An experimentally supported model of the Bacillus subtilis global transcriptional regulatory network

Augusta: From RNA-Seq to gene regulatory networks and Boolean models

Computational models of gene regulations help to understand regulatory mechanisms and are extensively used in a wide range of areas, e.g., biotechnology or medicine, with significant benefits. Unfortunately, there are only a few computational gene regulatory models of whole genomes allowing static and dynamic analysis due to the lack of sophisticated tools for their reconstruction. Here, we describe Augusta, an open-source Python package for Gene Regulatory Network (GRN) and Boolean Network (BN) inference from the high-throughput gene expression data. Augusta can reconstruct genome-wide models suitable for static and dynamic analyses. Augusta uses a unique approach where the first estimation of a GRN inferred from expression data is further refined by predicting transcription factor binding motifs in promoters of regulated genes and by incorporating verified interactions obtained from databases. Moreover, a refined GRN is transformed into a draft BN by searching in the curated model database and setting logical rules to incoming edges of target genes, which can be further manually edited as the model is provided in the SBML file format. The approach is applicable even if information about the organism under study is not available in the databases, which is typically the case for non-model organisms including most microbes. Augusta can be operated from the command line and, thus, is easy to use for automated prediction of models for various genomes. The Augusta package is freely available at github.com/JanaMus/Augusta. Documentation and tutorials are available at augusta.readthedocs.io.

Transcriptome data are insufficient to control false discoveries in regulatory network inference

Inference of causal transcriptional regulatory networks (TRNs) from transcriptomic data suffers notoriously from false positives. Approaches to control the false discovery rate (FDR), for example, via permutation, bootstrapping, or multivariate Gaussian distributions, suffer from several complications: difficulty in distinguishing direct from indirect regulation, nonlinear effects, and causal structure inference requiring "causal sufficiency," meaning experiments that are free of any unmeasured, confounding variables. Here, we use a recently developed statistical framework, model-X knockoffs, to control the FDR while accounting for indirect effects, nonlinear dose-response, and user-provided covariates. We adjust the procedure to estimate the FDR correctly even when measured against incomplete gold standards. However, benchmarking against chromatin immunoprecipitation (ChIP) and other gold standards reveals higher observed than reported FDR. This indicates that unmeasured confounding is a major driver of FDR in TRN inference. A record of this paper's transparent peer review process is included in the supplemental information.

Optimizing data integration improves gene regulatory network inference in Arabidopsis thaliana

MOTIVATIONS

Gene regulatory networks (GRNs) are traditionally inferred from gene expression profiles monitoring a specific condition or treatment. In the last decade, integrative strategies have successfully emerged to guide GRN inference from gene expression with complementary prior data. However, datasets used as prior information and validation gold standards are often related and limited to a subset of genes. This lack of complete and independent evaluation calls for new criteria to robustly estimate the optimal intensity of prior data integration in the inference process.

RESULTS

We address this issue for two regression-based GRN inference models, a weighted random forest (weigthedRF) and a generalized linear model estimated under a weighted LASSO penalty with stability selection (weightedLASSO). These approaches are applied to data from the root response to nitrate induction in Arabidopsis thaliana. For each gene, we measure how the integration of transcription factor binding motifs influences model prediction. We propose a new approach, DIOgene, that uses model prediction error and a simulated null hypothesis in order to optimize data integration strength in a hypothesis-driven, gene-specific manner. This integration scheme reveals a strong diversity of optimal integration intensities between genes, and offers good performance in minimizing prediction error as well as retrieving experimental interactions. Experimental results show that DIOgene compares favorably against state-of-the-art approaches and allows to recover master regulators of nitrate induction.

AVAILABILITY AND IMPLEMENTATION

The R code and notebooks demonstrating the use of the proposed approaches are available in the repository https://github.com/OceaneCsn/integrative_GRN_N_induction.

Regulation of late-acting operons by three transcription factors and a CRISPR-Cas component during Myxococcus xanthus development

Upon starvation, rod-shaped Myxococcus xanthus bacteria form mounds and then differentiate into round, stress-resistant spores. Little is known about the regulation of late-acting operons important for spore formation. C-signaling has been proposed to activate FruA, which binds DNA cooperatively with MrpC to stimulate transcription of developmental genes. We report that this model can explain regulation of the fadIJ operon involved in spore metabolism, but not that of the spore coat biogenesis operons exoA-I, exoL-P, and nfsA-H. Rather, a mutation in fruA increased the transcript levels from these operons early in development, suggesting negative regulation by FruA, and a mutation in mrpC affected transcript levels from each operon differently. FruA bound to all four promoter regions in vitro, but strikingly each promoter region was unique in terms of whether or not MrpC and/or the DNA-binding domain of Nla6 bound, and in terms of cooperative binding. Furthermore, the DevI component of a CRISPR-Cas system is a negative regulator of all four operons, based on transcript measurements. Our results demonstrate complex regulation of sporulation genes by three transcription factors and a CRISPR-Cas component, which we propose produces spores suited to withstand starvation and environmental insults.

Regulator DegU can remarkably influence alkaline protease AprE biosynthesis in Bacillus licheniformis 2709

Alkaline protease AprE, produced by Bacillus licheniformis 2709 is an important edible hydrolase, which has potential applications in nutrient acquisition and medicine. The expression of AprE is finely regulated by a complex transcriptional regulation system. However, there is little study on transcriptional regulation mechanism of AprE biosynthesis in Bacillus licheniformis, which limits system engineering and further enhancement of AprE. Here, the severely depressed expression of aprE in degU and degS deletion mutants illustrated that the regulator DegU and its phosphorylation played a crucial part in AprE biosynthesis. Further electrophoretic mobility shift assay (EMSA) in vitro indicated that phosphorylated DegU can directly bind to the regulatory region though the DNase I foot-printing experiments failed to observe protected region. The plasmid-mediated overexpression of degU32 (Hy) obviously improved the yield of AprE by 41.6 % compared with the control strain, which demonstrated the importance of phosphorylation state of DegU on the transcription of aprE in vivo. In this study, the putative binding sequence of aprE (5'-TAAAT……AAAAT…….AACAT…TAAAA-3') located upstream -91 to -87 bp, -101 to -97 bp, -195 to -191 bp, -215 to -211 bp of the transcription start site (TSS) in B. licheniformis was computationally identified based on the DNA-binding sites of DegU in Bacillus subtilis. Overall, we systematically investigated the influence of the interplay between phosphorylated DegU and its cognate DNA sequence on expression of aprE, which not only contributes to the further AprE high-production in a genetically modified host in the future, but also significantly increases our understanding of the aprE transcription mechanism.

PMF-GRN: a variational inference approach to single-cell gene regulatory network inference using probabilistic matrix factorization

Inferring gene regulatory networks (GRNs) from single-cell data is challenging due to heuristic limitations. Existing methods also lack estimates of uncertainty. Here we present Probabilistic Matrix Factorization for Gene Regulatory Network Inference (PMF-GRN). Using single-cell expression data, PMF-GRN infers latent factors capturing transcription factor activity and regulatory relationships. Using variational inference allows hyperparameter search for principled model selection and direct comparison to other generative models. We extensively test and benchmark our method using real single-cell datasets and synthetic data. We show that PMF-GRN infers GRNs more accurately than current state-of-the-art single-cell GRN inference methods, offering well-calibrated uncertainty estimates.

Spore germination: Two ion channels are better than one

Germination is the process by which spores emerge from dormancy. Although spores can remain dormant for decades, the study of germination is an active field of research. In this issue of Genes & Development, Gao and colleagues (pp. 31-45) address a perplexing question: How can a dormant spore initiate germination in response to environmental cues? Three distinct complexes are involved: GerA, a germinant-gated ion channel; 5AF/FigP, a second ion channel required for amplification; and SpoVA, a channel for dipicolinic acid (DPA). DPA release is followed by rehydration of the spore core, thus allowing the resumption of metabolic activity.

SpoVAF and FigP assemble into oligomeric ion channels that enhance spore germination

Bacterial spores can remain dormant for decades yet rapidly germinate and resume growth in response to nutrients. GerA family receptors that sense and respond to these signals have recently been shown to oligomerize into nutrient-gated ion channels. Ion release initiates exit from dormancy. Here, we report that a distinct ion channel, composed of SpoVAF (5AF) and its newly discovered partner protein, YqhR (FigP), amplifies the response. At high germinant concentrations, 5AF/FigP accelerate germination; at low concentrations, this complex becomes critical for exit from dormancy. 5AF is homologous to the channel-forming subunit of GerA family receptors and is predicted to oligomerize around a central pore. 5AF mutations predicted to widen the channel cause constitutive germination during spore formation and membrane depolarization in vegetative cells. Narrow-channel mutants are impaired in germination. A screen for suppressors of a constitutively germinating 5AF mutant identified FigP as an essential cofactor of 5AF activity. We demonstrate that 5AF and FigP interact and colocalize with GerA family receptors in spores. Finally, we show that 5AF/FigP accelerate germination in B. subtilis spores that have nutrient receptors from another species. Our data support a model in which nutrient-triggered ion release by GerA family receptors activates 5AF/FigP ion release, amplifying the response to germinant signals.

Structure-primed embedding on the transcription factor manifold enables transparent model architectures for gene regulatory network and latent activity inference

BACKGROUND

Modeling of gene regulatory networks (GRNs) is limited due to a lack of direct measurements of genome-wide transcription factor activity (TFA) making it difficult to separate covariance and regulatory interactions. Inference of regulatory interactions and TFA requires aggregation of complementary evidence. Estimating TFA explicitly is problematic as it disconnects GRN inference and TFA estimation and is unable to account for, for example, contextual transcription factor-transcription factor interactions, and other higher order features. Deep-learning offers a potential solution, as it can model complex interactions and higher-order latent features, although does not provide interpretable models and latent features.

RESULTS

We propose a novel autoencoder-based framework, StrUcture Primed Inference of Regulation using latent Factor ACTivity (SupirFactor) for modeling, and a metric, explained relative variance (ERV), for interpretation of GRNs. We evaluate SupirFactor with ERV in a wide set of contexts. Compared to current state-of-the-art GRN inference methods, SupirFactor performs favorably. We evaluate latent feature activity as an estimate of TFA and biological function in S. cerevisiae as well as in peripheral blood mononuclear cells (PBMC).

CONCLUSION

Here we present a framework for structure-primed inference and interpretation of GRNs, SupirFactor, demonstrating interpretability using ERV in multiple biological and experimental settings. SupirFactor enables TFA estimation and pathway analysis using latent factor activity, demonstrated here on two large-scale single-cell datasets, modeling S. cerevisiae and PBMC. We find that the SupirFactor model facilitates biological analysis acquiring novel functional and regulatory insight.

Identification and engineering of the aprE regulatory region and relevant regulatory proteins in Bacillus licheniformis 2709

Bacillus licheniformis 2709 is the main industrial producer of alkaline protease (AprE), but its biosynthesis is strictly controlled by a highly sophisticated transcriptional network. In this study, the UP elements of aprE located 74-98, 98-119 and 140-340 bp upstream of the transcriptional start site (TSS) were identified, which presented obvious effects on the transcription of aprE. To further analyze the transcriptional mechanism, the specific proteins binding to the approximately 500-bp DNA sequences were subsequently captured by reverse-chromatin immunoprecipitation (reverse-ChIP) and DNA pull-down (DPD) assays, which captured the transcriptional factors CggR, FruR, and YhcZ. The study demonstrated that CggR, FruR and YhcZ had no significant effect on cell growth and aprE expression. Then, aprE expression was significantly enhanced by deleting a potential negative regulatory factor binding site in the genome. The AprE enzyme activity in shake flasks of the genomic mutant BL ∆1 was 47% higher than in the original strain, while the aprE transcription level increased 3.16 times. The protocol established in this study provides a valuable reference for the high-level production of proteins in other Bacillus species. At the same time, it will help reveal the molecular mechanism of the transcriptional regulatory network of aprE and provide important theoretical guidance for further enhancing the yield of AprE.

Multiomics atlas-assisted discovery of transcription factors enables specific cell state programming

The same types of cells can assume diverse states with varying functionalities. Effective cell therapy can be achieved by specifically driving a desirable cell state, which requires the elucidation of key transcription factors (TFs). Here, we integrated epigenomic and transcriptomic data at the systems level to identify TFs that define different CD8 + T cell states in an unbiased manner. These TF profiles can be used for cell state programming that aims to maximize the therapeutic potential of T cells. For example, T cells can be programmed to avoid a terminal exhaustion state (Tex Term ), a dysfunctional T cell state that is often found in tumors or chronic infections. However, Tex Term exhibits high similarity with the beneficial tissue-resident memory T states (T RM ) in terms of their locations and transcription profiles. Our bioinformatic analysis predicted Zscan20 , a novel TF, to be uniquely active in Tex Term . Consistently, Zscan20 knock-out thwarted the differentiation of Tex Term in vivo , but not that of T RM . Furthermore, perturbation of Zscan20 programs T cells into an effector-like state that confers superior tumor and virus control and synergizes with immune checkpoint therapy. We also identified Jdp2 and Nfil3 as powerful Tex Term drivers. In short, our multiomics-based approach discovered novel TFs that enhance anti-tumor immunity, and enable highly effective cell state programming.

One sentence summary

Multiomics atlas enables the systematic identification of cell-state specifying transcription factors for therapeutic cell state programming.

Bacillus subtilis NrnB is expressed during sporulation and acts as a unique 3'-5' exonuclease

All cells employ a combination of endo- and exoribonucleases to degrade long RNA polymers to fragments 2-5 nucleotides in length. These short RNA fragments are processed to monoribonucleotides by nanoRNases. Genetic depletion of nanoRNases has been shown to increase abundance of short RNAs. This deleteriously affects viability, virulence, and fitness, indicating that short RNAs are a metabolic burden. Previously, we provided evidence that NrnA is the housekeeping nanoRNase for Bacillus subtilis. Herein, we investigate the biological and biochemical functions of the evolutionarily related protein, B. subtilis NrnB (NrnBBs). These experiments show that NrnB is surprisingly different from NrnA. While NrnA acts at the 5' terminus of RNA substrates, NrnB acts at the 3' terminus. Additionally, NrnA is expressed constitutively under standard growth conditions, yet NrnB is selectively expressed during endospore formation. Furthermore, NrnA processes only short RNAs, while NrnB unexpectedly processes both short RNAs and longer RNAs. Indeed, inducible expression of NrnB can even complement the loss of the known global 3'-5' exoribonucleases, indicating that it acts as a general exonuclease. Together, these data demonstrate that NrnB proteins, which are widely found in Firmicutes, Epsilonproteobacteria and Archaea, are fundamentally different than NrnA proteins and may be used for specialized purposes.

Simultaneous estimation of gene regulatory network structure and RNA kinetics from single cell gene expression

Cells respond to environmental and developmental stimuli by remodeling their transcriptomes through regulation of both mRNA transcription and mRNA decay. A central goal of biology is identifying the global set of regulatory relationships between factors that control mRNA production and degradation and their target transcripts and construct a predictive model of gene expression. Regulatory relationships are typically identified using transcriptome measurements and causal inference algorithms. RNA kinetic parameters are determined experimentally by employing run-on or metabolic labeling (e.g. 4-thiouracil) methods that allow transcription and decay rates to be separately measured. Here, we develop a deep learning model, trained with single-cell RNA-seq data, that both infers causal regulatory relationships and estimates RNA kinetic parameters. The resulting in silico model predicts future gene expression states and can be perturbed to simulate the effect of transcription factor changes. We acquired model training data by sequencing the transcriptomes of 175,000 individual Saccharomyces cerevisiae cells that were subject to an external perturbation and continuously sampled over a one hour period. The rate of change for each transcript was calculated on a per-cell basis to estimate RNA velocity. We then trained a deep learning model with transcriptome and RNA velocity data to calculate time-dependent estimates of mRNA production and decay rates. By separating RNA velocity into transcription and decay rates, we show that rapamycin treatment causes existing ribosomal protein transcripts to be rapidly destabilized, while production of new transcripts gradually slows over the course of an hour. The neural network framework we present is designed to explicitly model causal regulatory relationships between transcription factors and their genes, and shows superior performance to existing models on the basis of recovery of known regulatory relationships. We validated the predictive power of the model by perturbing transcription factors in silico and comparing transcriptome-wide effects with experimental data. Our study represents the first step in constructing a complete, predictive, biophysical model of gene expression regulation.

Identification of CgeA as a glycoprotein that anchors polysaccharides to the spore surface in Bacillus subtilis

The Bacillus subtilis spore is composed of a core, containing chromosomal DNA, surrounded by a cortex layer made of peptidoglycan, and a coat composed of concentric proteinaceous layers. A polysaccharide layer is added to the spore surface, and likely anchored to the crust, the coat outermost layer. However, the identity of the coat protein(s) to which the spore polysaccharides (SPS) are attached is uncertain. First, we showed that the crust proteins CotVWXYZ and CgeA were all contained in the peeled SPS layer obtained from a strain missing CotE, the outer coat morphogenetic protein, suggesting that the SPS is indeed bound to at least one of the spore surface proteins. Second, CgeA is known to be located at the most downstream position in the crust assembly pathway. An analysis of truncated variants of CgeA suggested that its N-terminal half is required for localization to the spore surface, while its C-terminal half is necessary for SPS addition. Third, an amino acid substitution strategy revealed that SPS was anchored at threonine 112 (T112), which constitutes a probable O-glycosylation site on CgeA. Our results indicated that CgeA is a glycoprotein required to initiate SPS assembly and serves as an anchor protein linking the crust and SPS layers.

Sporulation conditions influence the surface and adhesion properties of Bacillus subtilis spores

Spore-forming bacteria of the Bacillus subtilis group are responsible for recurrent contamination of processing lines in the food industry which can lead to food spoilage. The persistence of B. subtilis would be due to the high resistance of spores to extreme environmental condition and their propensity to contaminate surfaces. While it is well known that sporulation conditions modulate spore resistance properties, little is known about their effect on surface and adhesion properties. Here, we studied the impact of 13 sporulation conditions on the surface and adhesion properties of B. subtilis 168 spores. We showed that Ca2+ or Mg2+ depletion, lower oxygen availability, acidic pH as well as oxidative stresses during sporulation lead to the release of more hydrophobic and adherent spores. The consequences of these sporulation conditions on crust composition in carbohydrates and proteins were also evaluated. The crust glycans of spores produced in a sporulation medium depleted in Ca2+ or Mg2+ or oxygen-limited conditions were impaired and contained lower amounts of rhamnose and legionaminic acid. In addition, we showed that lower oxygen availability or addition of hydrogen peroxide during sporulation decreases the relative amount of two crust proteins (CgeA and CotY) and the changes observed in these conditions could be due to transcriptional repression of genes involved in crust synthesis in late stationary phase. The fact that sporulation conditions affect the ease with which spores can contaminate surfaces could explain the frequent and recurrent presence of B. subtilis spores in food processing lines.

Extensive cellular multi-tasking within Bacillus subtilis biofilms

Bacillus subtilis is a soil-dwelling bacterium that can form biofilms, or communities of cells surrounded by a self-produced extracellular matrix. In biofilms, genetically identical cells often exhibit heterogeneous transcriptional phenotypes, so that subpopulations of cells carry out essential yet costly cellular processes that allow the entire population to thrive. Surprisingly, the extent of phenotypic heterogeneity and the relationships between subpopulations of cells within biofilms of even in well-studied bacterial systems like B. subtilis remains largely unknown. To determine relationships between these subpopulations of cells, we created 182 strains containing pairwise combinations of fluorescent transcriptional reporters for the expression state of 14 different genes associated with potential cellular subpopulations. We determined the spatial organization of the expression of these genes within biofilms using confocal microscopy, which revealed that many reporters localized to distinct areas of the biofilm, some of which were co-localized. We used flow cytometry to quantify reporter co-expression, which revealed that many cells "multi-task," simultaneously expressing two reporters. These data indicate that prior models describing B. subtilis cells as differentiating into specific cell types, each with a specific task or function, were oversimplified. Only a few subpopulations of cells, including surfactin and plipastatin producers, as well as sporulating and competent cells, appear to have distinct roles based on the set of genes examined here. These data will provide us with a framework with which to further study and make predictions about the roles of diverse cellular phenotypes in B. subtilis biofilms. IMPORTANCE Many microbes differentiate, expressing diverse phenotypes to ensure their survival in various environments. However, studies on phenotypic differentiation have typically examined only a few phenotypes at one time, thus limiting our knowledge about the extent of differentiation and phenotypic overlap in the population. We investigated the spatial organization and gene expression relationships for genes important in B. subtilis biofilms. In doing so, we mapped spatial gene expression patterns and expanded the number of cell populations described in the B. subtilis literature. It is likely that other bacteria also display complex differentiation patterns within their biofilms. Studying the extent of cellular differentiation in other microbes may be important when designing therapies for disease-causing bacteria, where studying only a single phenotype may be masking underlying phenotypic differentiation relevant to infection outcomes.

New Thoughts on an Old Topic: Secrets of Bacterial Spore Resistance Slowly Being Revealed

The quest for bacterial survival is exemplified by spores formed by some Firmicutes members. They turn up everywhere one looks, and their ubiquity reflects adaptations to the stresses bacteria face. Spores are impactful in public health, food safety, and biowarfare. Heat resistance is the hallmark of spores and is countered principally by a mineralized gel-like protoplast, termed the spore core, with reduced water which minimizes macromolecular movement/denaturation/aggregation. Dry heat, however, introduces mutations into spore DNA. Spores have countermeasures to extreme conditions that are multifactorial, but the fact that spore DNA is in a crystalline-like nucleoid in the spore core, likely due to DNA saturation with small acid-soluble spore proteins (SASPs), suggests that reduced macromolecular motion is also critical in spore dry heat resistance. SASPs are also central in the radiation resistance characteristic of spores, where the contributions of four spore features-SASP; Ca2+, with pyridine-2,6-dicarboxylic acid (CaDPA); photoproduct lyase; and low water content-minimize DNA damage. Notably, the spore environment steers UV photochemistry toward a product that germinated spores can repair without significant mutagenesis. This resistance extends to chemicals and macromolecules that could damage spores. Macromolecules are excluded by the spore coat which impedes the passage of moieties of ≥10 kDa. Additionally, damaging chemicals may be degraded or neutralized by coat enzymes/proteins. However, the principal protective mechanism here is the inner membrane, a compressed structure lacking lipid fluidity and presenting a barrier to the diffusion of chemicals into the spore core; SASP saturation of DNA also protects against genotoxic chemicals. Spores are also resistant to other stresses, including high pressure and abrasion. Regardless, overarching mechanisms associated with resistance seem to revolve around reduced molecular motion, a fine balance between rigidity and flexibility, and perhaps efficient repair.

An atlas of gene regulatory networks for memory CD4 + T cells in youth and old age

Aging profoundly affects immune-system function, promoting susceptibility to pathogens, cancers and chronic inflammation. We previously identified a population of IL-10-producing, T follicular helper-like cells (" Tfh10 "), linked to suppressed vaccine responses in aged mice. Here, we integrate single-cell ( sc )RNA-seq, scATAC-seq and genome-scale modeling to characterize Tfh10 - and the full CD4 + memory T cell ( CD4 + TM ) compartment - in young and old mice. We identified 13 CD4 + TM populations, which we validated through cross-comparison to prior scRNA-seq studies. We built gene regulatory networks ( GRNs ) that predict transcription-factor control of gene expression in each T-cell population and how these circuits change with age. Through integration with pan-cell aging atlases, we identified intercellular-signaling networks driving age-dependent changes in CD4 + TM. Our atlas of finely resolved CD4 + TM subsets, GRNs and cell-cell communication networks is a comprehensive resource of predicted regulatory mechanisms operative in memory T cells, presenting new opportunities to improve immune responses in the elderly.

EnsInfer: a simple ensemble approach to network inference outperforms any single method

This study evaluates both a variety of existing base causal inference methods and a variety of ensemble methods. We show that: (i) base network inference methods vary in their performance across different datasets, so a method that works poorly on one dataset may work well on another; (ii) a non-homogeneous ensemble method in the form of a Naive Bayes classifier leads overall to as good or better results than using the best single base method or any other ensemble method; (iii) for the best results, the ensemble method should integrate all methods that satisfy a statistical test of normality on training data. The resulting ensemble model EnsInfer easily integrates all kinds of RNA-seq data as well as new and existing inference methods. The paper categorizes and reviews state-of-the-art underlying methods, describes the EnsInfer ensemble approach in detail, and presents experimental results. The source code and data used will be made available to the community upon publication.

Expression of the 2Duf protein in wild-type Bacillus subtilis spores stabilizes inner membrane proteins and increases spore resistance to wet heat and hydrogen peroxide

AIMS

This work aimed to characterize spore inner membrane (IM) properties and the mechanism of spore killing by wet heat and H2O2 with spores overexpressing the 2Duf protein, which is naturally encoded from a transposon found only in some Bacillus strains with much higher spore resistance than wild-type spores.

METHODS AND RESULTS

Killing of Bacillus subtilis spores by wet heat or hydrogen peroxide (H2O2) was slower when 2Duf was present, and Ca-dipicolinic acid release was slower than killing. Viabilities on rich plates of wet heat- or H2O2 -treated spores +/- 2Duf were lower when NaCl was added, but higher with glucose. Addition of glucose but not Casamino acids addition increased treated spores' viability on minimal medium plates. Spores with 2Duf required higher heat activation for germination, and their germination was more wet-heat resistant than that of wild-type spores, processes that involve IM proteins. IM permeability and lipid mobility were lower in spores with 2Duf, although IM phospholipid composition was similar in spores +/- 2Duf.

CONCLUSIONS

These results and previous work suggests that wet heat and H2O2 kill spores by damaging an IM enzyme or enzymes involved in oxidative phosphorylation.

etiBsu1209: A comprehensive multiscale metabolic model for Bacillus subtilis

Genome-scale metabolic models (GEMs) have been widely used to guide the computational design of microbial cell factories, and to date, seven GEMs have been reported for Bacillus subtilis, a model gram-positive microorganism widely used in bioproduction of functional nutraceuticals and food ingredients. However, none of them are widely used because they often lead to erroneous predictions due to their low predictive power and lack of information on regulatory mechanisms. In this work, we constructed a new version of GEM for B. subtilis (iBsu1209), which contains 1209 genes, 1595 metabolites, and 1948 reactions. We applied machine learning to fill gaps, which formed a relatively complete metabolic network able to predict with high accuracy (89.3%) the growth of 1209 mutants under 12 different culture conditions. In addition, we developed a visualization and code-free software, Model Tool, for multiconstraints model reconstruction and analysis. We used this software to construct etiBsu1209, a multiscale model that integrates enzymatic constraints, thermodynamic constraints, and transcriptional regulatory networks. Furthermore, we used etiBsu1209 to guide a metabolic engineering strategy (knocking out fabI and yfkN genes) for the overproduction of nutraceutical menaquinone-7, and the titer increased to 153.94 mg/L, 2.2-times that of the parental strain. To the best of our knowledge, etiBsu1209 is the first comprehensive multiscale model for B. subtilis and can serve as a solid basis for rational computational design of B. subtilis cell factories for bioproduction.

Structure primed embedding on the transcription factor manifold enables transparent model architectures for gene regulatory network and latent activity inference

The modeling of gene regulatory networks (GRNs) is limited due to a lack of direct measurements of regulatory features in genome-wide screens. Most GRN inference methods are therefore forced to model relationships between regulatory genes and their targets with expression as a proxy for the upstream independent features, complicating validation and predictions produced by modeling frameworks. Separating covariance and regulatory influence requires aggregation of independent and complementary sets of evidence, such as transcription factor (TF) binding and target gene expression. However, the complete regulatory state of the system, e.g. TF activity (TFA) is unknown due to a lack of experimental feasibility, making regulatory relations difficult to infer. Some methods attempt to account for this by modeling TFA as a latent feature, but these models often use linear frameworks that are unable to account for non-linearities such as saturation, TF-TF interactions, and other higher order features. Deep learning frameworks may offer a solution, as they are capable of modeling complex interactions and capturing higher-order latent features. However, these methods often discard central concepts in biological systems modeling, such as sparsity and latent feature interpretability, in favor of increased model complexity. We propose a novel deep learning autoencoder-based framework, StrUcture Primed Inference of Regulation using latent Factor ACTivity (SupirFactor), that scales to single cell genomic data and maintains interpretability to perform GRN inference and estimate TFA as a latent feature. We demonstrate that SupirFactor outperforms current leading GRN inference methods, predicts biologically relevant TFA and elucidates functional regulatory pathways through aggregation of TFs.

Prediction and validation of novel SigB regulon members in Bacillus subtilis and regulon structure comparison to Bacillales members

BACKGROUND

Sigma factor B (SigB) is the central regulator of the general stress response in Bacillus subtilis and regulates a group of genes in response to various stressors, known as the SigB regulon members. Genes that are directly regulated by SigB contain a promotor binding motif (PBM) with a previously identified consensus sequence.

RESULTS

In this study, refined SigB PBMs were derived and different spacer compositions and lengths (N₁₂-N₁₇) were taken into account. These were used to identify putative SigB-regulated genes in the B. subtilis genome, revealing 255 genes: 99 had been described in the literature and 156 genes were newly identified, increasing the number of SigB putative regulon members (with and without a SigB PBM) to > 500 in B. subtilis. The 255 genes were assigned to five categories (I-V) based on their similarity to the original SigB consensus sequences. The functionalities of selected representatives per category were assessed using promoter-reporter fusions in wt and ΔsigB mutants upon exposure to heat, ethanol, and salt stress. The activity of the P_rsbV (I) positive control was induced upon exposure to all three stressors. P_ytoQ (II) showed SigB-dependent activity only upon exposure to ethanol, whereas P_pucI (II) with a N₁₇ spacer and P_ylaL (III) with a N₁₆ spacer showed mild induction regardless of heat/ethanol/salt stress. P_ywzA (III) and P_yaaI (IV) displayed ethanol-specific SigB-dependent activities despite a lower-level conserved - 10 binding motif. P_gtaB (V) was SigB-induced under ethanol and salt stress while lacking a conserved - 10 binding region. The activities of P_ygaO and P_ykaA (III) did not show evident changes under the conditions tested despite having a SigB PBM that highly resembled the consensus. The identified extended SigB regulon candidates in B. subtilis are mainly involved in coping with stress but are also engaged in other cellular processes. Orthologs of SigB regulon candidates with SigB PBMs were identified in other Bacillales genomes, but not all showed a SigB PBM. Additionally, genes involved in the integration of stress signals to activate SigB were predicted in these genomes, indicating that SigB signaling and regulon genes are species-specific.

CONCLUSION

The entire SigB regulatory network is sophisticated and not yet fully understood even for the well-characterized organism B. subtilis 168. Knowledge and information gained in this study can be used in further SigB studies to uncover a complete picture of the role of SigB in B. subtilis and other species.

Identification and characterization of new proteins crucial for bacterial spore resistance and germination

2Duf, named after the presence of a transmembrane (TM) Duf421 domain and a small Duf1657 domain in its sequence, is likely located in the inner membrane (IM) of spores in some Bacillus species carrying a transposon with an operon termed spoVA ^2mob. These spores are known for their extreme resistance to wet heat, and 2Duf is believed to be the primary contributor to this trait. In this study, we found that the absence of YetF or YdfS, both Duf421 domain-containing proteins and found only in wild-type (wt) B. subtilis spores with YetF more abundant, leads to decreased resistance to wet heat and agents that can damage spore core components. The IM phospholipid compositions and core water and calcium-dipicolinic acid levels of YetF-deficient spores are similar to those of wt spores, but the deficiency could be restored by ectopic insertion of yetF, and overexpression of YetF increased wt spore resistance to wet heat. In addition, yetF and ydfS spores have decreased germination rates as individuals and populations with germinant receptor-dependent germinants and increased sensitivity to wet heat during germination, potentially due to damage to IM proteins. These data are consistent with a model in which YetF, YdfS and their homologs modify IM structure to reduce IM permeability and stabilize IM proteins against wet heat damage. Multiple yetF homologs are also present in other spore forming Bacilli and Clostridia, and even some asporogenous Firmicutes, but fewer in asporogenous species. The crystal structure of a YetF tetramer lacking the TM helices has been reported and features two distinct globular subdomains in each monomer. Sequence alignment and structure prediction suggest this fold is likely shared by other Duf421-containing proteins, including 2Duf. We have also identified naturally occurring 2duf homologs in some Bacilli and Clostridia species and in wt Bacillus cereus spores, but not in wt B. subtilis. Notably, the genomic organization around the 2duf gene in most of these species is similar to that in spoVA ^2mob, suggesting that one of these species was the source of the genes on this operon in the extremely wet heat resistant spore formers.

Network-based approaches for modeling disease regulation and progression

Molecular interaction networks lay the foundation for studying how biological functions are controlled by the complex interplay of genes and proteins. Investigating perturbed processes using biological networks has been instrumental in uncovering mechanisms that underlie complex disease phenotypes. Rapid advances in omics technologies have prompted the generation of high-throughput datasets, enabling large-scale, network-based analyses. Consequently, various modeling techniques, including network enrichment, differential network extraction, and network inference, have proven to be useful for gaining new mechanistic insights. We provide an overview of recent network-based methods and their core ideas to facilitate the discovery of disease modules or candidate mechanisms. Knowledge generated from these computational efforts will benefit biomedical research, especially drug development and precision medicine. We further discuss current challenges and provide perspectives in the field, highlighting the need for more integrative and dynamic network approaches to model disease development and progression.

NetREX-CF integrates incomplete transcription factor data with gene expression to reconstruct gene regulatory networks

The inference of Gene Regulatory Networks (GRNs) is one of the key challenges in systems biology. Leading algorithms utilize, in addition to gene expression, prior knowledge such as Transcription Factor (TF) DNA binding motifs or results of TF binding experiments. However, such prior knowledge is typically incomplete, therefore, integrating it with gene expression to infer GRNs remains difficult. To address this challenge, we introduce NetREX-CF-Regulatory Network Reconstruction using EXpression and Collaborative Filtering-a GRN reconstruction approach that brings together Collaborative Filtering to address the incompleteness of the prior knowledge and a biologically justified model of gene expression (sparse Network Component Analysis based model). We validated the NetREX-CF using Yeast data and then used it to construct the GRN for Drosophila Schneider 2 (S2) cells. To corroborate the GRN, we performed a large-scale RNA-Seq analysis followed by a high-throughput RNAi treatment against all 465 expressed TFs in the cell line. Our knockdown result has not only extensively validated the GRN we built, but also provides a benchmark that our community can use for evaluating GRNs. Finally, we demonstrate that NetREX-CF can infer GRNs using single-cell RNA-Seq, and outperforms other methods, by using previously published human data.

Network reconstruction for trans acting genetic loci using multi-omics data and prior information

BACKGROUND

Molecular measurements of the genome, the transcriptome, and the epigenome, often termed multi-omics data, provide an in-depth view on biological systems and their integration is crucial for gaining insights in complex regulatory processes. These data can be used to explain disease related genetic variants by linking them to intermediate molecular traits (quantitative trait loci, QTL). Molecular networks regulating cellular processes leave footprints in QTL results as so-called trans-QTL hotspots. Reconstructing these networks is a complex endeavor and use of biological prior information can improve network inference. However, previous efforts were limited in the types of priors used or have only been applied to model systems. In this study, we reconstruct the regulatory networks underlying trans-QTL hotspots using human cohort data and data-driven prior information.

METHODS

We devised a new strategy to integrate QTL with human population scale multi-omics data. State-of-the art network inference methods including BDgraph and glasso were applied to these data. Comprehensive prior information to guide network inference was manually curated from large-scale biological databases. The inference approach was extensively benchmarked using simulated data and cross-cohort replication analyses. Best performing methods were subsequently applied to real-world human cohort data.

RESULTS

Our benchmarks showed that prior-based strategies outperform methods without prior information in simulated data and show better replication across datasets. Application of our approach to human cohort data highlighted two novel regulatory networks related to schizophrenia and lean body mass for which we generated novel functional hypotheses.

CONCLUSIONS

We demonstrate that existing biological knowledge can improve the integrative analysis of networks underlying trans associations and generate novel hypotheses about regulatory mechanisms.

Time-Resolved Proteomics of Germinating Spores of Bacillus cereus

Bacillus cereus is a spore-forming human pathogen that is a burden to the food chain. Dormant spores are highly resistant to harsh environmental conditions, but lose resistance after germination. In this study, we investigate the B. cereus spore proteome upon spore germination and outgrowth so as to obtain new insights into the molecular mechanisms involved. We used mass spectrometry combined with co-expression network analysis and obtained a unique global proteome view of the germination and outgrowth processes of B. cereus spores by monitoring 2211 protein changeovers. We are the first to examine germination and outgrowth models of B. cereus spores experimentally by studying the dynamics of germinant receptors, other proteins involved in spore germination and resistance, and coat and exosporium proteins. Furthermore, through the co-expression analysis of 1175 proteins identified with high quality data, germination proteome data were clustered into eight modules (termed black, blue, brown, green, red, turquoise, grey, and yellow), whose associated functions and expression profiles were investigated. Germination related proteins were clustered into blue and brown modules, the abundances of which decreased after finishing germination. In the brown and blue we identified 124 proteins that could be vital during germination. These proteins will be very interesting to study in future genetic studies regarding their function in spore revival in B. cereus.

Systems-level identification of key transcription factors in immune cell specification

Transcription factors (TFs) are crucial for regulating cell differentiation during the development of the immune system. However, the key TFs for orchestrating the specification of distinct immune cells are not fully understood. Here, we integrated the transcriptomic and epigenomic measurements in 73 mouse and 61 human primary cell types, respectively, that span the immune cell differentiation pathways. We constructed the cell-type-specific transcriptional regulatory network and assessed the global importance of TFs based on the Taiji framework, which is a method we have previously developed that can infer the global impact of TFs using integrated transcriptomic and epigenetic data. Integrative analysis across cell types revealed putative driver TFs in cell lineage-specific differentiation in both mouse and human systems. We have also identified TF combinations that play important roles in specific developmental stages. Furthermore, we validated the functions of predicted novel TFs in murine CD8+ T cell differentiation and showed the importance of Elf1 and Prdm9 in the effector versus memory T cell fate specification and Kdm2b and Tet3 in promoting differentiation of CD8+ tissue resident memory (Trm) cells, validating the approach. Thus, we have developed a bioinformatic approach that provides a global picture of the regulatory mechanisms that govern cellular differentiation in the immune system and aids the discovery of novel mechanisms in cell fate decisions.

Role of novel polysaccharide layers in assembly of the exosporium, the outermost protein layer of the Bacillus anthracis spore

A fundamental question in cell biology is how cells assemble their outer layers. The bacterial endospore is a well-established model for cell layer assembly. However, the assembly of the exosporium, a complex protein shell comprising the outermost layer in the pathogen Bacillus anthracis, remains poorly understood. Exosporium assembly begins with the deposition of proteins at one side of the spore surface, followed by the progressive encirclement of the spore. We seek to resolve a major open question: the mechanism directing exosporium assembly to the spore, and then into a closed shell. We hypothesized that material directly underneath the exosporium (the interspace) directs exosporium assembly to the spore and drives encirclement. In support of this, we show that the interspace possesses at least two distinct layers of polysaccharide. Secondly, we show that putative polysaccharide biosynthetic genes are required for exosporium encirclement, suggesting a direct role for the interspace. These results not only significantly clarify the mechanism of assembly of the exosporium, an especially widespread bacterial outer layer, but also suggest a novel mechanism in which polysaccharide layers drive the assembly of a protein shell.

Using genome-wide expression compendia to study microorganisms

A gene expression compendium is a heterogeneous collection of gene expression experiments assembled from data collected for diverse purposes. The widely varied experimental conditions and genetic backgrounds across samples creates a tremendous opportunity for gaining a systems level understanding of the transcriptional responses that influence phenotypes. Variety in experimental design is particularly important for studying microbes, where the transcriptional responses integrate many signals and demonstrate plasticity across strains including response to what nutrients are available and what microbes are present. Advances in high-throughput measurement technology have made it feasible to construct compendia for many microbes. In this review we discuss how these compendia are constructed and analyzed to reveal transcriptional patterns.

The impact of PrsA over-expression on the Bacillus subtilis transcriptome during fed-batch fermentation of alpha-amylase production

The production of the alpha-amylase (AMY) enzyme in Bacillus subtilis at a high rate leads to the accumulation of unfolded AMY, which causes secretion stress. The over-expression of the PrsA chaperone aids enzyme folding and reduces stress. To identify affected pathways and potential mechanisms involved in the reduced growth, we analyzed the transcriptomic differences during fed-batch fermentation between a PrsA over-expressing strain and control in a time-series RNA-seq experiment. We observe transcription in 542 unannotated regions, of which 234 had significant changes in expression levels between the samples. Moreover, 1,791 protein-coding sequences, 80 non-coding genes, and 20 riboswitches overlapping UTR regions of coding genes had significant changes in expression. We identified putatively regulated biological processes via gene-set over-representation analysis of the differentially expressed genes; overall, the analysis suggests that the PrsA over-expression affects ATP biosynthesis activity, amino acid metabolism, and cell wall stability. The investigation of the protein interaction network points to a potential impact on cell motility signaling. We discuss the impact of these highlighted mechanisms for reducing secretion stress or detrimental aspects of PrsA over-expression during AMY production.

Comparative biological network analysis for differentially expressed proteins as a function of bacilysin biosynthesis in Bacillus subtilis

The Gram-positive bacterium Bacillus subtilis produces a diverse range of secondary metabolites with different structures and activities. Among them, bacilysin is an enzymatically synthesized dipeptide that consists of L-alanine and L-anticapsin. Previous research by our group has suggested bacilysin's role as a pleiotropic molecule in its producer, B. subtilis PY79. However, the nature of protein interactions in the absence of bacilysin has not been defined. In the present work, we constructed a protein-protein interaction subnetwork by using Omics Integrator based on our recent comparative proteomics data obtained from a bacilysin-silenced strain, OGU1. Functional enrichment analyses on the resulting networks pointed to certain putatively perturbed pathways such as citrate cycle, quorum sensing and secondary metabolite biosynthesis. Various molecules, which were absent from the experimental data, were included in the final network. We believe that this study can guide further experiments in the identification and confirmation of protein-protein interactions in B. subtilis.

Transcriptomic Analysis Reveals the Role of tmRNA on Biofilm Formation in Bacillus subtilis

Bacillus strains are widely distributed in terrestrial and marine environments, and some of them are used as biocontrol organisms for their biofilm-formation ability. In Bacillus subtilis, biofilm formation is fine-tuned by a complex network, a clear understanding of which still requires study. In bacteria, tmRNA, encoded by the ssrA gene, catalyzes trans-translation that can rescue ribosomes stalled on mRNA transcripts lacking a functional stop codon. tmRNA also affects physiological bioprocesses in some bacteria. In this study, we constructed a ssrA mutant in B. subtilis and found that the biofilm formation in the ssrA mutant was largely impaired. Moreover, we isolated a biofilm-formation suppressor of ssrA, in which the biofilm formation was restored to a level even stronger than that in the wild type. We further performed RNAseq assays with the wild type, ssrA mutant, and suppressor of ssrA for comparisons of their transcriptomes. By analyzing the transcriptomic data, we predicted the possible functions of some differentially expressed genes (DEGs) in the tmRNA regulation of biofilm formation in B. subtilis. Finally, we found that the overexpression of two DEGs, acoA and yhjR, could restore the biofilm formation in the ssrA mutant, indicating that AcoA and YhjR were immediate regulators involved in the tmRNA regulatory web controlling biofilm formation in B. subtilis. Our data can improve the knowledge about the molecular network involved in Bacillus biofilm formation and provide new targets for manipulation of Bacillus biofilms for future investigation.

Seed banks alter the molecular evolutionary dynamics of Bacillus subtilis

Fluctuations in the availability of resources constrain the growth and reproduction of individuals, which subsequently affects the evolution of their respective populations. Many organisms contend with such fluctuations by entering a reversible state of reduced metabolic activity, a phenomenon known as dormancy. This pool of dormant individuals (i.e. a seed bank) does not reproduce and is expected to act as an evolutionary buffer, though it is difficult to observe this effect directly over an extended evolutionary timescale. Through genetic manipulation, we analyze the molecular evolutionary dynamics of Bacillus subtilis populations in the presence and absence of a seed bank over 700 days. The ability of these bacteria to enter a dormant state increased the accumulation of genetic diversity over time and altered the trajectory of mutations, findings that were recapitulated using simulations based on a mathematical model of evolutionary dynamics. While the ability to form a seed bank did not alter the degree of negative selection, we found that it consistently altered the direction of molecular evolution across genes. Together, these results show that the ability to form a seed bank can affect the direction and rate of molecular evolution over an extended evolutionary timescale.

A Sporulation-Specific sRNA Bvs196 Contributing to the Developing Spore in Bacillus velezensis

Many putative sRNAs have been characterized using bioinformatic analysis and high-throughput sequencing in Gram-positive Bacillus strains, but there are only a few functional studies on the sRNAs involved in the spore formation developmental process. In particular, there is no sRNA confirmed experimentally to regulate the late stages of sporulation. Bvs196 is an sRNA with a length of 294 nucleotides that is abundantly expressed in the stationary phase of several media and independently transcribed in Bacillus velezensis strain PEBA20, as validated by RNA-seq and Northern blot,. It is also confirmed, by qRT-PCR, that Bvs196 is transcribed abundantly throughout the intermediate and late stages of sporulation. Using the gfpmut3a gene transcriptional reporter demonstrates that Bvs196 is expressed specifically in the forespore during sporulation and controlled by σF and σG (mainly by σG). This was observed by fluorescence microscopy and multi-function microplate reader. Further evolutionary conservation analysis found that Bvs196 is widely present in Bacillus with a strongly conserved and stable secondary structure. Resistance phenotypic assays of spores formed from the Bvs196 deletion mutant, the overexpressed Bvs196 mutant, and the wild-type strain revealed that the absence of Bvs196 led to reduced heat and UV resistance and enhanced formaldehyde resistance. We determined, by MST analysis, that Bvs196 can directly interact with spo0A and sspN-tlp mRNAs in vitro, and that short incomplete complementary paired bases affect the binding affinity of Bvs196 to target mRNAs. Our results suggest that Bvs196 is a novel sporulation-specific sRNA of B. velezensis, 294 nt in length, independently transcribed under the control of σF and σG in the forespore during sporulation, and that it affects spore resistance, and is able to directly interact with spo0A and sspN-tlp mRNAs. The remarkable conservation and impressive expression level of Bvs196 imply that it acts as an important conservative regulator, presumably by interacting with many other unknown targets in the forespore, and therefore contributing to spore properties. This work provides new clues for further understanding of the spore formation regulatory network.

Mathematical models to study the biology of pathogens and the infectious diseases they cause

Mathematical models have many applications in infectious diseases: epidemiologists use them to forecast outbreaks and design containment strategies; systems biologists use them to study complex processes sustaining pathogens, from the metabolic networks empowering microbial cells to ecological networks in the microbiome that protects its host. Here, we (1) review important models relevant to infectious diseases, (2) draw parallels among models ranging widely in scale. We end by discussing a minimal set of information for a model to promote its use by others and to enable predictions that help us better fight pathogens and the diseases they cause.

High performance single-cell gene regulatory network inference at scale: The Inferelator 3.0

Motivation

Gene regulatory networks define regulatory relationships between transcription factors and target genes within a biological system, and reconstructing them is essential for understanding cellular growth and function. Methods for inferring and reconstructing networks from genomics data have evolved rapidly over the last decade in response to advances in sequencing technology and machine learning. The scale of data collection has increased dramatically; the largest genome-wide gene expression datasets have grown from thousands of measurements to millions of single cells, and new technologies are on the horizon to increase to tens of millions of cells and above.

Results

In this work, we present the Inferelator 3.0, which has been significantly updated to integrate data from distinct cell types to learn context-specific regulatory networks and aggregate them into a shared regulatory network, while retaining the functionality of the previous versions. The Inferelator is able to integrate the largest single-cell datasets and learn cell-type-specific gene regulatory networks. Compared to other network inference methods, the Inferelator learns new and informative Saccharomyces cerevisiae networks from single-cell gene expression data, measured by recovery of a known gold standard. We demonstrate its scaling capabilities by learning networks for multiple distinct neuronal and glial cell types in the developing Mus musculus brain at E18 from a large (1.3 million) single-cell gene expression dataset with paired single-cell chromatin accessibility data.

Availability and implementation

The inferelator software is available on GitHub (https://github.com/flatironinstitute/inferelator) under the MIT license and has been released as python packages with associated documentation (https://inferelator.readthedocs.io/).

Supplementary information

Supplementary data are available at Bioinformatics online.

Gene regulatory network inference in long-lived C. elegans reveals modular properties that are predictive of novel aging genes

We design a “wisdom-of-the-crowds” GRN inference pipeline and couple it to complex network analysis to understand the organizational principles governing gene regulation in long-lived glp-1/Notch Caenorhabditis elegans. The GRN has three layers (input, core, and output) and is topologically equivalent to bow-tie/hourglass structures prevalent among metabolic networks. To assess the functional importance of structural layers, we screened 80% of regulators and discovered 50 new aging genes, 86% with human orthologues. Genes essential for longevity—including ones involved in insulin-like signaling (ILS)—are at the core, indicating that GRN's structure is predictive of functionality. We used in vivo reporters and a novel functional network covering 5,497 genetic interactions to make mechanistic predictions. We used genetic epistasis to test some of these predictions, uncovering a novel transcriptional regulator, sup-37, that works alongside DAF-16/FOXO. We present a framework with predictive power that can accelerate discovery in C. elegans and potentially humans.

•

Gene-regulatory inference provides global network of long-lived animals

•

The large-scale topology of the network has an hourglass structure

•

Membership to the core of the hourglass is a good predictor of functionality

•

Discovered 50 novel aging genes, including sup-37, a DAF-16 dependent gene

Prediction of whole-cell transcriptional response with machine learning

MOTIVATION

Applications in synthetic and systems biology can benefit from measuring whole-cell response to biochemical perturbations. Execution of experiments to cover all possible combinations of perturbations is infeasible. In this paper, we present the host response model (HRM), a machine learning approach that maps response of single perturbations to transcriptional response of the combination of perturbations.

RESULTS

The HRM combines high-throughput sequencing with machine learning to infer links between experimental context, prior knowledge of cell regulatory networks, and RNASeq data to predict a gene's dysregulation. We find that the HRM can predict the directionality of dysregulation to a combination of inducers with an accuracy of >90% using data from single inducers. We further find that the use of prior, known cell regulatory networks doubles the predictive performance of the HRM (an R2 from 0.3 to 0.65). The model was validated in two organisms, Escherichia coli and Bacillus subtilis, using new experiments conducted after training. Finally, while the HRM is trained with gene expression data, the direct prediction of differential expression makes it possible to also conduct enrichment analyses using its predictions. We show that the HRM can accurately classify >95% of the pathway regulations. The HRM reduces the number of RNASeq experiments needed as responses can be tested in silico prior to the experiment.

AVAILABILITY AND IMPLEMENTATION

The HRM software and tutorial are available at https://github.com/sd2e/CDM and the configurable differential expression analysis tools and tutorials are available at https://github.com/SD2E/omics_tools.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Developmental gene regulatory network connections predicted by machine learning from gene expression data alone

Gene regulatory network (GRN) inference can now take advantage of powerful machine learning algorithms to complement traditional experimental methods in building gene networks. However, the dynamical nature of embryonic development-representing the time-dependent interactions between thousands of transcription factors, signaling molecules, and effector genes-is one of the most challenging arenas for GRN prediction. In this work, we show that successful GRN predictions for a developmental network from gene expression data alone can be obtained with the Priors Enriched Absent Knowledge (PEAK) network inference algorithm. PEAK is a noise-robust method that models gene expression dynamics via ordinary differential equations and selects the best network based on information-theoretic criteria coupled with the machine learning algorithm Elastic Net. We test our GRN prediction methodology using two gene expression datasets for the purple sea urchin, Stronglyocentrotus purpuratus, and cross-check our results against existing GRN models that have been constructed and validated by over 30 years of experimental results. Our results find a remarkably high degree of sensitivity in identifying known gene interactions in the network (maximum 81.58%). We also generate novel predictions for interactions that have not yet been described, which provide a resource for researchers to use to further complete the sea urchin GRN. Published ChIPseq data and spatial co-expression analysis further support a subset of the top novel predictions. We conclude that GRN predictions that match known gene interactions can be produced using gene expression data alone from developmental time series experiments.

Quantitative prediction of conditional vulnerabilities in regulatory and metabolic networks using PRIME

The ability of Mycobacterium tuberculosis (Mtb) to adopt heterogeneous physiological states underlies its success in evading the immune system and tolerating antibiotic killing. Drug tolerant phenotypes are a major reason why the tuberculosis (TB) mortality rate is so high, with over 1.8 million deaths annually. To develop new TB therapeutics that better treat the infection (faster and more completely), a systems-level approach is needed to reveal the complexity of network-based adaptations of Mtb. Here, we report a new predictive model called PRIME (Phenotype of Regulatory influences Integrated with Metabolism and Environment) to uncover environment-specific vulnerabilities within the regulatory and metabolic networks of Mtb. Through extensive performance evaluations using genome-wide fitness screens, we demonstrate that PRIME makes mechanistically accurate predictions of context-specific vulnerabilities within the integrated regulatory and metabolic networks of Mtb, accurately rank-ordering targets for potentiating treatment with frontline drugs.

Taiji-reprogram: a framework to uncover cell-type specific regulators and predict cellular reprogramming cocktails

Cellular reprogramming is a promising technology to develop disease models and cell-based therapies. Identification of the key regulators defining the cell type specificity is pivotal to devising reprogramming cocktails for successful cell conversion but remains a great challenge. Here, we present a systems biology approach called Taiji-reprogram to efficiently uncover transcription factor (TF) combinations for conversion between 154 diverse cell types or tissues. This method integrates the transcriptomic and epigenomic data to construct cell-type specific genetic networks and assess the global importance of TFs in the network. Comparative analysis across cell types revealed TFs that are specifically important in a particular cell type and often tightly associated with cell-type specific functions. A systematic search of TFs with differential importance in the source and target cell types uncovered TF combinations for desired cell conversion. We have shown that Taiji-reprogram outperformed the existing methods to better recover the TFs in the experimentally validated reprogramming cocktails. This work not only provides a comprehensive catalog of TFs defining cell specialization but also suggests TF combinations for direct cell conversion.

Predictive regulatory and metabolic network models for systems analysis of Clostridioides difficile

We present predictive models for comprehensive systems analysis of Clostridioides difficile, the etiology of pseudomembranous colitis. By leveraging 151 published transcriptomes, we generated an EGRIN model that organizes 90% of C. difficile genes into a transcriptional regulatory network of 297 co-regulated modules, implicating genes in sporulation, carbohydrate transport, and metabolism. By advancing a metabolic model through addition and curation of metabolic reactions including nutrient uptake, we discovered 14 amino acids, diverse carbohydrates, and 10 metabolic genes as essential for C. difficile growth in the intestinal environment. Finally, we developed a PRIME model to uncover how EGRIN-inferred combinatorial gene regulation by transcription factors, such as CcpA and CodY, modulates essential metabolic processes to enable C. difficile growth relative to commensal colonization. The C. difficile interactive web portal provides access to these model resources to support collaborative systems-level studies of context-specific virulence mechanisms in C. difficile.

Resistance properties and the role of the inner membrane and coat of Bacillus subtilis spores with extreme wet heat resistance

AIMS

A protein termed 2Duf greatly increases wet heat resistance of Bacillus subtilis spores. The current work examines the effects of 2Duf on spore resistance to other sporicides, including chemicals that act on or must cross spores' inner membrane (IM), where 2Duf is likely present. The overall aim was to gain a deeper understanding of how 2Duf affects spore resistance, and of spore resistance itself.

METHODS AND RESULTS

2Duf's presence increased spore resistance to chemicals that damage or must cross the IM to kill spores. Spore coat removal decreased 2Duf-spore resistance to chemicals and wet heat, and 2Duf-spores made at higher temperatures were more resistant to wet heat and chemicals. 2Duf-less spores lacking coats and Ca-dipicolinic acid were also extremely sensitive to wet heat and chemicals that transit the IM to kill spores.

CONCLUSIONS

The new work plus previous results lead to a number of important conclusions as follows. (1) 2Duf may influence spore resistance by decreasing the permeability of and lipid mobility in spores' IM. (2) Since wet heat-killed spores that germinate do not accumulate ATP, wet heat may inactivate some spore IM protein essential in ATP production which is stabilized in a more rigid IM. (3) Both Ca-dipicolinic acid and the spore coat play an important role in the permeability of the spore IM, and thus in many spore resistance properties.

SIGNIFICANCE AND IMPACT OF THE STUDY

The work in this manuscript gives a new insight into mechanisms of spore resistance to chemicals and wet heat, to the understanding of spore wet heat killing, and the role of Ca-dipicolinic acid and the coat in spore resistance.

A new role for SR1 from Bacillus subtilis: regulation of sporulation by inhibition of kinA translation

SR1 is a dual-function sRNA from Bacillus subtilis. It inhibits translation initiation of ahrC mRNA encoding the transcription activator of the arginine catabolic operons. Base-pairing is promoted by the RNA chaperone CsrA, which induces a slight structural change in the ahrC mRNA to facilitate SR1 binding. Additionally, SR1 encodes the small protein SR1P that interacts with glyceraldehyde-3P dehydrogenase A to promote binding to RNase J1 and enhancing J1 activity. Here, we describe a new target of SR1, kinA mRNA encoding the major histidine kinase of the sporulation phosphorelay. SR1 and kinA mRNA share 7 complementary regions. Base-pairing between SR1 and kinA mRNA decreases kinA translation without affecting kinA mRNA stability and represses transcription of the KinA/Spo0A downstream targets spoIIE, spoIIGA and cotA. The initial interaction between SR1 and kinA mRNA occurs 10 nt downstream of the kinA start codon and is decisive for inhibition. The sr1 encoded peptide SR1P is dispensable for kinA regulation. Deletion of sr1 accelerates sporulation resulting in low quality spores with reduced stress resistance and altered coat protein composition which can be compensated by sr1 overexpression. Neither CsrA nor Hfq influence sporulation or spore properties.

High Resolution Analysis of Proteome Dynamics during Bacillus subtilis Sporulation

Bacillus subtilis vegetative cells switch to sporulation upon nutrient limitation. To investigate the proteome dynamics during sporulation, high-resolution time-lapse proteomics was performed in a cell population that was induced to sporulate synchronously. Here, we are the first to comprehensively investigate the changeover of sporulation regulatory proteins, coat proteins, and other proteins involved in sporulation and spore biogenesis. Protein co-expression analysis revealed four co-expressed modules (termed blue, brown, green, and yellow). Modules brown and green are upregulated during sporulation and contain proteins associated with sporulation. Module blue is negatively correlated with modules brown and green, containing ribosomal and metabolic proteins. Finally, module yellow shows co-expression with the three other modules. Notably, several proteins not belonging to any of the known transcription regulons were identified as co-expressed with modules brown and green, and might also play roles during sporulation. Finally, levels of some coat proteins, for example morphogenetic coat proteins, decreased late in sporulation.

Identification of an RNA sponge that controls the RoxS riboregulator of central metabolism in Bacillus subtilis

sRNAs are a taxonomically-restricted but transcriptomically-abundant class of post-transcriptional regulators. While of major importance for adaption to the environment, we currently lack global-scale methodology enabling target identification, especially in species without known RNA hub proteins (e.g. Hfq). Using psoralen RNA cross-linking and Illumina-sequencing we identify RNA-RNA interacting pairs in vivo in Bacillus subtilis, resolving previously well-described interactants. Although sRNA-sRNA pairings are rare (compared with sRNA-mRNA), we identify a robust example involving the conserved sRNA RoxS and an unstudied sRNA RosA (Regulator of sRNA A). We show RosA to be the first confirmed RNA sponge described in a Gram-positive bacterium. RosA interacts with at least two sRNAs, RoxS and FsrA. The RosA/RoxS interaction not only affects the levels of RoxS but also its processing and regulatory activity. We also found that the transcription of RosA is repressed by CcpA, the key regulator of carbon-metabolism in B. subtilis. Since RoxS is already known to be transcriptionally controlled by malate via the transcriptional repressor Rex, its post-transcriptional regulation by CcpA via RosA places RoxS in a key position to control central metabolism in response to varying carbon sources.