Unité de Microbiologie Structurale / Recherches / Envelope stress response

Envelope stress response

Team: O. Subrini, N. Sassoon, J.M. Betton

Our group, which has a long-term interest in protein folding, studies protein quality control systems in the periplasm of Escherichia coli. Our main objective is to understand the mechanism of misfolded protein sensing by the three-component Cpx system that signals and regulates an envelope stress response in E. coli. In order to investigate the phosphotransfer reactions, the whole system was reconstituted from purified Cpx proteins. The membrane CpxA histidine kinase was produced in a cell-free system in the presence of Brij35 micelles and purified in two chromatography steps with a yield of 1.5 mg/ml of cell-free reaction. The autokinase, phosphotransferase and phophatase activity of CpxA synthesized in vitro were compared with that of the protein produced in bacterial membranes, and confirmed that both membrane proteins were functionally identical. However, when the function of CpxP was tested on both solubilized CpxA, no inhibitory effect was observed. It appears that CpxP did not recognize the conformation of the sensor domain of CpxA in detergent. A further detailed analysis of CpxA variants carrying substitutions in this domain and defective for the phosphatase activity was compatible with this hypothesis. To avoid this problem we are reconstituting CpxA in nanodiscs. This bilayer model will provide a more native environment for the Cpx signaling system than that brought by detergent micelles. We have also developed a cell-based system for overproducing membrane proteins. The overproduction of CpxA in E. coli, like most of recombinant integral membrane proteins, is highly toxic. It was known that the over-expression of atpF gene (or uncF) encoding the β-subunit  of the E. coli ATP synthase could induce a massive internal membrane synthesis, thus decreasing the cellular toxicity. Therefore, we transcriptionally fused this gene to cpxA and tested the co-expression in the classical BL21(DE3) strain and its derivatives, C41(DE3) and C43(DE3). The former strain was only viable, and allowed a high level production of CpxA. After solubilization, about 50 mg of purified CpxA was obtained by liter of culture. Bacteria co-producing AtpF and CpxA were examined by transmission electron microscopy after fixation and staining (Figure). Large invaginations of the inner membrane were observed. It seems that this phenotype linked to the selection of C41(DE3) and C43(DE3) strains can be acquired in the parental BL21(DE3) strain without genetic selection. A possible molecular mechanism would be that these two integral membrane proteins form protein-protein interactions stabilizing the invaginations. The technological interest of this observation will be pursued by producing other membrane proteins with the help of the same type of host/plasmid.


Electron micrograph of thin sections of BL21(DE3) cells co-expressing atpF and cpxA, encoding both inner membrane proteins.  Membrane invaginations appear localized at the septum of dividing bacteria.