Release of the lipopolysaccharide deacylase PagL from latency compensates for a lack of lipopolysaccharide aminoarabinose modification-dependent resistance to the antimicrobial peptide polymyxin B in Salmonella enterica

Abstract

Salmonella enterica modifies its lipopolysaccharide (LPS), including the lipid A portion, to adapt to its environments. The lipid A 3-O-deacylase PagL exhibits latency; deacylation of lipid A is not usually observed in vivo despite the expression of PagL, which is under the control of a two-component regulatory system, PhoP-PhoQ. In contrast, PagL is released from latency in pmrA and pmrE mutants, both of which are deficient in aminoarabinose-modified lipid A, although the biological significance of this is not clear. The attachment of aminoarabinose to lipid A decreases the net anionic charge at the membrane's surface and reduces electrostatic repulsion between neighboring LPS molecules, leading to increases in bacterial resistance to cationic antimicrobial peptides, including polymyxin B. Here we examined the effects of the release of PagL from latency on resistance to polymyxin B. The pmrA pagL and pmrE pagL double mutants were more susceptible to polymyxin B than were the parental pmrA and pmrE mutants, respectively. Furthermore, introduction of the PagL expression plasmid into the pmrA pagL double mutant increased the resistance to polymyxin B. In addition, PagL-dependent deacylation of lipid A was observed in a mutant in which lipid A could not be modified with phosphoethanolamine, which partly contributes to the PmrA-dependent resistance to polymyxin B. These results, taken together, suggest that the release of PagL from latency compensates for the loss of resistance to polymyxin B that is due to a lack of other modifications to LPS.

FIG. 1.

PhoP-PhoQ- and PmrA-PmrB-regulated lipid A modifications in S. enterica serovar Typhimurium. (A) Prototype lipid A of S. enterica serovar Typhimurium. (B) Modified lipid A of S. enterica serovar Typhimurium. The phosphate residues and acyl chains of lipid A of S. enterica can be derivatized in a PhoP-PhoQ- or PmrA-PmrB-regulated manner (reviewed in reference 7). Aminoarabinose and/or phosphoethanolamine groups (shown in blue) can be attached to phosphate residues, under the control of PmrA-PmrB (18, 45). Minor species were present in which the locations of the aminoarabinose and phosphoethanolamine groups were reversed or in which both phosphates were modified with the same substituent (45). Both the pmrF operon and pmrE are necessary for the PmrA-PmrB-regulated attachment of aminoarabinose to lipid A (15, 45). pmrC mediates the PmrA-PmrB-regulated attachment of phosphoethanolamine to lipid A (27). The addition of the palmitate chain is catalyzed by PagP (3, 19), the formation of the 2-hydroxymyristate group requires LpxO (13), and the deacylation at position 3 of lipid A is catalyzed by PagL (39) (shown in red). The pagL and pagP genes are regulated by PhoP-PhoQ (2), and the lpxO gene is partly regulated by PhoP-PhoQ (12, 13). PhoP-PhoQ also activates PmrA-PmrB; therefore, the aminoarabinose and phosphoethanolamine modifications occur under PhoP-PhoQ-activating conditions (18, 44).

FIG. 2.

MALDI-TOF mass spectrometry of lipid A purified from S. enterica serovar Typhimurium strains. The wild-type (ATCC 14028s), pmrA (JSG421), pmrE (KCS041), ΔpmrC (KCS180), ΔpagL (KCS216), pmrA ΔpagL (KCS208), pmrE ΔpagL (KCS209), and ΔpmrC ΔpagL (KCS210) strains were cultivated in growth medium at pH 7.4 (A) or 5.8 (B to H). The m/z values of lipid A species are shown, and those that represent deacylated lipid A species are denoted by asterisks. Insets in panels A, B, E, and H show results of MALDI-TOF mass spectrometry of lipid A using 2-5-dihydroxybenzoic acid matrices. The structural interpretations of lipid A species are summarized in Table 2.

FIG. 3.

Expression levels of PagL protein were similar among Salmonella wild-type, pmrA, pmrE, and ΔpmrC strains grown in the mild acid medium. Ten-microgram samples of membrane proteins prepared from strains cultivated in mild acid medium (pH 5.8) containing 10 μM MgCl₂ were subjected to SDS-polyacrylamide gel (12.5%) electrophoresis and analyzed with staining (A) or by Western blotting (B). Lanes: 1, wild-type strain; 2, pmrA strain (JSG421); 3, pmrE strain (KCS041); 4, ΔpmrC strain (KCS180); 5, ΔpagL strain (KCS216).

FIG. 4.

The pagL mutants were more susceptible to polymyxin B than were the parental mutants. (A) The wild-type (14028s), ΔpagL (KCS216), pmrA (JSG421), pmrA ΔpagL (KCS208), pmrE (KCS041), pmrE ΔpagL (KCS209), ΔpmrC (KCS180), and ΔpmrC ΔpagL (KCS210) strains grown in the growth medium (pH 5.8) were treated with 1 μg/ml of polymyxin B. Percent survival was determined as described in Materials and Methods. The results shown are for two independent sets of experiments (experiments 1 and 2). (B) The pmrA and pmrA ΔpagL mutant strains cultivated in the growth medium (pH 5.8) were treated with the indicated concentrations of polymyxin B. Percent survival is the average of triplicate measurements, and error bars indicate standard deviations.

FIG. 5.

Growth of S. enterica serovar Typhimurium strains in the presence of polymyxin B. The wild-type (14028s), pmrA (JSG421), and pmrA ΔpagL (KCS208) strains cultivated in the growth medium (pH 5.8) as described in Materials and Methods were diluted 1:10 with fresh growth medium (pH 5.8). Then, the cells were grown at 37°C in the presence (A) or absence (B) of 1 μg/ml of polymyxin B. The results shown are representative of at least two independent experiments. OD₆₀₀, optical density at 600 nm.

FIG. 6.

MALDI-TOF mass spectrometry of lipid A purified from pagL-null mutant strains transformed with a PagL expression plasmid. Wild-type (ATCC 14028s), pmrA (JSG421), and pmrA ΔpagL (KCS208) strains transformed with a PagL expression construct pWLP23 or the control vector pWKS30 were grown in the growth medium (pH 5.8), and their lipid A was analyzed with a MALDI-TOF mass spectrometer. The m/z values of lipid A species are shown, and those that represent deacylated lipid A species are denoted by asterisks. The structural interpretations of lipid A species are summarized in Table 2.

FIG. 7.

Introduction of PagL expression plasmids increased the resistance of the pagL-null mutant strain to polymyxin B. Wild-type (ATCC 14028s), pmrA (JSG421), and pmrA ΔpagL (KCS208) strains transformed with the control vector (pWKS30) or PagL expression construct (pWLP23) were cultivated in the growth medium (pH 5.8) and then treated with 0.9 or 1.0 μg/ml of polymyxin B. Percent survival is the average of triplicate measurements, and the error bars indicate standard deviations.

FIG. 8.

The attachment of phosphoethanolamine inhibited the deacylation of lipid A. The pmrE (KCS041) or pmrA (JSG421) strain transformed with pBAD24 (control vector) or pKK28 (PmrC) was cultivated in growth medium (pH 7.4) containing 0.2% (wt/vol) arabinose to induce PmrC expression under the control of the P_BAD promoter, and then its lipid A was analyzed with a MALDI-TOF mass spectrometer. The m/z values of lipid A species are shown, and those that represent deacylated lipid A species are denoted by asterisks. The structural interpretations of lipid A species are summarized in Table 2.