Identification and characterization of a high-affinity choline uptake system of Brucella abortus

Abstract

Phosphatidylcholine (PC), a common phospholipid of the eukaryotic cell membrane, is present in the cell envelope of the intracellular pathogen Brucella abortus, the etiological agent of bovine brucellosis. In this pathogen, the biosynthesis of PC proceeds mainly through the phosphatidylcholine synthase pathway; hence, it relies on the presence of choline in the milieu. These observations imply that B. abortus encodes an as-yet-unknown choline uptake system. Taking advantage of the requirement of choline uptake for PC synthesis, we devised a method that allowed us to identify a homologue of ChoX, the high-affinity periplasmic binding protein of the ABC transporter ChoXWV. Disruption of the choX gene completely abrogated PC synthesis at low choline concentrations in the medium, thus indicating that it is a high-affinity transporter needed for PC synthesis via the PC synthase (PCS) pathway. However, the synthesis of PC was restored when the mutant was incubated in media with higher choline concentrations, suggesting the presence of an alternative low-affinity choline uptake activity. By means of a fluorescence-based equilibrium-binding assay and using the kinetics of radiolabeled choline uptake, we show that ChoX binds choline with an extremely high affinity, and we also demonstrate that its activity is inhibited by increasing choline concentrations. Cell infection assays indicate that ChoX activity is required during the first phase of B. abortus intracellular traffic, suggesting that choline concentrations in the early and intermediate Brucella-containing vacuoles are limited. Altogether, these results suggest that choline transport and PC synthesis are strictly regulated in B. abortus.

Fig 1

B. abortus choline uptake kinetics. (A) Samples containing 5 × 10⁸ B. abortus 2308 cells pregrown in GW medium supplemented with 0, 10, or 100 μM cold choline chloride (Cho) were incubated in the presence of [³H]choline. At the indicated times, the samples were pelleted and washed, and the incorporated radioactivities were measured as indicated in Materials and Methods. (B) Samples of B. abortus 2308 (2.5 × 10⁸ cells) grown in GW medium without choline chloride were incubated for 7 min in the presence of increasing concentrations of [³H]choline-cold choline (1:100 [vol/vol]), and the incorporated radioactivities were measured as indicated in Materials and Methods. Each graphic shows a representative of three experiments.

Fig 2

Analysis of choline-binding protein candidates. (A) Comparison of B. abortus proteins containing an OpuAC domain. The E values are used to indicate the significance of the OpuAC prediction for each protein. The second column shows the percentage of identity between B. abortus OpuAC proteins and the S. meliloti high-affinity choline-binding protein, ChoX. The relevant amino acids for the overall fold of S. meliloti ChoX are highlighted, including those that participate in one disulfide bridge (Cys³³ to Cys²⁴⁷) and two cis-peptide bonds (Met⁹¹-Pro⁹² and Glu²⁰⁶-Pro²⁰⁷). Aromatic residues which are key determinants of the binding pocket (Trp⁴³, Trp⁹⁰, Tyr¹¹⁹, and Trp²⁰⁵) are indicated by asterisks above the sequences. (B) Comparison of the amino acid sequences of the B. abortus choX (BaChoX) and the proline-glycine-betaine substrate-binding protein of A. tumefaciens (AtProX) (AT5A_08575), ChoX of S. meliloti (SmChoX) (SMc02737), and BetT of Pseudomonas syringae (PsBetT) (PSPTO_5269).

Fig 3

PC synthesis is defective in the B. abortus ΔchoX mutant. (A) Lipid profile of mutants at low choline concentrations. The bacteria were grown in GW medium with sodium [¹⁴C]acetate supplemented with 0, 1, or 10 μM choline (Cho), and total lipid extracts were analyzed by monodimensional TLC. (B) Lipid profile of mutants at high choline concentrations. The bacteria were grown in GW medium with sodium [¹⁴C]acetate supplemented with 0, 100, or 500 μM Cho and processed as described above. WT, wild-type B. abortus 2308; Δ1_0226, ΔchoX, and Δ2_0502, B. abortus single-deletion mutants carrying a deletion in the gene BAB1_0226, BAB1_1593, or BAB2_0502, respectively; Δ1_0226 ΔchoX, Δ2_0502 ΔchoX, and Δ1_0226 Δ2_0502, B. abortus double-deletion mutants with mutations in the same genes mentioned above; pcs, B. abortus pcs mutant; PE, phosphatidylethanolamine; PC, phosphatidylcholine; CL, cardiolipin; PG, phosphatidylglycerol; OL, ornithine lipid.

Fig 4

Choline uptake is severely impaired in B. abortus ΔchoX. The amounts of radiolabeled choline incorporated by B. abortus 2308, the B. abortus ΔchoX mutant, the B. abortus ΔchoX (pDK51-choX) complemented strain, and the B. abortus pcs mutant are shown. The strains were incubated at 37°C with [³H]choline (0.1 μCi/reaction) for the indicated times. The curves indicate the mean values from two independent experiments. The standard error bars are smaller than the symbols.

Fig 5

Ligand-binding analysis of B. abortus ChoX. The binding affinities of B. abortus ChoX for choline and betaine were determined via intrinsic tryptophan fluorescence quenching after the addition of different ligand concentrations to purified ChoX. (A) Emission spectra of ChoX (200 nM) in the absence or presence of saturating amounts of choline (upper panel). Changes in fluorescence intensities at a fixed wavelength (317 nm) were plotted against the ligand concentration (lower panel) and were used for calculation of the dissociation constants (K_D). (B) Binding of betaine to ChoX (1 μM) under saturating concentrations of the substrate (upper panel). The changes in emissions at 340 nm were used to determine the binding affinities (lower panel). All data sets were fitted to the equation of a single-binding-site model, and each K_D was calculated as the mean value ± the standard deviation (SD) from at least three independent experiments. a.u, arbitrary units.

Fig 6

ChoX is required for the early intracellular steps of B. abortus. Intracellular replication of B. abortus 2308, B. abortus pcs, and B. abortus ΔchoX in J774.A1 macrophages (MOI of 50:1). The viabilities of the intracellular bacteria were determined at 4 h, 24 h, and 48 h postinfection by enumeration of the CFU. The data represent the relative changes in the numbers of CFU/ml at the indicated times over the CFU/ml at 4 h p.i. The values are the mean relative change of CFU/ml ± the SD from a representative experiment (of three independent measurements) performed in duplicate.