Resistance to oxidative stress by inner membrane protein ElaB is regulated by OxyR and RpoS

Abstract

C-tail anchored inner membrane proteins are a family of proteins that contain a C-terminal transmembrane domain but lack an N-terminal signal sequence for membrane targeting. They are widespread in eukaryotes and prokaryotes and play critical roles in membrane traffic, apoptosis and protein translocation in eukaryotes. Recently, we identified and characterized in Escherichia coli a new C-tail anchored inner membrane, ElaB, which is regulated by the stationary phase sigma factor RpoS. ElaB is important for resistance to oxidative stress but the exact mechanism is unclear. Here, we show that ElaB functions as part of the adaptive oxidative stress response by maintaining membrane integrity. Production of ElaB is induced by oxidative stress at the transcriptional level. Moreover, elaB expression is also regulated by the key regulator OxyR via an OxyR binding site in the promoter of elaB. OxyR induces the expression of elaB in the exponential growth phase, while excess OxyR reduces elaB expression in an RpoS-dependent way in the stationary phase. In addition, deletion of elaB reduced fitness compared to wild-type cells after prolonged incubation. Therefore, we determined how ElaB is regulated under oxidative stress: RpoS and OxyR coordinately control the expression of inner membrane protein ElaB.

Figure 1

Expression of ElaB is induced by oxidative stress. A. Overnight cultures of BW25113 wild type (WT) were diluted to a turbidity of 0.05 at 600 nm and cultured at 37°C to a turbidity of 1.0; then, 10 mM H₂O₂ was added for 10 min. The expression levels of elaB, oxyR and elaA were quantified, and fold changes were calculated. All the fold changes in genes were normalized to oxyR in cells without H₂O₂ treatment. For statistical analysis, P < 0.01 is shown in **. B. ElaB was fused with 2× Flag before the stop codon, and cells were cultured and treated with 5 mM H₂O₂ at the indicated time points. The expression levels of ElaB‐Flag and OxyR‐Flag were determined with Western blotting with the same amount of total protein (upper and middle panel). The expression levels of ElaA‐Flag under the same conditions were used as a negative control (lower panel).

Figure 2

ElaB mutation reduces cell membrane integrity during oxidative stress. BW25113 wild‐type (WT) cells were cultured in the same condition as shown in Fig. 1A. A. Live/Dead staining was performed (live cells appear green, and dead cells appear red/yellow), and the percentages of dead cells were calculated. Cells that were not treated with H₂O₂ were used as controls. B. The cells were stained with lipid peroxidation‐specific dye C11‐BODIPY. The upper panels indicate lipid oxidation in the cell membrane, and the lower panels indicate bright‐field views of corresponding upper panels. Percentages of cells with lipid peroxidation were calculated. In A and B, 1000 cells in each culture were observed, and only one representative image for each strain is shown.

Figure 3

ElaB is regulated by OxyR in E. coli. A. The promoter region of elaB and the sequences of the probe containing the putative OxyR binding sites are shown. The numbers indicate the locations relative to the start codon A of elaB. The predicted binding sites of OxyR are marked. The −10 and −35 regions are highlighted in green and light blue. The transcriptional start site (TSS) is marked with an arrow. The ribosome binding site (RBS) is also highlighted in grey. The start codon of elaB is shown in red letters. For the promoter activity assay, the open reading frame (ORF) of elaB was replaced by lacZ ORF. B. WT and ΔoxyR harbouring pHGR01‐PelaB‐L (containing OxyR binding sites 1 and 2), pHGR01‐PelaB‐S (only containing OxyR binding site 2) and pHGR01‐PelaB‐SM (mutation of OxyR binding site 2 in pHGR01‐PelaB‐S) cells in the exponential growth phase were collected, and β‐galactosidase activities were evaluated. C. Complementation of oxyR via pCA24N‐oxyR restored the promoter activity of elaB during the exponential growth phase rather than during the stationary phase. For statistical analysis, P < 0.01 is marked as **. D. OxyR binds to the DNA probe (PelaB‐S) containing the binding site 2 in a concentration‐dependent manner (lanes 1–6). The addition of unlabelled probe reduced the binding of OxyR to the labelled probe in a concentration‐dependent manner (lanes 7–10). E. OxyR was unable to bind to the mutant DNA probe (PelaB‐SM) under the same conditions.

Figure 4

Promoter activity of elaB is regulated by OxyR in an RpoS‐dependent manner. (A) Production of ElaB‐Flag was determined using Western blotting for the BW25113 wild type (WT), ΔoxyR, ΔrpoS and ΔΔ cells. Same amount of total protein was loaded in each lane. The expression plasmids pCA24N‐oxyR and pCA24N‐rpoS were transferred into WT (B) and ΔoxyR (C) cells. Production of OxyR‐His (red arrows) and RpoS‐His (green triangles) was induced by 0.5 mM IPTG at OD ₆₀₀ ~ 0.1 for 2 h and 6 h. Cm indicates the chloramphenicol resistance protein. The levels of OxyR‐His, RpoS‐His and ElaB‐Flag were determined using Western blotting. Same amount of total protein was loaded in each lane. (D) ΔoxyR/pHGR01‐PelaB‐L cells expressing RpoS were induced with 0.5 mM IPTG for exponential phase and stationary phase, and β‐galactosidase activities were tested. The pCA24N vector was used as a negative control. (E) The ΔrpoS/pHGR01‐PelaB‐L cells expressing oxyR and rpoS were induced, and β‐galactosidase activities were determined as described in D. For statistical analysis, P < 0.01 is marked as **.

Figure 5

ElaB increases cell fitness. A. Growth of BW25113 wild type (WT), ΔelaB, ΔoxyR and ΔrpoS. B. Overnight cultures of WT and ΔelaB::km were diluted to OD ₆₀₀ 0.1 and were cultured till OD ₆₀₀ 1.0. Then, different ratios of ΔelaB::km and WT were mixed, and the percentages of ΔelaB::km in total cells were determined at different time points. C. The ΔrpoS::km cells were mixed with WT and ΔelaB at the ratio of 1:1, and the percentages of ΔrpoS::km in total cells were determined at different time points. D. The ΔrpoS::km cells in (C) were replaced by ΔoxyR::km, and the percentages of ΔoxyR::km in total cells were determined at different time points.