of renal transplantation revealed by high-throughput single-cell RNA sequencing.

Background: The role of the cellular level in kidney transplant rejection is unclear, and single-cell RNA sequencing (scRNA-seq) can reveal the single-cell landscape behind rejection of human kidney allografts at the single-cell level. Methods: High-quality transcriptomes were generated from scRNA-seq data from five human kidney transplantation biopsy cores. Cluster analysis was performed on the scRNA-seq data by known cell marker genes in order to identify different cell types. In addition, pathways, pseudotime developmental trajectories and transcriptional regulatory networks involved in different cell subpopulations were explored. Next, we systematically analyzed the scoring of gene sets regarding single-cell expression profiles based on biological processes associated with oxidative stress. Results: We obtained 81,139 single cells by scRNA-seq from kidney transplant tissue biopsies of three antibody-mediated rejection (ABMR) patients and two acute kidney injury (AKI) patients with non-rejection causes and identified 11 cell types, including immune cells, renal cells and several stromal cells. Immune cells such as macrophages showed inflammatory activation and antigen presentation and complement signaling, especially in rejection where some subpopulations of cells specifically expressed in rejection showed specific pro-inflammatory responses. In addition, patients with rejection are characterized by an increased number of fibroblasts, and further analysis of subpopulations of fibroblasts revealed their involvement in inflammatory and fibrosis-related pathways leading to increased renal rejection and fibrosis. Notably, the gene set score for response to oxidative stress was higher in patients with rejection. Conclusion: Insight into histological differences in kidney transplant patients with or without rejection was gained by assessing differences in cellular levels at single-cell resolution. In conclusion, we applied scRNA-seq to rejection after renal transplantation to deconstruct its heterogeneity and identify new targets for personalized therapeutic approaches.

Detection of TALEN-mediated genome cleavage during the early embryonic stage of the starfish Patiria pectinifera.

Echinoderms have long been utilized as experimental materials to study the genetic control of developmental processes and their evolution. Among echinoderms, the molecular study of starfish embryos has received considerable attention across research topics such as gene regulatory network evolution and larval regeneration. Recently, experimental techniques to manipulate gene functions have been gradually established in starfish as the feasibility of genome editing methods was reported. However, it is still unclear when these techniques cause genome cleavage during the development of starfish, which is critical to understand the timeframe and applicability of the experiment during early development of starfish. We herein reported that gene functions can be analyzed by the genome editing method TALEN in early embryos, such as the blastula of the starfish Patiria pectinifera. We injected the mRNA of TALEN targeting rar, which was previously constructed, into eggs of P. pectinifera and examined the efficiency of genome cleavage through developmental stages from 6 to 48 hours post fertilization. The results will be key knowledge not only when designing TALEN-based experiments but also when assessing the results.

CastNet: a systems-level sequence evolution simulator

BackgroundSimulating DNA evolution has been done through coevolution-agnostic probabilistic frameworks for the past 3 decades. The most common implementation is by using the converse of the probabilistic approach used to infer phylogenies which, in the simplest form, simulates a single sequence at a time. However, biological systems are multi-genic, and gene products can affect each other’s evolutionary paths through coevolution. These crucial evolutionary dynamics still remain to be simulated, and we believe that modelling them can lead to profound insights for comparative genomics.ResultsHere we present CastNet, a genome evolution simulator that assumes each genome is a collection of genes with constantly evolving regulatory interactions in between them. The regulatory interactions produce a phenotype in the form of gene expression profiles, upon which fitness is calculated. A genetic algorithm is then used to evolve a population of such entities through a user-defined phylogeny. Importantly, the regulatory mutations are a response to sequence mutations, thus making a 1–1 relationship between the rate of evolution of sequences and of regulatory parameters. This is, to our knowledge, the first time the evolution of sequences and regulation have been explicitly linked in a simulation, despite there being a multitude of sequence evolution simulators, and a handful of models to simulate Gene Regulatory Network (GRN) evolution. In our test runs, we see a coevolutionary signal among genes that are active in the GRN, and neutral evolution in genes that are not included in the network, showing that selective pressures imposed on the regulatory output of the genes are reflected in their sequences.ConclusionWe believe that CastNet represents a substantial step for developing new tools to study genome evolution, and more broadly, coevolutionary webs and complex evolving systems. This simulator also provides a new framework to study molecular evolution where sequence coevolution has a leading role.

A single-cell trajectory atlas of striatal development

The striatum integrates dense neuromodulatory inputs from many brain regions to coordinate complex behaviors. This integration relies on the coordinated responses from distinct striatal cell types. While previous studies have characterized the cellular and molecular composition of the striatum using single-cell RNA-sequencing at distinct developmental timepoints, the molecular changes spanning embryonic through postnatal development at the single-cell level have not been examined. Here, we combine published mouse striatal single-cell datasets from both embryonic and postnatal timepoints to analyze the developmental trajectory patterns and transcription factor regulatory networks within striatal cell types. Using this integrated dataset, we found that dopamine receptor-1 expressing spiny projection neurons have an extended period of transcriptional dynamics and greater transcriptional complexity over postnatal development compared to dopamine receptor-2 expressing neurons. Moreover, we found the transcription factor, FOXP1, exerts indirect changes to oligodendrocytes. These data can be accessed and further analyzed through an interactive website (https://mouse-striatal-dev.cells.ucsc.edu).

Development's Inaugural Outstanding Paper Prize.

Development's Inaugural Outstanding Paper Prize.

A Reduction of Transcriptional Regulation in Aquatic Oligotrophic Microorganisms Enhances Fitness in Nutrient-Poor Environments.

In this review, we consider the regulatory strategies of aquatic oligotrophs, microbial cells that are adapted to thrive under low-nutrient concentrations in oceans, lakes, and other aquatic ecosystems. Many reports have concluded that oligotrophs use less transcriptional regulation than copiotrophic cells, which are adapted to high nutrient concentrations and are far more common subjects for laboratory investigations of regulation. It is theorized that oligotrophs have retained alternate mechanisms of regulation, such as riboswitches, that provide shorter response times and smaller amplitude responses and require fewer cellular resources. We examine the accumulated evidence for distinctive regulatory strategies in oligotrophs. We explore differences in the selective pressures copiotrophs and oligotrophs encounter and ask why, although evolutionary history gives copiotrophs and oligotrophs access to the same regulatory mechanisms, they might exhibit distinctly different patterns in how these mechanisms are used. We discuss the implications of these findings for understanding broad patterns in the evolution of microbial regulatory networks and their relationships to environmental niche and life history strategy. We ask whether these observations, which have emerged from a decade of increased investigation of the cell biology of oligotrophs, might be relevant to recent discoveries of many microbial cell lineages in nature that share with oligotrophs the property of reduced genome size.

Uncovering the involvement of DoDELLA1-interacting proteins in development by characterizing the DoDELLA gene family in Dendrobium officinale

BackgroundGibberellins (GAs) are widely involved in plant growth and development. DELLA proteins are key regulators of plant development and a negative regulatory factor of GA. Dendrobium officinale is a valuable traditional Chinese medicine, but little is known about D. officinale DELLA proteins. Assessing the function of D. officinale DELLA proteins would provide an understanding of their roles in this orchid’s development.ResultsIn this study, the D. officinale DELLA gene family was identified. The function of DoDELLA1 was analyzed in detail. qRT-PCR analysis showed that the expression levels of all DoDELLA genes were significantly up-regulated in multiple shoots and GA3-treated leaves. DoDELLA1 and DoDELLA3 were significantly up-regulated in response to salt stress but were significantly down-regulated under drought stress. DoDELLA1 was localized in the nucleus. A strong interaction was observed between DoDELLA1 and DoMYB39 or DoMYB308, but a weak interaction with DoWAT1.ConclusionsIn D. officinale, a developmental regulatory network involves a close link between DELLA and other key proteins in this orchid’s life cycle. DELLA plays a crucial role in D. officinale development.

Single-cell transcriptomic analysis reveals the developmental trajectory and transcriptional regulatory networks of pigment glands in Gossypium bickii

Comprehensive utilization of cottonseeds is limited by the presence of pigment glands and its inclusion gossypol. The ideal cotton has glandless seeds but a glanded plant, a trait found in only a few Australian wild cotton species, including Gossypium bickii. Introgression of this trait into cultivated species has proved to be difficult. Understanding the biological processes toward pigment gland morphogenesis and the associated underlying molecular mechanisms will facilitate breeding of cultivated cotton varieties with the trait of glandless seeds and glanded plant. In this study, single-cell RNA sequencing (scRNA-seq) was performed on 12 222 protoplasts isolated from cotyledons of germinating G. bickii seeds 48 h after imbibition. Clustered into 14 distinct clusters unsupervisedly, these cells could be grouped into eight cell populations with the assistance of known cell marker genes. The pigment gland cells were well separated from others and could be separated into pigment gland parenchyma cells, secretory cells, and apoptotic cells. By integrating the pigment gland cell developmental trajectory, transcription factor regulatory networks, and core transcription factor functional validation, we established a model for pigment gland formation. In this model, light and gibberellin were verified to promote the formation of pigment glands. In addition, three novel genes, GbiERF114 (ETHYLENE RESPONSE FACTOR 114), GbiZAT11 (ZINC FINGER OF ARABIDOPSIS THALIANA 11), and GbiNTL9 (NAC TRANSCRIPTION FACTOR-LIKE 9), were found to affect pigment gland formation. Collectively, these findings provide new insights into pigment gland morphogenesis and lay the cornerstone for future cotton scRNA-seq investigations.

Inferring cell developmental stage-specific lncRNA regulation in the developing human neocortex with CDSlncR.

Noncoding RNAs (ncRNAs) occupy ~98% of the transcriptome in human, and are usually not translated into proteins. Among ncRNAs, long non-coding RNAs (lncRNAs, >200 nucleotides) are important regulators to modulate gene expression, and are involved in many biological processes (e.g., cell development). To study lncRNA regulation, many computational approaches or tools have been proposed by using bulk transcriptomics data. Nevertheless, previous bulk data-driven methods are mostly limited to explore the lncRNA regulation regarding all of cells, instead of the lncRNA regulation specific to cell developmental stages. Fortunately, recent advance in single-cell sequencing data has provided a way to investigate cell developmental stage-specific lncRNA regulation. In this work, we present a novel computational method, CDSlncR (Cell Developmental Stage-specific lncRNA regulation), which combines putative lncRNA-target binding information with single-cell transcriptomics data to infer cell developmental stage-specific lncRNA regulation. For each cell developmental stage, CDSlncR constructs a cell developmental stage-specific lncRNA regulatory network in the cell developmental stage. To illustrate the effectiveness of CDSlncR, we apply CDSlncR into single-cell transcriptomics data of the developing human neocortex for exploring lncRNA regulation across different human neocortex developmental stages. Network analysis shows that the lncRNA regulation is unique in each developmental stage of human neocortex. As a case study, we also perform particular analysis on the cell developmental stage-specific lncRNA regulation related to 18 known lncRNA biomarkers in autism spectrum disorder. Finally, the comparison result indicates that CDSlncR is an effective method for predicting cell developmental stage-specific lncRNA targets. CDSlncR is available at https://github.com/linxi159/CDSlncR.

PI3K signaling specifies proximal-distal fate by driving a developmental gene regulatory network in SOX9+ mouse lung progenitors.

The tips of the developing respiratory buds are home to important progenitor cells marked by the expression of SOX9 and ID2. Early in embryonic development (prior to E13.5), SOX9+progenitors are multipotent, generating both airway and alveolar epithelium, but are selective progenitors of alveolar epithelial cells later in development. Transcription factors, including Sox9, Etv5, Irx, Mycn, and Foxp1/2 interact in complex gene regulatory networks to control proliferation and differentiation of SOX9+progenitors. Molecular mechanisms by which these transcription factors and other signaling pathways control chromatin state to establish and maintain cell-type identity are not well-defined. Herein, we analyze paired gene expression (RNA-Seq) and chromatin accessibility (ATAC-Seq) data from SOX9+ epithelial progenitor cells (EPCs) during embryonic development in Mus musculus. Widespread changes in chromatin accessibility were observed between E11.5 and E16.5, particularly at distal cis-regulatory elements (e.g. enhancers). Gene regulatory network (GRN) inference identified a common SOX9+ progenitor GRN, implicating phosphoinositide 3-kinase (PI3K) signaling in the developmental regulation of SOX9+ progenitor cells. Consistent with this model, conditional ablation of PI3K signaling in the developing lung epithelium in mouse resulted in an expansion of the SOX9+ EPC population and impaired airway epithelial cell differentiation. These data demonstrate that PI3K signaling is required for epithelial patterning during lung organogenesis, and emphasize the combinatorial power of paired RNA and ATAC seq in defining regulatory networks in development.

Evolution of miRNA-Binding Sites and Regulatory Networks in Cichlids.

The divergence of regulatory regions and gene regulatory network (GRN) rewiring is a key driver of cichlid phenotypic diversity. However, the contribution of miRNA-binding site turnover has yet to be linked to GRN evolution across cichlids. Here, we extend our previous studies by analyzing the selective constraints driving evolution of miRNA and transcription factor (TF)–binding sites of target genes, to infer instances of cichlid GRN rewiring associated with regulatory binding site turnover. Comparative analyses identified increased species-specific networks that are functionally associated to traits of cichlid phenotypic diversity. The evolutionary rewiring is associated with differential models of miRNA- and TF-binding site turnover, driven by a high proportion of fast-evolving polymorphic sites in adaptive trait genes compared with subsets of random genes. Positive selection acting upon discrete mutations in these regulatory regions is likely to be an important mechanism in rewiring GRNs in rapidly radiating cichlids. Regulatory variants of functionally associated miRNA- and TF-binding sites of visual opsin genes differentially segregate according to phylogeny and ecology of Lake Malawi species, identifying both rewired, for example, clade-specific and conserved network motifs of adaptive trait associated GRNs. Our approach revealed several novel candidate regulators, regulatory regions, and three-node motifs across cichlid genomes with previously reported associations to known adaptive evolutionary traits.

Snake venom gene expression is coordinated by novel regulatory architecture and the integration of multiple co-opted vertebrate pathways.

Understanding how regulatory mechanisms evolve is critical for understanding the processes that give rise to novel phenotypes. Snake venom systems represent a valuable and tractable model for testing hypotheses related to the evolution of novel regulatory networks, yet the regulatory mechanisms underlying venom production remain poorly understood. Here, we use functional genomics approaches to investigate venom regulatory architecture in the prairie rattlesnake and identify cis-regulatory sequences (enhancers and promoters), trans-regulatory transcription factors, and integrated signaling cascades involved in the regulation of snake venom genes. We find evidence that two conserved vertebrate pathways, the extracellular signal-regulated kinase and unfolded protein response pathways, were co-opted to regulate snake venom. In one large venom gene family (snake venom serine proteases), this co-option was likely facilitated by the activity of transposable elements. Patterns of snake venom gene enhancer conservation, in some cases spanning 50 million yr of lineage divergence, highlight early origins and subsequent lineage-specific adaptations that have accompanied the evolution of venom regulatory architecture. We also identify features of chromatin structure involved in venom regulation, including topologically associated domains and CTCF loops that underscore the potential importance of novel chromatin structure to coevolve when duplicated genes evolve new regulatory control. Our findings provide a model for understanding how novel regulatory systems may evolve through a combination of genomic processes, including tandem duplication of genes and regulatory sequences, cis-regulatory sequence seeding by transposable elements, and diverse transcriptional regulatory proteins controlled by a co-opted regulatory cascade.

Omics Studies for the Identification of Ascidian Peptides, Cognate Receptors, and Their Relevant Roles in Ovarian Follicular Development.

Omics studies contribute to the elucidation of genomes and profiles of gene expression. In the ascidian Ciona intestinalis Type A (Ciona robusta), mass spectrometry (MS)-based peptidomic studies have detected numerous Ciona-specific (nonhomologous) neuropeptides as well as Ciona homologs of typical vertebrate neuropeptides and hypothalamic peptide hormones. Candidates for cognate G protein-coupled receptors (GPCRs) for these peptides have been found in the Ciona transcriptome by two ways. First, Ciona homologous GPCRs of vertebrate counterparts have been detected by sequence homology searches of cognate transcriptomes. Second, the transcriptome-derived GPCR candidates have been used for machine learning-based systematic prediction of interactions not only between Ciona homologous peptides and GPCRs but also between novel Ciona peptides and GPCRs. These data have ultimately led to experimental evidence for various Ciona peptide-GPCR interactions. Comparative transcriptomics between the wildtype and Ciona vasopressin (CiVP) gene-edited Ciona provide clues to the biological functions of CiVP in ovarian follicular development and whole body growth. Furthermore, the transcriptomes of follicles treated with peptides, such as Ciona tachykinin and cionin (a Ciona cholecystokinin homolog), have revealed key regulatory genes for Ciona follicle growth, maturation, and ovulation, eventually leading to the verification of essential and novel molecular mechanisms underlying these biological events. These findings indicate that omics studies, combined with artificial intelligence and single-cell technologies, pave the way for investigating in greater details the nervous, neuroendocrine, and endocrine systems of ascidians and the molecular and functional evolution and diversity of peptidergic regulatory networks throughout chordates.

Common Themes and Future Challenges in Understanding Gene Regulatory Network Evolution.

A major driving force behind the evolution of species-specific traits and novel structures is alterations in gene regulatory networks (GRNs). Comprehending evolution therefore requires an understanding of the nature of changes in GRN structure and the responsible mechanisms. Here, we review two insect pigmentation GRNs in order to examine common themes in GRN evolution and to reveal some of the challenges associated with investigating changes in GRNs across different evolutionary distances at the molecular level. The pigmentation GRN in Drosophila melanogaster and other drosophilids is a well-defined network for which studies from closely related species illuminate the different ways co-option of regulators can occur. The pigmentation GRN for butterflies of the Heliconius species group is less fully detailed but it is emerging as a useful model for exploring important questions about redundancy and modularity in cis-regulatory systems. Both GRNs serve to highlight the ways in which redeployment of trans-acting factors can lead to GRN rewiring and network co-option. To gain insight into GRN evolution, we discuss the importance of defining GRN architecture at multiple levels both within and between species and of utilizing a range of complementary approaches.

Enabling Studies of Genome-Scale Regulatory Network Evolution in Large Phylogenies with MRTLE.

Transcriptional regulatory networks specify context-specific patterns of genes and play a central role in how species evolve and adapt. Inferring genome-scale regulatory networks in non-model species is the first step for examining patterns of conservation and divergence of regulatory networks. Transcriptomic data obtained under varying environmental stimuli in multiple species are becoming increasingly available, which can be used to infer regulatory networks. However, inference and analysis of multiple gene regulatory networks in a phylogenetic setting remains challenging. We developed an algorithm, Multi-species Regulatory neTwork LEarning (MRTLE), to facilitate such studies of regulatory network evolution. MRTLE is a probabilistic graphical model-based algorithm that uses phylogenetic structure, transcriptomic data for multiple species, and sequence-specific motifs in each species to simultaneously infer genome-scale regulatory networks across multiple species. We applied MRTLE to study regulatory network evolution across six ascomycete yeasts using transcriptomic measurements collected across different stress conditions. MRTLE networks recapitulated experimentally derived interactions in the model organism S. cerevisiae as well as non-model species, and it was more beneficial for network inference than methods that do not use phylogenetic information. We examined the regulatory networks across species and found that regulators associated with significant expression and network changes are involved in stress-related processes. MTRLE and its associated downstream analysis provide a scalable and principled framework to examine evolutionary dynamics of transcriptional regulatory networks across multiple species in a large phylogeny.