Conserved Units of Co-Expression in Bacterial Genomes: An Evolutionary Insight into Transcriptional Regulation

Abstract

Genome-wide measurements of transcriptional activity in bacteria indicate that the transcription of successive genes is strongly correlated beyond the scale of operons. Here, we analyze hundreds of bacterial genomes to identify supra-operonic segments of genes that are proximal in a large number of genomes. We show that these synteny segments correspond to genomic units of strong transcriptional co-expression. Structurally, the segments contain operons with specific relative orientations (co-directional or divergent) and nucleoid-associated proteins are found to bind at their boundaries. Functionally, operons inside a same segment are highly co-expressed even in the apparent absence of regulatory factors at their promoter regions. Remote operons along DNA can also be co-expressed if their corresponding segments share a transcriptional or sigma factor, without requiring these factors to bind directly to the promoters of the operons. As evidence that these results apply across the bacterial kingdom, we demonstrate them both in the Gram-negative bacterium Escherichia coli and in the Gram-positive bacterium Bacillus subtilis. The underlying process that we propose involves only RNA-polymerases and DNA: it implies that the transcription of an operon mechanically enhances the transcription of adjacent operons. In support of a primary role of this regulation by facilitated co-transcription, we show that the transcription en bloc of successive operons as a result of transcriptional read-through is strongly and specifically enhanced in synteny segments. Finally, our analysis indicates that facilitated co-transcription may be evolutionary primitive and may apply beyond bacteria.

Supra-operonic clusters of functionally related genes (SOCs) are a source of horizontal gene co-transfers

“…Such larger functional genomic units have been reported previously131415161718, but are still poorly characterized. SOCs were first inferred from the conservation of gene clusters between species14.…”

“…Genes within SOCs are expected to be co-functioning131415161718. As an alternative to CO-AC, we also approximated the pairwise distance distribution of genes in SOCs through the normalized autocovariance of functionally similar gene-pairs (GO-AC), considering two genes to be functionally similar if they had at least one GO term30 in common.…”

Bacterial genome architecture shapes global transcriptional regulation by DNA supercoiling

“…4B) when the genes are moderately expressed in these simulations (one transcript every 2.5 minutes), as expected for most of them in the genome. In bacteria, the regulatory effect of SC is generally analysed from transcriptomes obtained shortly after chromosome relaxation by antibiotics (novobiocin, norfloxacin) which inhibit the DNA gyrase [7,8,9,10]. Based on the previous observations, the picture of a global relaxation might yet appear as slightly misleading, if the effect of gyrase inhibition is very different in convergent vs divergent regions.…”

“…Inheritable changes in DNA topology are also under positive selection during evolution experiments with bacteria, in which SC-modifying mutations can provide a substantial fitness gain [6]. The regulatory action of SC is usually analysed from transcriptomes obtained after treatment by DNA gyrase inhibitors, causing global relaxation of the chromosome and changes in the transcription level of hundreds of genes [7,8,9,10]. Since topoisomerases are found in all bacterial species, including those almost devoid of transcription factors such as Mycoplasma or Buchnera [11,10], they might underpin an ancestral and widespread global regulation mode.…”

“…The regulatory action of SC is usually analysed from transcriptomes obtained after treatment by DNA gyrase inhibitors, causing global relaxation of the chromosome and changes in the transcription level of hundreds of genes [7,8,9,10]. Since topoisomerases are found in all bacterial species, including those almost devoid of transcription factors such as Mycoplasma or Buchnera [11,10], they might underpin an ancestral and widespread global regulation mode. However, our understanding of the underlying mechanisms remains limited, as there is currently no quantitative or even qualitative systematic model of the features defining a supercoiling-sensitive gene, and how it responds to SC changes.…”