
Canonical and non‐canonical EcfG sigma factors control the general stress response in Rhizobium etli
Author(s) -
Jans Ann,
Vercruysse Maarten,
Gao Shanjun,
Engelen Kristof,
Lambrichts Ivo,
Fauvart Maarten,
Michiels Jan
Publication year - 2013
Publication title -
microbiologyopen
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.881
H-Index - 36
ISSN - 2045-8827
DOI - 10.1002/mbo3.137
Subject(s) - sigma factor , biology , genetics , gene , genome , computational biology , phylogenetic tree , plasmid , gene expression , promoter
A core component of the α ‐proteobacterial general stress response (GSR) is the extracytoplasmic function ( ECF ) sigma factor EcfG, exclusively present in this taxonomic class. Half of the completed α ‐proteobacterial genome sequences contain two or more copies of genes encoding σ EcfG ‐like sigma factors, with the primary copy typically located adjacent to genes coding for a cognate anti‐sigma factor (NepR) and two‐component response regulator (PhyR). So far, the widespread occurrence of additional, non‐canonical σ EcfG copies has not satisfactorily been explained. This study explores the hierarchical relation between Rhizobium etli σ EcfG1 and σ EcfG2 , canonical and non‐canonical σ EcfG proteins, respectively. Contrary to reports in other species, we find that σ EcfG1 and σ EcfG2 act in parallel, as nodes of a complex regulatory network, rather than in series, as elements of a linear regulatory cascade. We demonstrate that both sigma factors control unique yet also shared target genes, corroborating phenotypic evidence. σ EcfG1 drives expression of rpoH2 , explaining the increased heat sensitivity of an ecfG1 mutant, while katG is under control of σ EcfG2 , accounting for reduced oxidative stress resistance of an ecfG2 mutant. We also identify non‐coding RNA genes as novel σ EcfG targets. We propose a modified model for GSR regulation in R. etli , in which σ EcfG1 and σ EcfG2 function largely independently. Based on a phylogenetic analysis and considering the prevalence of α ‐proteobacterial genomes with multiple σ EcfG copies, this model may also be applicable to numerous other species.