The Burkholderia cenocepacia BDSF quorum sensing fatty acid is synthesized by a bifunctional crotonase homologue having both dehydratase and thioesterase activities

Abstract

Signal molecules of the diffusible signal factor (DSF) family have been shown recently to be involved in regulation of pathogenesis and biofilm formation in diverse Gram-negative bacteria. DSF signals are reported to be active not only on their cognate bacteria, but also on unrelated bacteria and the pathogenic yeast, Candida albicans. DSFs are monounsaturated fatty acids of medium chain length containing an unusual cis-2 double bond. Although genetic analyses had identified genes involved in DSF synthesis, the pathway of DSF synthesis was unknown. The DSF of the important human pathogen Burkholderia cenocepacia (called BDSF) is cis-2-dodecenoic acid. We report that BDSF is synthesized from a fatty acid synthetic intermediate, the acyl carrier protein (ACP) thioester of 3-hydroxydodecanoic acid. This intermediate is intercepted by protein Bcam0581 and converted to cis-2-dodecenoyl-ACP. Bcam0581 is annotated as a homologue of crotonase, the first enzyme of the fatty acid degradation pathway. We demonstrated Bcam0581to be a bifunctional protein that not only catalysed dehydration of 3-hydroxydodecanoyl-ACP to cis-2-dodecenoyl-ACP, but also cleaved the thioester bond to give the free acid. Both activities required the same set of active-site residues. Although dehydratase and thioesterase activities are known activities of the crotonase superfamily, Bcam0581 is the first protein shown to have both activities.

Fig. 1

Structures, alignments, the crotonase and proposed Bcam 0581 dehydratase reactions. A. Structures of BDSF (cis-2-dodecenoic acid) and DSF (11-methyl-cis-2-dodecenoic acid). B. Relevant partial alignments of rat liver mitochondrial crotonase, Bcam0581 and RpfF. The solid circles denote the catalytic glutamate residues whereas the arrows denote the oxyanion hole residues. C. The crotonase reaction. The physiological crotonase reaction is to the right. D. The proposed Bcam0581 dehydratase reaction. The formal names of BDSF and DSF are 2-(Z)-dodecenoic acid and 11-methyl-2-(Z)-dodecenoic acid, respectively.

Fig. 2. Purification and solution structure of recombinant Bcam0581 protein

A. Size exclusion chromatography of the His-tagged Bcam0581 protein expressed in E. coli. The inset is a 10% SDS-PAGE analysis of the proteins of the peaks 1 and 2 where M, is a protein standard marker (BioRad). The elution peaks of the molecular weight standards are given at the top of the panel. B. Chemical cross-linking analyses of the purified Bcam0581 protein where EGS denotes ethylene glycol bis succinimidylsuccinate (Experimental procedures). The samples were separated by 10% SDS-PAGE where M is the Precision Plus Protein Standard (BioRad).

Fig. 3. Bcam0581 catalyzes the hydrolysis of acyl-ACP thioesters

A. Acyl-ACPs were first prepared as described in Experimental procedures. The reaction mixture for assays of Bcam0581 thioesterase activity contained 0.1 M Tris-HCl (pH7.5), 2 mM β-mercaptoethanol, 0.2 µg Bcam0581 and 20 µM acyl-ACP (3-hydroxydodecanoyl-ACP, trans-2-dodecenoyl-ACP or cis-2-dodecenoyl-ACP). The assay mixtures were incubated at 37°C for 30 min and the reaction products were resolved by conformationally sensitive gel electrophoresis on 18% polyacrylamide gels containing a concentration of urea optimized for the separation (Post-Beittenmiller et al., 1991). B. Bcam0581thioesterase activity.

Fig. 4. Mutagenesis indicates that thioesterase activity is intrinsic to Bcam0581 and involves essential residues conserved in crotonase

Acyl-ACPs were prepared and Bcam0581 variants were purified as described in Experimental procedures. Bcam0581 thioesterase activity assays were performed using (Panel A) 3-hydroxydodecanoyl-ACP (3-OH-C12:0-ACP), (Panel B) cis-2-dodecenoyl-ACP (cis-2-C12:1-ACP), (Panel C) trans-2-dodecenoyl-ACP (trans-2-C12:1-ACP) and (panel D) dodecanoyl-ACP (C12:0-ACP) as substrates. The reactions were initiated by the addition of 0.2 µg of wild type Bcam0581 or one of the the mutant proteins carrying amino acid substitutions E158A, E158Q, G83P, G135P, E138A or E138Q. In panels A and B the reaction mixtures were incubated at 37°C for 10 min whereas panels C and D are 30 min incubations. The reaction products were resolved by conformationally sensitive gel electrophoresis as in Fig. 3.

Fig. 5. Functional characterization of Bcam0581 and its mutant derivatives in E. coli

Panel A. Argentation thin-layer chromatographic analysis of [1-¹⁴C] acetate-labeled esters present in the medium from E. coli strain K19 carrying either pBHK06 (Bcam0581m) or its mutant derivatives as given. Cultures of plasmid-containing strains were induced with IPTG. The fatty acid methyl esters were obtained from the medium as described in Experimental procedures. The methyl esters were then separated by argentation thin layer chromatography followed by autoradiography. The migration positions of the fatty acid species are shown. Sat, saturated fatty acid esters; UFA, unsaturated fatty acid esters. Panel B. BDSF bioassay of E. coli DH5α expressing Bcam0581 or the mutant derivatives. The formation of a blue halo due to hydrolysis of 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid around the site of inoculation indicates the presence of DSF-like activity. pQE is the vector (pQE-2) control whereas pQEBcam expressed wild type Bcam0581. The mutant proteins were also expressed in pQE-2.

Fig. 6

AasS allows production of cis-2-C12:1-ACP by Bcam0581by reversal of thioesterase action. A. in vitro Bcam0581 reaction with the addition of AasS using 3-hydroxydodecanoyl-ACP (3-OH-C12:0-ACP) as substrate. Lanes 3–8 are reactions in which Bcam0581 and AasS were added together. The reaction mixture contained 0.1 M Tris-HCl (pH7.5), 2 mM β-mercaptoethanol, 20 µM 3-hydroxydodecanoyl-ACP, 10 mM ATP, 10 mM MgCl₂ and 0.2 µg purified Bcam0581 and various amounts of AasS in a final volume of 30 µl. The assay mixtures were incubated at 37°C for either 90 min or 4 h. Lane 9 is a reaction as described above except with Bcam0581 alone (AasS was not added) and incubation at 37°C for 10 min. Lanes 10 and 11 are boiled samples of the reactions of Lanes 8 and 9, respectively, in which additional AasS was added to convert any free fatty acid to acyl-ACPs. These assay mixtures were incubated at 37°C for additional 2 h after addition of 4 µg AasS, 10 mM ATP and 10 mM MgCl₂. Boiling precipitated the high molecular weight proteins whereas the ACP species remained soluble. B. Reconstruction of fatty acid synthesis in vitro showing the new band is derived from 3-hydroxydodecanoyl-ACP. Lanes 2–4, ¹⁴C-labeled 3-hydroxydodecanoyl-ACP was synthesized in 20 min reactions as described in Experimental procedures. Then either Bcam0581 (0.2 µg) alone or Bcam0581 (0.2 µg) plus AasS (4 µg) were added to the reaction mixtures of lanes 3 and 4, respectively, followed by incubation at 37°C for additional 30 min. Note that the reaction mixture in lane 4 also contains 10 mM ATP and 10 mM MgCl₂. The reaction products were analyzed as in Figs. 3 and 4. Note that two irrelevant lanes were excised from the middle of this radiogram to simplify the figure. C. Mutagenesis of the Bcam0581 active site residues impairs dehydratase activity in vitro. In vitro Bcam0581 or its variants (E158A, G83P, G135P and E138A) reactions with 3-hydroxydodecanoyl-ACP (3-OH-C12:0-ACP) as substrate. All reactions contained 0.2 µg of a Bcam0581 protein and 4 µg of AasS and were performed as in A.

Fig. 7. Identification of the reaction products formed in vitro

A. NADH oxidation activity of EcFabI using different substrates. NADH oxidation was monitored at 340 nm for the EcFabI reaction using trans-2-dodecenoyl-ACP(■), synthetic cis-2-dodecenoyl-ACP (●) or purified cis-2-dodecenoyl-ACP (○) from in vitro reactions run with simultaneous addition of Bcam0581 and AasS as described in Experimental procedures. Note that the ● and ○ symbols largely overlap. The □ symbol denotes background without addition of substrates. The controls without substrate showed no significant change in absorbance. The curves have been adjusted to the same zero time absorbance in the figure. The data are the means± standard error of the mean of three independent assays. B. BDSF bioassay. Each well contains 10 µl of BDSF (either extracted from reaction mixtures or synthetic), trans-2-dodecenoic acid or 3-hydroxydodecanoic acid at a concentration of 60 µM. Synthetic BDSF was the positive control. C. BDSF induction of endoglucanase expression in the DSF biosensor strain, Xcc8523 (pL6engGUS). The biosensor strain in which E. coli gusA encoding β-D-glucuronidase is fused to the engXCA promoter was cultured and exposed to BDSF samples. Samples of the cultures were collected at different time points after the addition (final concentration 60 µM) of trans-2-dodecenoic acid (■), synthetic BDSF (●) or BDSF extracted (○) from in vitro reactions. The in vitro reactions received simultaneous additions of Bcam0581 and AasS. □ denotes background without any additions. The β-D-glucuronidase activities were determined as described in Experimental procedures. The data are the means± standard error of the mean of three independent assays. D. GC-MS chromatogram from analyses of the fatty acid extracted from a reaction mixture and synthetic BDSF. The fatty acids were derivatized first to their methyl esters and then to their dimethyl disulfide adducts which were analyzed by gas chromatography-mass spectroscopy. The elution profiles of DMDS adducts of the methyl esters show mass chromatogram peaks of m/z 187 and m/z 306. E. Mass spectroscopy of the cleavage products of the dimethyl disulfide adducts of fatty acid methyl esters either extracted from a reaction mixture or synthetic BDSF. The unsaturated esters gave a cleavage fragment of m/z 187 (numbers in ovals) corresponding to the methyl end of the molecule plus a second fragment m/z 120 (numbers in squares) corresponding to the ester end of the molecule (expected m/z 119).

Fig. 8. Phylogeny of Bcam0581

The minimum evolution tree of selected Pfam00378 protein sequences with bootstrap percentage confidence values shown for each branch is given. Phylogenetic analyses were conducted as described in Experimental procedures. The scale bar corresponds to a 50% difference in compared residues, on average, per branch length. Protein sequences from the malonyl-CoA decarboxylase superfamily, Pfam06833, were used as a related out-group. Each branch is labeled with the Uniprot sequence identifier followed by the species identifier. “B4EKM5 BURCJ” corresponds to Bcam0581 and is in bold.

Fig. 9. Current models of BDSF biosynthesis in B. cenocepacia

BDSF biosynthesis is catalyzed by Bcam0581. A. In the presence of acyl-ACP synthetase (AasS) Bcam0581 functions as a dehydratase, Bcam (D), to convert 3-hydroxydodecanoyl-ACP to cis-2-dodecenoyl-ACP and then as a thioesterase, Bcam(T) to release free BDSF. Although we did not detect cis-2-C12:1 in the absence of AasS, a coupled reaction occurred in vivo in E. coli. B. The intersection of BDSF synthesis and fatty acid synthesis. Bcam0581 taps off a portion of the 3-hydroxydodecanoyl-ACP intermediate of long chain fatty acid synthesis