SoxRS-mediated lipopolysaccharide modification enhances resistance against multiple drugs in Escherichia coli

Abstract

Lipopolysaccharide (LPS) is a major constituent of the outer membrane of gram-negative bacteria that serves as a barrier against harmful molecules, including antibiotics. The waaYZ locus that encodes the LPS core biosynthetic function in Escherichia coli was found to be induced strongly by superoxide generators but not by H(2)O(2), ethanol, or heat shock. This induction was dependent on SoxRS, a superoxide and nitric oxide sensing system, through a soxbox in the waaY promoter that binds SoxS. A DeltawaaYZ mutant became more sensitive to some superoxide generators, and the activation of SoxR by these drugs became more sensitized in the mutant. Through phenotypic microarray analysis, we found that the mutant became sensitive to a wide variety of chemicals not restricted to oxidizing agents. We found that the mutant is under envelope stress and is altered in LPS composition, as monitored by the level of sigma(E) activation and changes in the electrophoretic mobility of LPS, respectively. waaY expression was also regulated by MarA (multiple-antibiotic resistance regulator), which shares a binding site (soxbox) with SoxS, and was induced by salicylate, a nonoxidative compound. These results demonstrate a novel way of protecting gram-negative bacteria against various compounds by modifying LPS, possibly through phosphorylation. Since either oxidant or nonoxidant compounds elicit resistance toward themselves and other toxic drugs, this mechanism could serve as an efficient way for pathogenic bacteria to enhance survival during antibiotic treatment within an oxidant-rich host immune environment.

FIG. 1.

waa gene cluster of E. coli K-12, location of the paraquat-inducible promoter, and structure of the LPS core region and action site of WaaY. (A) waa gene cluster and waaYZU region. The small arrow in waaR indicates the location of the H73 promoter. (B) Structure of the LPS core region of E. coli K-12 and genes involved in biosynthesis, shown at their approximate sites of action. P, phosphate; Hep, heptose; Glc, glucose; Gal, galactose; KDO, 3-deoxy-d-manno-2-octulosonic acid; PEtN, 2-aminoethyl phosphate. The proposed site of waaY action is circled.

FIG. 2.

Primer extension analysis and sequence of waaYp region. (A) Primer extension was carried out with total RNAs from paraquat-treated (0.8 mM) and untreated wild-type cells. +1 sites are indicated by arrows and asterisks. Growth conditions for RNA extraction were the same as those for the β-galactosidase assay. Induction was quantified as 19-fold by phosphorimager analysis (Bio-Rad). (B) Sequence of the waaY promoter region and details.

FIG. 3.

Response of waaYp to various chemicals. Single-copy waaYp-lacZ fusion cells (JH101) in early exponential phase (OD₆₀₀ = 0.2) were treated aerobically with various concentrations of chemicals for 1 h at 37°C and then assayed for β-galactosidase activity (Miller units). The concentration of ethanol is indicated separately (%). These are the most representative results from multiple measurements.

FIG. 4.

soxRS-dependent induction of waaY. (A) waaYp activity was assayed in various soxRS and oxyR mutant backgrounds, using single-copy lacZ fusions. The isogenic wild type (WT) is JH101, and all mutants are derivatives of JH101: the sox-8::cat mutant (JH201) has a deletion mutation of soxRS, the soxS3::Tn10 mutant (JH301) has an insertion mutation of soxS, the soxR4::cat mutant (JH401) has a constitutive soxR mutation, and the oxyR::kan mutant (JH501) has an insertion mutation of oxyR. β-Galactosidase activity was assayed after paraquat (PQ) treatment for 1 h with vigorous aeration. (B) Northern analysis was done with RNAs from the wild type (GC4468) and the soxRS mutant (BW829). An SspI-SspI fragment including the entire waaY open reading frame was used as a probe (Fig. 1A). A 1.6-kb transcript was induced 11.3-fold (quantified by phosphorimaging). (C) A 98-bp AluI-AluI fragment, from positions −101 to −4 (Fig. 2B), was used as a probe for a gel shift assay with increasing amounts of purified SoxS. Lanes 1 to 3, 0, 65, and 130 ng of purified SoxS, respectively (0, 250, and 500 nM, respectively); lanes 4 to 8, 130 ng of SoxS (500 nM), with nonspecific competitor in 65-, 130-, and 650-fold molar excess (lanes 4 to 6, respectively) or with specific competitor in 5- and 10-fold molar excess (lanes 7 and 8, respectively) over the labeled probe.

FIG. 5.

Sensitivity of waaY mutant to superoxide stress. (A) Different amounts of wild-type (GC4468), ΔwaaYZ (JH1003), and ΔwaaY_inf cells were grown on LB plates containing superoxide generators for 16 h, and growth was compared. For complementation, a plasmid expressing WaaY (pWaaY) was transformed into the ΔwaaYZ and ΔwaaY_inf mutants, and the growth of the transformants was compared. Vec, empty plasmid control. (B) soxSp-lacZ fusions in the wild-type (MS1343) and ΔwaaYZ (JH2003) backgrounds were assayed for β-galactosidase activity to monitor SoxR activity after menadione, plumbagin, and paraquat treatment for 1 h. For complementation, a plasmid carrying waaYZ (pWaaYZ) was transformed into the wild-type and ΔwaaYZ strains, and cells were assayed for β-galactosidase activity. To see the SoxR dependence of the activation, a soxR null mutation (BW900) was introduced into both the wild-type and ΔwaaYZ strains and assayed for β-galactosidase activity.

FIG. 6.

Changes in outer membrane structure of waaYZ mutant. (A) β-Galactosidase activity of the envelope stress reporter strain (rpoH P3-lacZ single-copy fusion) was assayed in the wild-type background (CAG16037) or ΔwaaYZ mutant background (JH3003). (B) LPSs were extracted from the wild-type (wt) (GC4468), ΔwaaY (JH1001), and ΔwaaYZ (JH1003) strains, separated by 15% SDS-PAGE, and visualized by silver staining.

FIG. 7.

waaY promoter is also induced by antibiotics through the mar system. Wild-type single-copy waaYp-lacZ fusion cells (JH103) and derivatives of the ΔsoxRS (JH203) and ΔmarA (JH603) mutants were treated with paraquat (0.1 mM) and sodium salicylate (10 mM) at early exponential phase (OD₆₀₀ = 0.2), and after 1 h of incubation with vigorous aeration, β-galactosidase activity was assayed. For soxbox deletion in the wild-type background, JH104 was used. We confirmed that the soxbox was correctly deleted and the −35 box remained intact by sequencing.