Global transcriptional regulatory network for Escherichia coli robustly connects gene expression to transcription factor activities

Abstract

Transcriptional regulatory networks (TRNs) have been studied intensely for >25 y. Yet, even for the TRN-probably the best characterized TRN-several questions remain. Here, we address three questions: () How complete is our knowledge of the TRN; () how well can we predict gene expression using this TRN; and () how robust is our understanding of the TRN? First, we reconstructed a high-confidence TRN (hiTRN) consisting of 147 transcription factors (TFs) regulating 1,538 transcription units (TUs) encoding 1,764 genes. The 3,797 high-confidence regulatory interactions were collected from published, validated chromatin immunoprecipitation (ChIP) data and RegulonDB. For 21 different TF knockouts, up to 63% of the differentially expressed genes in the hiTRN were traced to the knocked-out TF through regulatory cascades. Second, we trained supervised machine learning algorithms to predict the expression of 1,364 TUs given TF activities using 441 samples. The algorithms accurately predicted condition-specific expression for 86% (1,174 of 1,364) of the TUs, while 193 TUs (14%) were predicted better than random TRNs. Third, we identified 10 regulatory modules whose definitions were robust against changes to the TRN or expression compendium. Using surrogate variable analysis, we also identified three unmodeled factors that systematically influenced gene expression. Our computational workflow comprehensively characterizes the predictive capabilities and systems-level functions of an organism's TRN from disparate data types.

The Escherichia coli Transcriptome Mostly Consists of Independently Regulated Modules

“…Previous studies have compiled transcriptomics data from independent research groups to study the transcriptional states and regulation of E. coli [21][22][23][24] . Even after resolving the significant normalization challenge with such disparate datasets, many sources of variation remain that obscure biological signals [25][26][27] . These datasets mostly contain microarray data; RNA sequencing (RNA-seq) data yields higher quality data with less noise and larger dynamic range 28 .…”

“…We compiled the global TRN using all interactions from RegulonDB 10.0 3 for both transcription factor and sRNA binding sites. Binding sites were added from recent studies, as described in Fang et al 27 , in addition to binding sites for Nac and NtrC 60 and binding sites for 10 uncharacterized transcription factors 38 . We also included sigma factor binding sites, riboswitch information, and transcriptional attenuation from Ecocyc 61 .…”

Control of metabolite efflux in microbial cell factories: current advances and future prospects

“…Transcription factor engineering rather lags behind in prokaryotes (see e.g. [197,302,303]), but an example of present interest is the use of marA to improve solvent tolerance (geraniol) in E. coli [304], as this acts, at least in part, by increasing the expression level of the enormous (770kDa) [305] and otherwise somewhat intractable (but see [306]) tolC/acrAB efflux transporter. Mutations in marR [307] and σ 70 (rpoD) [308] can have similar effects.…”

Systematic discovery of uncharacterized transcription factors inEscherichia coliK-12 MG1655

“…Third, several computational approaches have been developed for the reconstruction of the TRN, including the use of gene expression data (21,22) , regulon-based associations (23) , and integrated analysis with metabolic models (24) . The expression data-driven approach for TRN reconstruction was widely used to predict transcription factor activities in E. coli K-12.…”