Differential effects of yfgL mutation on Escherichia coli outer membrane proteins and lipopolysaccharide

Abstract

YfgL together with NlpB, YfiO, and YaeT form a protein complex to facilitate the insertion of proteins into the outer membrane of Escherichia coli. Without YfgL, the levels of OmpA, OmpF, and LamB are significantly reduced, while OmpC levels are slightly reduced. In contrast, the level of TolC significantly increases in a yfgL mutant. When cells are depleted of YaeT or YfiO, levels of all outer membrane proteins examined, including OmpC and TolC, are severely reduced. Thus, while the assembly pathways of various nonlipoprotein outer membrane proteins may vary through the step involving YfgL, all assembly pathways in Escherichia coli converge at the step involving the YaeT/YfiO complex. The negative effect of yfgL mutation on outer membrane proteins may in part be due to elevated sigma E activity, which has been shown to downregulate the synthesis of various outer membrane proteins while upregulating the synthesis of periplasmic chaperones, foldases, and lipopolysaccharide. The data presented here suggest that the yfgL effect on outer membrane proteins also stems from a defective assembly apparatus, leading to aberrant outer membrane protein assembly, except for TolC, which assembles independent of YfgL. Consistent with this view, the simultaneous absence of YfgL and the major periplasmic protease DegP confers a synthetic lethal phenotype, presumably due to the toxic accumulation of unfolded outer membrane proteins. The results support the hypothesis that TolC and major outer membrane proteins compete for the YaeT/YfiO complex, since mutations that adversely affect synthesis or assembly of major outer membrane proteins lead to elevated TolC levels.

FIG. 1.

SDS-PAGE analysis of proteins from outer membranes (OM) and inner membranes (IM) fractionated by sucrose density gradients. Envelopes, prepared from cultures of yfgL⁺ and ΔyfgL strains grown overnight in LB, were resuspended in lysis buffer. After centrifugation, 20 fractions of 0.25 ml each were removed from the top of the gradient. Samples from fractions 1 to 17 were analyzed by SDS-PAGE, and protein bands were visualized after staining the gel with Coomassie blue. Because the running gel did not contain urea, OmpC and OmpF bands were not separated.

FIG. 2.

Analysis of proteins from the outer and inner membrane fractions of yfgL⁺ and ΔyfgL strains. Membrane fractions were obtained from sucrose density gradients, as described in the legend to Fig. 1. (A) Protein samples from the four peak outer membrane and inner membrane fractions, 13 to 16 and 3 to 6, respectively, in Fig. 1, were analyzed by SDS (urea)-PAGE. OmpC, OmpF, and OmpA were detected in the outer membrane fractions after staining the gel with Coomassie blue. LamB and TolC from the outer membrane fractions and AcrA from the inner membrane fractions were detected by Western blot analysis using antibodies specific to these proteins. (B) Protein levels from all four fractions were quantified, averaged, and graphed. Multiple fractions were analyzed to eliminate any fraction-specific bias in protein quantification.

FIG. 3.

TolC levels in various strains either expressing (+) or lacking (−) various OMPs. Proteins from whole-cell extracts of cultures grown overnight at 37°C in LB were analyzed using Western blots. TolC levels were quantified relative to maltose binding protein (MBP) and averaged from two independent Western blots. The TolC/MBP ratio from MC4100 (lane 1) was taken as 1, and values from other strains were relative to that of MC4100. Antibodies used in the Western blots were raised against the TolC-MBP chimeric protein; hence, they recognize both proteins.

FIG. 4.

Analysis of LPS from outer membranes (OM) and inner membranes (IM) of yfgL⁺ and ΔyfgL strains. Three sucrose density gradient fractions, corresponding to the peak regions of the outer (fractions 13 to 15) and inner (fractions 3 to 5) membranes in Fig. 1 from each strain, were analyzed. The first lane contains LPS isolated from wild-type (rfa⁺) cells. LPS was visualized by silver staining.

FIG. 5.

Western blot analysis of OMPs from periplasmic fractions obtained from various bacterial strains, with relevant protein compositions shown at the bottom of the figure. Bacterial cultures were grown to mid-log phase in LB at 30°C. To avoid casual contamination of OMPs in the periplasmic fraction, samples obtained after periplasmic extraction were centrifuged for 1 h at 100,000 × g. Prior to SDS-PAGE, protein samples were treated with proteinase K (PK) or left untreated. After inhibiting the protease activity, protein samples were mixed with SDS sample buffer, boiled for 5 min, and analyzed by SDS-PAGE. Membrane blots were incubated with antibodies that either recognize OmpC, OmpF, OmpA, and an unknown protein band labeled Unk (A) or TolC and MBP (B). TolC* is a characteristic proteinase K-cleaved band generated from assembled TolC.

FIG. 6.

Effects of DegP_S210A overexpression on various OMPs. The relevant characteristics of various strains used are shown at the bottom of the figure. Periplasmic fractions from bacterial cultures, grown to mid-log phase at 30°C in LB, carrying just the vector plasmid (odd-numbered lanes) or a plasmid expressing DegP_S210A (even-numbered lanes) were isolated and analyzed by Western blots to detect OmpC, OmpF, OmpA (A and B), and MBP (C). Prior to SDS-PAGE, protein samples were heated (A and C) or were left unheated (B). Expression of DegP_S210A was induced by isopropyl-β-d-thiogalactopyranoside (0.4 mM final concentration). OmpA* and Unk refer to folded OmpA and an unknown protein band, respectively.