Specificity of Acyl-Homoserine Lactone Synthases Examined by Mass Spectrometry
http://www.100md.com
《细菌学杂志》
Department of Pharmacology, Program in Biomolecular Structure, The University of Colorado Health Sciences Center, P.O. Box 8511 MS8303, Aurora, Colorado 80045
ABSTRACT
Many gram-negative bacteria produce a specific set of N-acyl-L-homoserine-lactone (AHL) signaling molecules for the purpose of quorum sensing, which is a means of regulating coordinated gene expression in a cell-density-dependent manner. AHLs are produced from acylated acyl-carrier protein (acyl-ACP) and S-adenosyl-L-methionine by the AHL synthase enzyme. The appearance of specific AHLs is due in large part to the intrinsic specificity of the enzyme for subsets of acyl-ACP substrates. Structural studies of the Pantoea stewartii enzyme EsaI and AHL-sensitive bioassays revealed that threonine 140 in the acyl chain binding pocket directs the enzyme toward production of 3-oxo-homoserine lactones. Mass spectrometry was used to examine the range of AHL molecular species produced by AHL synthases under a variety of conditions. An AHL selective normal-phase chromatographic purification with addition of a deuterated AHL internal standard was followed by reverse-phase liquid chromatography-tandem mass spectrometry in order to obtain estimates of the relative amounts of different AHLs from biological samples. The AHLs produced by wild-type and engineered EsaI and LasI AHL synthases show that intrinsic specificity and different cellular conditions influence the production of AHLs. The threonine at position 140 in EsaI is important for the preference for 3-oxo-acyl-ACPs, but the role of the equivalent threonine in LasI is less clear. In addition, LasI expressed in Escherichia coli produces a high proportion of unusual AHLs with acyl chains consisting of an odd number of carbons. Furthermore, these studies offer additional methods that will be useful for surveying and quantitating AHLs from different sources.
INTRODUCTION
Many gram-negative bacteria produce acyl-homoserine lactones (AHLs) for a type of intercellular signaling known as quorum sensing, which is commonly associated with their pathogenic or communal behavior (reviewed in references 39 and 64). At least 70 different bacterial species produce AHLs, which can vary in acyl chain length from C4 to C18. In addition, AHLs can have different degrees of unsaturation at the C-7 or C-8 position, as well as oxidation at the 3 position (Fig. 1) (15, 36). Most bacterial species produce a restricted range of AHLs that is strongly coupled to the AHL selectivity of their own detection machinery, which uses transcriptional regulators, known as R-proteins, that selectively bind the AHL to either activate or repress many downstream target genes (53, 55, 60). This apparent coupling has led to the notion that AHL production is relatively restricted so that signaling within the species can be distinguished from interspecies signaling (21).
The intrinsic specificity of the AHL synthase for subsets of acyl-acyl carrier protein (acyl-ACP) substrates is an important factor in the production of particular AHLs. The AHLs are synthesized in the cell by the AHL synthase enzyme from substrates that are common components of cellular metabolism, acyl-ACP and S-adenosyl-L-methionine (40, 46). The Pantoea stewartii enzyme EsaI and Pseudomonas aeruginosa enzyme LasI preferentially produce 3-oxo-C6-homoserine lactone (HSL) and 3-oxo-C12-HSL, respectively (1, 48). Structural studies of these two enzymes revealed a common binding site for the acyl-ACP phosphopantetheine prosthetic group, and modeling studies implicated a particular threonine in the acyl-chain binding site as being important for hydrogen bonding to the 3-oxo position of acyl-ACP (17, 62). Furthermore, comparison of AHL synthase sequences and the AHLs they preferentially produce suggested that a threonine at that position correlates with the production of 3-oxo-HSLs, whereas alanine and glycine correlate with unsubstituted AHLs and serine correlates with the production of 3-hydroxy-HSLs. The AHLs produced by EsaI and a threonine 140-to-alanine substitution mutant, identified using two sensitive reporter bioassays, show a dramatic shift in AHL production from 3-oxo-C6-HSL in the wild type (WT) to C6-HSL in the mutant (62). Although this assay confirmed the role of this residue in AHL synthase specificity for 3-oxo-acyl-ACP, the true specificity could not be assessed using this AHL detection method.
AHLs can be detected effectively using AHL-sensitive bioassays in solution or in a thin-layer-chromatography (TLC) overlay format. The TLC overlay assay is widely used to determine the types of AHLs produced by a particular bacterium, or more specifically by a particular AHL synthase (37, 55). These bioassays are highly sensitive, with the ability to detect sub-picomole amounts of particular AHLs. However, there is an inherent detection bias, because they rely on the particular specificity of the reporter strain that is used for AHL detection, such as Chromobacterium violaceum (37), Agrobacterium tumefaciens (TraR) (55), or Pseudomonas aeruginosa (LasR) (48). Recent developments in reporter strains have provided some relatively promiscuous detection systems that recognize a much wider variety of AHLs (49, 66). However, for complete coverage of known AHLs, multiple reporter strains must be used with a full complement of standards to assess which spots correspond to each particular signal (7).
Mass spectrometry (MS) and gas chromatography (GC)-MS offer alternative methods of AHL detection, which are based on the physical/chemical properties of the compounds, such as the mass/charge ratio of molecular ions, collisionally induced product ions, and chromatographic retention properties. Derivatization of AHLs has been used for the examination of constellations of 3-oxo-HSLs by GC-MS (8). Direct analysis of AHLs has been reported using GC-MS (6) and high-performance liquid chromatography (HPLC) coupled to ion trap mass spectrometry (LC-MS) (36, 41, 44, 48, 59). The addition of a preconcentration step has been reported to result in highly sensitive methods of AHL detection (13, 14).
Here we describe a combination of approaches that show how distributions of AHLs change with mutations introduced into two AHL synthases, EsaI and LasI. The distribution of AHLs from experimental samples was determined by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC/MS/MS) using a triple-quadrupole mass spectrometer. We examined the influence of the active site threonines of EsaI and LasI in the specificity of AHL synthesis, as well as the contributions of different bacterial backgrounds to the variety of AHLs produced by these AHL synthases in vivo.
MATERIALS AND METHODS
AHL nomenclature. For brevity, an acyl chain of a specific length is denoted by "Cn" for the number of carbons in the chain, rather than the chemical names (i.e., hexanoyl is C6), and the substitution type and position as 3-oxo or 3-hydroxy (eg., 3-oxo-C6-HSL). D3 indicates that the AHL contains three deuterium atoms at the terminal position of the acyl chain.
Synthesis of D3-C6-homoserine lactone. Solid-phase carbodiimide chemistry was used to synthesize D3-C6-HSL, using N-cyclohexylcarbodiimide-N'-propyloxymethyl polystyrene resin (PS-carbodiimide; Argonaut Technologies) (12, 45). The fatty acid D3-hexanoic acid (34.5 μmol [4.11 mg]) (Cambridge Isotope) was added to 400 μl of methylene chloride (CH2Cl2) with 10% N,N-dimethylformamide in a separate glass reaction vial. To this reaction vial, 40 mg of PS-carbodiimide was added and the mixture was stirred for 5 min at room temperature. After the fatty acid had been allowed to equilibrate with the resin, 4.2 mg (23 μmol) of -amino--butyrolactone hydrobromide (HSL-HBr) (Aldrich) dissolved in 200 μl of CH2Cl2 with 10% N,N-dimethylformamide and 4.6 μmol (6.4 μl) of triethylamine (subsequent analysis of products revealed no significant lactonolysis when triethylamine was used as the base [data not shown]) was added and stirred for 30 h at room temperature. The reaction workup consisted of filtering out the resin-bound substrate/intermediate, coupled to solid-phase extraction as described below for AHL extracts. The D3-C6-HSL product was resuspended in methanol to a final concentration of 200 μM. This dilution was shown by LC/MS to be equivalent to 0.2 nmol of C6-HSL and was used for all remaining experiments (see Fig. 4B).
Mutagenesis of the LasIG plasmid. LasIG codon substitutions were made from pLasIG using the QuikChange mutagenesis method (Stratagene) with the complementary primers for each mutation shown in Table 1. The mutations were confirmed by restriction digestion analysis, where possible, and by DNA sequencing of the entire gene.
Extraction and purification of acyl-homoserine lactones for mass spectrometry analysis. E. coli strains (Table 1), harboring expression plasmids for the AHL synthases EsaI from P. stewartii and pLasI and pLasIG from P. aeruginosa (18, 61), were grown in 5-ml cultures in Luria-Bertani (LB) broth with 100 μg/ml of ampicillin at 37°C with shaking for 12 to 16 h. Each culture was diluted 1:50 into 10 ml of fresh LB broth with 100 μg/ml ampicillin and incubated until the cell density reached an optical density at 600 nm (OD600) of 0.6 to 0.8. The temperature was lowered to 25°C for 15 min, and after equilibration, each culture was induced by the addition of IPTG (isopropyl--D-thiogalactopyranoside) to 0.5 mM. Cultures were incubated at 25°C for 6 to 8 h with shaking. The P. aeruginosa cultures were grown in the same way, except that strain PAO1 required no antibiotic and strain PAO214 with pEX30-las required 500 μg/ml carbenicillin. Culture supernatants were processed by centrifugation at 3,200 x g for 10 min, followed by decanting into a 20-ml syringe and passing through a 0.2-μm filter. For samples being prepared for quantitative analysis, 0.4 nmol of the synthetic D3-C6-HSL was added per 10 ml of culture after harvesting the supernatant but prior to filtering it. The culture supernatants were extracted two times, with 10 ml of acidified ethyl acetate (0.1 ml/liter acetic acid), and 9 ml from each extraction was pooled and taken to complete dryness in 2-ml glass sample vials. This extraction procedure was expected to extract at least 75% of the 3-oxo-C6-HSLs. AHLs were also obtained from commercial sources (Sigma, Quorum Sciences Inc.).
Initial purification of the AHL molecular species was accomplished by solid-phase extraction (SPE). Each sample, which had been redissolved in 100 μl of methanol (Optima, Sigma), was applied to activated Sep Pak Plus silica cartridges (Waters), which were fitted with a 6-ml glass reaction vial and attached to the vacuum manifold. Cartridges had been activated by successive washings of 6 ml of each solvent in the following sequence: equal volumes of isooctane and ethyl ether (ethyl ether must not have the preservative butylated hydroxytoluene), acidified ethyl acetate, and isooctane and ethyl ether. The sample in methanol was added to 5 ml isooctane-ethyl ether, and was pipetted into the reaction vial and then loaded onto the SPE cartridge. The SPE cartridge was washed twice with 6 ml of isooctane-ether and then eluted into a glass fraction tube with 5 to 8 ml of acidified ethyl acetate. The purified samples were taken to near dryness by vacuum evaporation and then transferred to a glass autosampler vial to be evaporated completely. The samples were redissolved in 50 or 100 μl of methanol and capped.
Hexane can be used in place of isooctane as a solvent in the purification procedure, but impurities in the reagent-grade hexane yielded electrospray ionization ions of m/z 284 and 228 in the AHL analysis. Therefore, hexane was prepurified by passage over the SPE cartridge prior to being used in the purification of AHLs from cell culture extracts. The true AHLs were identified by LC/MS/MS in a multiple reaction monitoring (MRM) experiment, where the transition from the parent ion to both the acyl and lactone moiety peaks overlapped by comparison to known standards and by evaluation of chromatographic retention times. Retention time analysis was used to distinguish between the C(n)-HSL and the 3-oxo-C(n-1)-HSL molecular species, which have similar masses (data not shown).
LasR reporter assay. The LasR reporter strain E. coli MG4/pKDT17a (48, 52) was grown overnight in A medium (52), supplemented with 1 mM MgSO4 and 100 μg/ml ampicillin at 30°C. Each extract was incubated with 1 ml of diluted culture (diluted to an OD600 of 0.1) for 5 to 6 h at 30°C. The samples were then subjected to analysis of -galactosidase activity using the Miller assay (16, 38).
High-performance liquid chromatography. Samples for LC/MS and LC/MS/MS analysis were resuspended in 100 μl of methanol, and 10 μl was injected onto a 2.0-mm by 150.0-mm Columbus C18 reverse-phase column (Phenomenex) operated at a flow rate of 200 μl/min with the effluent flowing directly into the mass spectrometer. Solvent A consisted of H2O containing 0.1% glacial acetic acid, and solvent B consisted of methanol containing 0.1% glacial acetic acid. A gradient elution method was utilized which started at 5% solvent B for 5 min, went to 95% solvent B over 30 min, and remained isocratic at 95% solvent B for 15 min. The column was reequilibrated for 5 min, and a blank run was performed between each analysis. This gradient was optimized for broad-range detection by lowering the initial organic concentration in the system, but it could easily be adjusted to better separate both shorter- and longer-acyl-chain AHLs.
Mass spectrometry. Mass spectrometric analyses were performed on a PE Sciex API-3000 triple-quadrupole tandem mass spectrometer (Perkin-Elmer Life Sciences, Thornhill, Ontario, Canada). Precursor ion-scanning experiments were performed in positive-ion mode with Q3 set to monitor for m/z 102.2 and Q1 set to scan a mass range of m/z 50 to m/z 400 over 5 s. The instrument parameters were as follows: ion spray voltage of 4,200 V, declustering potential of 50 V, focusing potential of 200 V, and collision energy of 25 V. Nitrogen was used as the collision gas.
Multiple reaction monitoring experiments were conducted using the same HPLC conditions and MS parameters as previously described. The ions monitored in Q1 and Q3 are shown in Table 2 for each AHL. These ions correspond to the transition from the parent ion of each AHL to both the acyl moiety [M + H-101]+, as well as the lactone moiety at m/z 102 (Fig. 1B) for each of the AHLs identified in the precursor ion-scanning experiments, as well as other AHLs predicted to be present.
Quantitative analysis of AHLs. Standard curves were prepared by making 1-mg/ml stock solutions of each of the following AHLs in methanol: 3-oxo-C6-HSL (Sigma), C6-HSL (Fluka), and C12-HSL (Fluka). The stock solutions (5 mM in methanol) were serially diluted to yield concentrations of 1,000, 200, 40, 8, and 1.6 μM. For each of the points in the standard curve, 10 μl of each dilution was added to an autosampler vial containing 0.4 nmol of the D3-C6-HSL internal standard in methanol. The standard curve samples were analyzed by MRM using the same HPLC conditions and instrument parameters as previously described. The analyte peak areas were integrated using a quantitation software package (Analyst 1.2; Perkin-Elmer). The standard curves generated compared the ratio of the areas of the analyte and internal standard peaks to the ratio of the amount of analyte and internal standard in each sample (see Results). Each integrated peak was inspected manually, and the data were normalized to the internal standard peak areas.
RESULTS
To understand how AHL synthases achieve such high specificity in AHL signal production, we examined the role of threonine 140 of EsaI and threonines 142 and 144 of LasI by comparing the AHL products of the wild-type and mutant enzymes. Determination of the AHL synthase intrinsic specificity with respect to utilization of different acyl-ACPs as substrates requires measurements of affinity and Km values for a wide range of acyl-ACPs. Not only is it difficult to make sufficient quantities of acyl-ACPs for such studies (26, 32), but it would also be necessary to know the identities and be able to make all of the relevant AHLs. Although AHLs can be surveyed using commonly used reporter strain biosensors, a complete analysis requires multiple reporter strains and numerous standards to detect and confirm many diverse AHLs. Furthermore, the reporter strains are biased for a specific range of acyl chain lengths as well as substitution at the 3 position. In order to achieve a more rapid and less-biased analysis of AHL synthase specificity than can be achieved using purified acyl-ACPs or AHL reporter bioassays, we used mass spectrometry to examine the range of AHLs produced by each enzyme expressed in E. coli under a variety of conditions.
Purification of AHLs from extracts. To analyze the AHLs produced by either native or mutated AHL synthases, it was necessary to first isolate the signals made by these enzymes. AHLs produced by gram-negative bacteria share a homoserine lactone ring and readily diffuse or are exported by the bacteria into the cell culture supernatant (30). Due to their inherent lipophilicity, AHLs were extracted from the cell culture supernatant by organic solvents such as ethyl acetate (37). To reduce the amount of lactone hydrolysis during workup and subsequent storage, the ethyl acetate was slightly acidified by the addition of 0.1 ml/liter acetic acid prior to the extraction (55, 65). Alternatively, AHLs can be extracted with other solvents, such as methylene chloride, which can extract 3-oxo and 3-hydroxy-HSLs more efficiently than does ethyl acetate, or AHLs can be extracted from whole cells using the Bligh and Dyer lipid extraction procedure (2). The mixtures of compounds of relevant molecular weight in either case were quite complex, and it was helpful to purify the AHLs from contaminating lipids and small molecules.
SPE under normal-phase conditions was effective in separating AHLs from contaminants based on a common chemical feature, the polar nature of the homoserine lactone ring. Development of the purification protocol used commercial AHLs and ethyl acetate extracts of Escherichia coli culture supernatants from cells expressing AHL synthase enzymes. Normal-phase TLC analysis was used to optimize the solvent system for SPE with purified 3-oxo-C6-HSL, 3-oxo-C12-HSL, and the ethyl acetate extracts of a 100-ml LasIG culture. The LasR reporter bioassay was used to track the elution of the AHL signal, as shown in Fig. 2. Ethyl acetate elution was sufficient to recover the input signal, which was satisfactory for this analysis, since an internal standard would correct for any loss from extraction and no further significant losses of AHL were seen in the SPE washing steps. This process eliminated the need to perform TLC analysis before mass spectrometric analysis and reduced the complexity of the samples to be examined while concentrating the AHLs.
Mass spectrometry of AHLs. Previous reports described that, when collisionally activated, the [M + H]+ ion derived from AHLs decomposed into specific ions, including an ion corresponding to the lactone moiety at m/z 102 and an ion derived from the acyl chain moiety [M + H-101]+ (see scheme in Fig. 1B) (41, 55). The ions generated by collisional activation provided precursor and product ion pairs, which were used to identify AHLs in precursor ion-scanning and MRM experiments. Using a triple-quadrupole mass spectrometer in precursor ion-scanning mode, AHLs were detected from a cell culture extract of the LasIG protein expressed in E. coli by scanning for parent ions that decompose to m/z 102, which correspond to the amino-butyrolactone (Fig. 3A). This analysis revealed that a significant proportion of odd-chain-length AHLs were produced and necessitated the use of a stable-isotope-labeled internal standard rather than an odd-chain-length AHL. No stable-isotope-labeled AHLs were available for any of the 10 to 15 AHLs expected to be present in the biological samples. Therefore, a suitable internal standard, D3-C6-HSL, was synthesized using a new method with the resin-bound coupling reagent PS-carbodiimide and D3-hexanoic acid (Fig. 4A). This product was tested by mass spectrometry (Fig. 4B).
Mass spectrometry can be used as a quantitative tool to determine accurate amounts of lipids present in a biological extract. The general approach has been to use a stable isotope dilution strategy to convert ion-abundant signals into ratio measurements that can be directly converted into absolute quantities of lipids using a calibration curve of isotope dilution against reference standard quantity (42). The use of the isotope-labeled internal standard corrects for changes in instrument sensitivity as well as ion chemistry effects, which determine the exact yield of product ions from collisional activation of a precursor ion. In the present study, an isotope-labeled internal standard was available for the C6-HSL, D3-C6-HSL, which was used to generate calibration curves for analysis of commercially available AHL standards, 3-oxo-C6-HSL, C6-HSL, and C12-HSL, and for subsequent analysis of biological samples. The calibration curves used to assess relationships between ion signal ratios and the quantities of AHLs (Fig. 5A) were linear, but for 3-oxo-C6-HSL and C12-HSL, the slopes were quite different (Fig. 5A). This difference in slope was likely caused by an alteration in the basic ion chemistry between the 3-oxo-C6-HSL and C12-HSL compared to the D3-C6-HSL internal standard, such as the presence of the keto group in 3-oxo species (41). Therefore, the ion currents for the two major product ions observed following collisional activation of each of the individual AHLs [M + H]+, i.e., [M + H-101]+ and m/z 102, were summed. When compared to that of the summed ion abundance of the internal standard [m/z 203102], the relationships between total ion response ratio to quantity of AHL were observed to be quite similar for all three AHLs investigated (Fig. 5B). Considering the similarity in the summed ion responses to the concentration of each AHL, a reasonably accurate assessment of individual AHL species quantity entering the mass spectrometer could be determined using this approach and the internal standard calibration curve (Fig. 5B).
The measurements of the AHLs presented here were comparative. The internal standard, D3-C6-HSL, was an isotopimer of only one AHL, namely C6-HSL, so that only this AHL was measured in an accurate and precise manner (Fig. 5). D3-C6-HSL was an AHL analog and, as such, did not correct for extraction and purification efficiencies of each individual AHL as does a stable-isotope-labeled internal standard. Thus, the values of AHLs other than C6-HSL do not represent exact amounts, and in particular the amounts of 3-oxo species may be underrepresented because the unsubstituted AHLs and AHLs with long chains (C12, for example) are extracted from the sample with higher efficiency than the 3-oxo-AHLs (52, 55). Although the use of a chemical analog as an internal standard cannot correct for all of the potential problems in extraction and purification, it does correct to some extent for extraction efficiency, and thus while the exact quantity of the AHLs (except for C6-HSL) cannot be precisely stated, it is highly likely they are in the rank order presented.
Specificity of AHL synthases. Based on structural analysis of the AHL synthase EsaI, a threonine (Thr140) was identified as a potential contributor to the specificity of the AHL synthase family for 3-oxo-acyl-ACPs (62). The TLC overlay bioassays showed a shift toward greater production of C6-HSL with concomitant lower production of 3-oxo-C6-HSL. However, the reporter assay did not reveal whether this was due to a shift in specificity or merely a loss of specificity. The AHLs produced by E. coli expressing the wild-type EsaI and the threonine-to-alanine mutant (T140A) were examined using the methods described above. There was a dramatic difference in the production of AHLs by the wild-type EsaI compared to the T140A EsaI mutant (in Fig. 6A). The level of 3-oxo-C6-HSL remained approximately the same, but the amount of C6-HSL increased dramatically. In addition, a number of unsubstituted AHLs of lengths C4, C5, C7, and C8 that were not seen before for either the wild-type or mutant EsaI were observed. Therefore, it appeared that mutation of threonine 140 to alanine led to a loss of specificity with a strong bias toward acyl-ACP substrates of length C6 without any loss of enzyme activity. This indicates that in EsaI, the role of threonine 140 is in restricting the acyl-ACPs that can bind to the enzyme, rather than enhancing the affinity of the enzyme for 3-oxo-acyl-ACPs.
In LasI from P. aeruginosa, the residue equivalent to EsaI Thr140, Thr142, was replaced with alanine, glycine, and serine. The T142G mutant produced smaller amounts of AHLs than the other mutants and WT LasI (Fig. 6B), but all of the T142 mutants did produce slightly more C12-HSL relative to 3-oxo-C12 than the WT LasI. However, the shift in specificity was very small compared to that observed for EsaI. This surprising result prompted our analysis of the only other amino acid side chain in a position to interact with the 3-oxo position of the acyl chain in the LasI acyl chain binding tunnel, threonine 144. The T144V mutant produced lower levels of AHLs in general than the WT LasI, but did not have any shift in the predominance of 3-oxo-AHLs compared to unsubstituted AHLs. Notably all of the LasI proteins produced AHLs with acyl chain lengths of odd and even numbers of carbons and unsaturated AHLs (Table 3) in addition to the AHLs, which are normally observed to have even acyl chain lengths. In addition, unsaturated AHLs were observed as abundant products. The unit of unsaturation was determined to be a single double bond, because the species were labile to catalytic reduction (data not shown). The unsaturated AHLs present at approximately 10% or above of the amount of the 3-oxo-C12 were 3-oxo-C14-HSL, C14-HSL, 3-oxo-C12-HSL, and C12-HSL. However, no precise quantitation was performed, because of the lack of a suitable reference standard.
The bacterial host of the AHL synthase altered the types of AHLs produced. AHL extracts from P. aeruginosa strain PAO214 expressing LasIG and the wild-type strain PAO1 grown for different times were compared to extracts of the same proteins expressed in E. coli, using this mass spectrometric method. It is well known that LasI makes predominantly 3-oxo-C12-HSL and, to a lesser degree, 3-oxo-C10-HSL, and that RhlI produces C4-HSL in vivo (Fig. 3B to D) (48). The extracts from E. coli (in Fig. 3A) showed predominantly the following ions, based on the observed transitions to m/z 102 (see also Table 2): m/z 242 for 3-oxo-C8-HSL, m/z 270 for 3-oxo-C10-HSL, m/z 284 for 3-oxo-C11-HSL, m/z 298 for 3-oxo-C12-HSL, m/z 312 for 3-oxo-C13-HSL, and m/z 326 for 3-oxo-C14-HSL. The retention time analysis with reference standards confirmed the identification of the odd-chain-length AHLs (data not shown). In contrast, the AHLs detected from extracts of LasIG expressed in P. aeruginosa (in Fig. 3B) included significant amounts only of 3-oxo-C10-HSL and 3-oxo-C12-HSL. The AHLs produced by the WT strain of P. aeruginosa PAO1 are also predominantly 3-oxo-C12-HSL and 3-oxo-C10-HSL at early times during bacterial growth (Fig. 3C) and show a clear shift to greater production of C4-HSL later in the progression of the culture, due to the increased activity of RhlI (Fig. 3D). Therefore, despite comparable levels of AHL production between species, the AHL extracts from LasI proteins expressed in E. coli gave a much broader range of AHLs, including the unusual appearance of odd-length AHLs than wild-type P. aeruginosa or LasIG expressed in an AHL synthase-null background of P. aeruginosa.
DISCUSSION
The specificities of two AHL synthases, EsaI and LasI, were examined using techniques that would define structural characteristics of the homoserine lactone products. EsaI had an apparent loss of enzyme specificity with mutation to a key threonine residue in the acyl chain binding pocket of EsaI. Among AHL synthases, there is a strong correlation between the identity of the residue at the position equivalent to Thr140 in EsaI and the AHL signal produced. In EsaI, this threonine appears to block the activity of the enzyme against unsubstituted acyl-ACPs, in contrast to what was previously believed to be a positive selection of 3-oxo-acyl-ACPs through a hydrogen bond between the ThrO1 and the 3-oxo position of acyl-ACP (62). Although this effect was much more subtle for the LasI mutants, there was also a slight loss of specificity, when the codon for the threonine residue was mutated to specify different amino acids that have been observed at that position in other AHL synthases. The neighboring T144, when mutated to valine, did not have any effect on the ratio of 3-oxo- to unsubstituted AHLs. This analysis confirms the importance of a specific threonine identified as a potential contributor to the specificity of the AHL synthase family for 3-oxo-acyl-ACPs. For EsaI, a number of other AHLs were produced, which suggests that this mutation also caused the enzyme to become less specific not only with respect to the 3 position of the acyl chain, but also the length of the acyl chain.
The mass spectrometric analysis of AHLs extracted from bacteria expressing recombinant or endogenous AHL synthases yielded unexpected results for the LasI enzyme. The AHLs produced by LasI expressed in P. aeruginosa and E. coli were vastly different (Fig. 3). In contrast, WT Yersinia pestis grown in culture and the YspI AHL synthase expressed in E. coli (31), produced similar AHLs, and EsaI expressed in E. coli produced only the AHLs it is known to produce in the WT strain (not shown). LasI may deplete its favored acyl-ACP but still be able to synthesize AHLs from other acyl-ACPs that have sufficiently high concentrations relative to their Km values for the enzyme. This is possible because the tunnel shape of the acyl chain binding site of the enzyme places no steric restriction on the length, long or short, of acyl chains that can be recognized by the enzyme. In contrast, the acyl binding pocket of EsaI restricts access of acyl-ACPs with acyl chain lengths significantly longer than C6 to the active site. Therefore, even though the preferred acyl donor for LasI is 3-oxo-C12-ACP, it is still able to synthesize a variety of shorter- and longer-acyl-chain AHLs in the E. coli environment.
In addition to differences in intrinsic AHL synthase specificity, heterologous expression of LasI could lead to changes in AHL distribution through differences in the pools of acyl-ACP that exist in a cell at a given time. P. aeruginosa has at least three ACP genes, and it is not known which of these three ACPs LasI uses under different conditions in vivo. In contrast, RhlI has a preference for AcpP Acp1 > Acp3, based on Km values of 5.9 μM, 7.4 μM, and 283 μM, respectively (50). AcpP is most similar to E. coli ACP and is thought to be essential for fatty acid biosynthesis, but Acp1 and Acp3 are upregulated by quorum sensing (22, 53, 60), which suggests that they may have specific functions in Pseudomonas pathogenicity that are still not understood. Therefore, the E. coli acyl-ACPs may not provide the specificity in the interactions with LasI that is seen with P. aeruginosa ACPs.
The available acyl-ACP pools in bacteria may be susceptible to metabolic changes. The fatty acid biosynthetic pathways in bacteria mostly produce even-chain fatty acids by extending the chain length of acyl-ACP through the addition of two carbons from malonyl-coenzyme A (reviewed in references 10 and 63). It has been recognized that the specific AHLs produced by a bacterium could be altered through modulation of the fatty acid biosynthetic pathway. In these studies (26), the amount of FabG (-ketoacyl-ACP reductase) was decreased in the cell and this impaired the ability of the bacterium to elongate acyl-ACPs. As a result, the mutant P. aeruginosa produced AHLs that had shorter acyl chains than the WT strain because the pool of acyl-ACPs most likely had shorter acyl chain lengths. Recent analyses of mutants in the P. aeruginosa VqsR (a novel quorum-sensing response regulator) gene also show the importance of the metabolic state and fatty acid biosynthesis cycle in AHL production. VqsR-null mutants fail to produce and/or secrete AHLs, and their ACP genes are significantly downregulated (28, 29).
Cellular metabolism may also be dramatically altered by AHLs. In addition to the lower apparent specificity of LasI in E. coli, a significant proportion of the AHLs have acyl chains with an odd number of carbon atoms and unsaturation. The pools of acyl-ACP in E. coli typically do not even contain acyl-ACPs of odd chain lengths (reviewed in references 10 and 63), which suggests that there must be another explanation for the appearance of the odd-chain-length AHLs. The most likely reason for the appearance of the unusual AHLs is that propionyl-ACP is being used as the precursor in the acyl-ACP elongation cycles in addition to the normal substrate acetyl-ACP. Propionyl-ACP is formed from propionyl-CoA by the acetyl-transacylase-catalyzed reaction, which normally produces acetyl-ACP but also has the ability to produce other ACPs. Propionate is formed and metabolized predominantly by the threonine degradation and methylcitrate cycles (4, 19, 20). In E. coli, anaerobic pathways could increase the levels of propionate in the cell (23). Therefore, the presence of the 3-oxo-C12-HSL or a receptor such as SdiA, or some other consequence of LasI expression in E. coli, either enhances the amount or activity of enzymes that produce propionate or alternatively inhibits the enzymes that deplete it from the cell. AHLs with odd acyl chain lengths have been reported previously, which suggests that these modulations of the cell's metabolic state may be more common than expected (3, 34).
AHLs have their main role in intraspecies communication, where the consequence of their synthesis and detection is relatively well understood (reviewed in reference 57). However, AHL-mediated signaling also functions in mixed bacterial populations (33), as well as in bacterial interactions with eukaryotic hosts such as the bobtail squid (11, 35, 43), plants, yeast, and also in patients infected with P. aeruginosa (9, 27, 51, 58). Our understanding of the importance and complexity of interspecies and interkingdom signaling by AHLs is still in its infancy. This work supports the idea that metabolic and environmental changes may alter the "language" spoken by a particular bacterium, which presents the potential to alter the consequences of that AHL signaling in vivo.
In conclusion, LC/MS/MS using a triple-quadrupole tandem mass spectrometer was an effective and sensitive approach for analysis of AHLs when used with an internal standard. A convenient method was developed to synthesize unsubstituted AHLs using carbodiimide chemistry for synthesis of deuterium-labeled internal standard. A method of AHL purification from cell culture supernatants, based on the lactone moiety, was effective at producing the appropriate samples for mass spectrometric analysis. LC/MS/MS analysis using a triple-quadrupole tandem mass spectrometer permitted precursor ion-scanning analysis, based on the fragment ion at m/z 102, to identify the AHLs in the sample. The summed MRM approach provided a semiquantitative method to compare the amounts of AHLs obtained from biological samples. The reference standards representing both the unsubstituted AHLs and the unsubstituted C6-HSL compared to the 3-oxo-C6-AHLs behave similarly to one another. Therefore, it can be expected that, with respect to chain length, unsubstituted AHLs between C6 and C12 and the 3-oxo-HSLs in this range will behave similarly enough to the D3-C6-HSL standard, in terms of ionization efficiency and of CID, that this standard will be sufficient for their quantitative analysis. This method revealed changes in AHL production that were useful in developing a better mechanistic understanding of the intrinsic specificity of AHL synthases and revealed underlying metabolic changes that also critically influence AHL synthesis in vivo.
ACKNOWLEDGMENTS
We thank Adam Barker, Kee-Yung Choi, and Herbert Schweizer for the P. aeruginosa strains; Don Zapien for assistance with plasmid mutagenesis; and Joe Hankin, Chris Johnson, and Wesley Martin for mass spectrometry assistance.
REFERENCES
Beck von Bodman, S., and S. K. Farrand. 1995. Capsular polysaccharide biosynthesis and pathogenicity in Erwinia stewartii require induction by an N-acylhomoserine lactone autoinducer. J. Bacteriol. 177:5000-5008.
Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911-917.
Brader, G., S. Sjoblom, H. Hyytiainen, K. Sims-Huopaniemi, and E. T. Palva. 2005. Altering substrate chain length specificity of an acylhomoserine lactone synthase in bacterial communication. J. Biol. Chem. 280:10403-10409.
Brock, M., C. Maerker, A. Schutz, U. Volker, and W. Buckel. 2002. Oxidation of propionate to pyruvate in Escherichia coli. Involvement of methylcitrate dehydratase and aconitase. Eur. J. Biochem. 269:6184-6194.
Buckman, B. O., M. M. Morrissey, and R. Mohan. 1998. Solution-phase parallel synthesis of benzoxazines using a polymer-supported carbodiimide. Tetrahedron Lett. 39:1487-1488.
Cataldi, T. R., G. Bianco, M. Frommberger, and P. Schmitt-Kopplin. 2004. Direct analysis of selected N-acyl-l-homoserine lactones by gas chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 18:1341-1344.
Cha, C., P. Gao, Y. C. Chen, P. D. Shaw, and S. K. Farrand. 1998. Production of acyl-homoserine lactone quorum-sensing signals by gram-negative plant-associated bacteria. Mol. Plant-Microbe Interact. 11:1119-1129.
Charlton, T. S., R. de Nys, A. Netting, N. Kumar, M. Hentzer, M. Givskov, and S. Kjelleberg. 2000. A novel and sensitive method for the quantification of N-3-oxoacyl homoserine lactones using gas chromatography-mass spectrometry: application to a model bacterial biofilm. Environ. Microbiol. 2:530-541.
Chhabra, S. R., C. Harty, D. S. Hooi, M. Daykin, P. Williams, G. Telford, D. I. Pritchard, and B. W. Bycroft. 2003. Synthetic analogues of the bacterial signal (quorum sensing) molecule N-(3-oxododecanoyl)-l-homoserine lactone as immune modulators. J. Med. Chem. 46:97-104.
Cronan, J. E. 2003. Bacterial membrane lipids: where do we stand Annu. Rev. Microbiol. 57:203-224.
Eberhard, A., A. L. Burlingame, C. Eberhard, G. L. Kenyon, K. H. Nealson, and N. J. Oppenheimer. 1981. Structural identification of autoinducer of Photobacterium fischeri. Biochemistry 20:2444-2449.
Flynn, D. L., R. V. Devraj, W. Naing, J. J. Parlow, J. J. Weidner, and S. L. Yang. 1998. Polymer-assisted solution phase (PASP) chemical library synthesis. Med. Chem. Res. 8:219-243.
Frommberger, M., P. Schmitt-Kopplin, F. Menzinger, V. Albrecht, M. Schmid, L. Eberl, A. Hartmann, and A. Kettrup. 2003. Analysis of N-acyl-l-homoserine lactones produced by Burkholderia cepacia with partial filling micellar electrokinetic chromatography-electrospray ionization-ion trap mass spectrometry. Electrophoresis 24:3067-3074.
Frommberger, M., P. Schmitt-Kopplin, G. Ping, H. Frisch, M. Schmid, Y. Zhang, A. Hartmann, and A. Kettrup. 2004. A simple and robust set-up for on-column sample preconcentration-nano-liquid chromatography-electrospray ionization mass spectrometry for the analysis of N-acylhomoserine lactones. Anal. Bioanal. Chem. 378:1014-1020.
Fuqua, C., and A. Eberhard. 1999. Signal generation in autoinduction systems: synthesis of acylated homoserine lactones by LuxI-type proteins, p.211-230. In G. Dunny and S. C. Winans (ed.), Cell-cell signaling in bacteria. ASM Press, Washington, D.C.
Giacomini, A., V. Corich, F. J. Ollero, A. Squartini, and M. P. Nuti. 1992. Experimental conditions may affect reproducibility of the beta-galactosidase assay. FEMS Microbiol. Lett. 79:87-90.
Gould, T. A., H. P. Schweizer, and M. E. A. Churchill. 2004. Structure of the Pseudomonas aeruginosa acyl-homoserinelactone synthase LasI. Mol. Microbiol. 53:1135-1146.
Gould, T. A., W. T. Watson, S. H. Choi, H. P. Schweizer, and M. E. A. Churchill. 2004. Crystallization of Pseudomonas aeruginosa AHL synthase LasI using beta-turn crystal engineering. Acta Crystallogr. D 60:518-520.
Haller, T., T. Buckel, J. Retey, and J. A. Gerlt. 2000. Discovering new enzymes and metabolic pathways: conversion of succinate to propionate by Escherichia coli. Biochemistry 39:4622-4629.
Heath, R. J., and C. O. Rock. 1996. Inhibition of beta-ketoacyl-acyl carrier protein synthase III (FabH) by acyl-acyl carrier protein in Escherichia coli. J. Biol. Chem. 271:10996-11000.
Henke, J. M., and B. L. Bassler. 2004. Bacterial social engagements. Trends Cell Biol. 14:648-656.
Hentzer, M., H. Wu, J. B. Andersen, K. Riedel, T. B. Rasmussen, N. Bagge, N. Kumar, M. A. Schembri, Z. Song, P. Kristoffersen, M. Manefield, J. W. Costerton, S. Molin, L. Eberl, P. Steinberg, S. Kjelleberg, N. Hoiby, and M. Givskov. 2003. Attenuation of Pseudomonas aeruginosa virulence by quorum sensing inhibitors. EMBO J. 22:3803-3815.
Hesslinger, C., S. A. Fairhurst, and G. Sawers. 1998. Novel keto acid formate-lyase and propionate kinase enzymes are components of an anaerobic pathway in Escherichia coli that degrades l-threonine to propionate. Mol. Microbiol. 27:477-492.
Hoang, T. T., Y. Ma, R. J. Stern, M. R. McNeil, and H. P. Schweizer. 1999. Construction and use of low-copy number T7 expression vectors for purification of problem proteins: purification of Mycobacterium tuberculosis RmlD and Pseudomonas aeruginosa LasI and RhlI proteins, and functional analysis of purified RhlI. Gene 237:361-371.
Hoang, T. T., A. J. Kutchma, A. Becher, and H. P. Schweizer. 2000. Integration-proficient plasmids for Pseudomonas aeruginosa: site-specific integration and use for engineering of reporter and expression strains. Plasmid 43:59-72.
Hoang, T. T., S. A. Sullivan, J. K. Cusick, and H. P. Schweizer. 2002. Beta-ketoacyl acyl carrier protein reductase (FabG) activity of the fatty acid biosynthetic pathway is a determining factor of 3-oxo-homoserine lactone acyl chain lengths. Microbiology 148:3849-3856.
Hooi, D. S., B. W. Bycroft, S. R. Chhabra, P. Williams, and D. I. Pritchard. 2004. Differential immune modulatory activity of Pseudomonas aeruginosa quorum-sensing signal molecules. Infect. Immun. 72:6463-6470.
Juhas, M., L. Eberl, and B. Tummler. 2005. Quorum sensing: the power of cooperation in the world of Pseudomonas. Environ. Microbiol. 7:459-471.
Juhas, M., L. Wiehlmann, P. Salunkhe, J. Lauber, J. Buer, and B. Tummler. 2005. GeneChip expression analysis of the VqsR regulon of Pseudomonas aeruginosa TB. FEMS Microbiol. Lett. 242:287-295.
Kaplan, H. B., and E. P. Greenberg. 1985. Diffusion of autoinducer is involved in the regulation of the Vibrio fischeri luminescence system. J.Bacteriol. 163:1210-1214.
Kirwan, J. P., T. A. Gould, H. P. Schweizer, S. W. Bearden, R. C. Murphy, and M. E. A. Churchill. 2006. Quorum-sensing signal synthesis by the Yersinia pestis acyl-homoserine lactone synthase YspI. J. Bacteriol. 188:784-788.
Kutchma, A. J., T. T. Hoang, and H. P. Schweizer. 1999. Characterization of a Pseudomonas aeruginosa fatty acid biosynthetic gene cluster: purification of acyl carrier protein (ACP) and malonyl-coenzyme A:ACP transacylase (FabD). J. Bacteriol. 181:5498-5504.
Lewenza, S., M. B. Visser, and P. A. Sokol. 2002. Interspecies communication between Burkholderia cepacia and Pseudomonas aeruginosa. Can. J. Microbiol. 48:707-716.
Lithgow, J. K., A. Wilkinson, A. Hardman, B. Rodelas, F. Wisniewski-Dye, P. Williams, and J. A. Downie. 2000. The regulatory locus cinRI in Rhizobium leguminosarum controls a network of quorum-sensing loci. Mol. Microbiol. 37:81-97.
Lupp, C., and E. G. Ruby. 2004. Vibrio fischeri LuxS and AinS: comparative study of two signal synthases. J. Bacteriol. 186:3873-3881.
Marketon, M. M., M. R. Gronquist, A. Eberhard, and J. E. Gonzalez. 2002. Characterization of the Sinorhizobium meliloti sinR/sinI locus and the production of novel N-acyl homoserine lactones. J. Bacteriol. 184:5686-5695.
McClean, K. H., M. K. Winson, L. Fish, A. Taylor, S. R. Chhabra, M. Camara, M. Daykin, J. H. Lamb, S. Swift, B. W. Bycroft, G. Stewart, and P. Williams. 1997. Quorum sensing and Chromobacterium violaceum: exploitation of violacein production and inhibition for the detection of N-acylhomoserine lactones. Microbiology 143:3703-3711.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
Miller, M. B., and B. L. Bassler. 2001. Quorum sensing in bacteria. Annu. Rev. Microbiol. 55:165-199.
More, M. I., L. D. Finger, J. L. Stryker, C. Fuqua, A. Eberhard, and S. C. Winans. 1996. Enzymatic synthesis of a quorum-sensing autoinducer through use of defined substrates. Science 272:1655-1658.
Morin, D., B. Grasland, K. Vallee-Rehel, C. Dufau, and D. Haras. 2003. On-line high-performance liquid chromatography-mass spectrometric detection and quantification of N-acylhomoserine lactones, quorum sensing signal molecules, in the presence of biological matrices. J. Chromatogr. A 1002:79-92.
Murphy, R. C., R. M. Barkley, K. Zemski Berry, J. A. Hankin, K. Harrison, C. Johnson, J. Krank, A. McAnoy, C. Uhlson, and S. Zarini. 2005. Electrospray ionization and tandem mass spectrometry of eicosanoids. Anal. Biochem. 346:1-42.
Nyholm, S. V., and M. J. McFall-Ngai. 2004. The winnowing: establishing the squid-vibrio symbiosis. Nat. Rev. Microbiol. 2:632-642.
Orsatti, L., S. Di Marco, C. Volpari, A. Vannini, P. Neddermann, and F. Bonelli. 2002. Determination of the stoichiometry of noncovalent complexes using reverse-phase high-performance liquid chromatography coupled with electrospray ion trap mass spectrometry. Anal. Biochem. 309:11-18.
Parlow, J. J., D. A. Mischke, and S. S. Woodard. 1997. Utility of complementary molecular reactivity and molecular recognition (CMR/R) technology and polymer-supported reagents in the solution-phase synthesis of heterocyclic carboxamides. J. Org. Chem. 62:5908-5919.
Parsek, M. R., D. L. Val, B. L. Hanzelka, J. E. J. Cronan, and E. P. Greenberg. 1999. Acyl homoserine-lactone quorum-sensing signal generation. Proc. Natl. Acad. Sci. USA 96:4360-4365.
Passador, L., K. D. Tucker, K. R. Guertin, M. P. Journet, A. S. Kende, and B. H. Iglewski. 1996. Functional analysis of the Pseudomonas aeruginosa autoinducer PAI. J. Bacteriol. 178:5995-6000.
Pearson, J. P., K. M. Gray, L. Passador, K. D. Tucker, A. Eberhard, B. H. Iglewski, and E. P. Greenberg. 1994. Structure of the autoinducer required for expression of Pseudomonas aeruginosa virulence genes. Proc. Natl. Acad. Sci. USA 91:197-201.
Ravn, L., A. B. Christensen, S. Molin, M. Givskov, and L. Gram. 2001. Methods for detecting acylated homoserine lactones produced by Gram-negative bacteria and their application in studies of AHL-production kinetics. J. Microbiol. Methods 44:239-251.
Raychaudhuri, A., A. Jerga, and P. A. Tipton. 2005. Chemical mechanism and substrate specificity of RhlI, an acylhomoserine lactone synthase from Pseudomonas aeruginosa. Biochemistry 44:2974-2981.
Ritchie, A. J., A. Jansson, J. Stallberg, P. Nilsson, P. Lysaght, and M. A. Cooley. 2005. The Pseudomonas aeruginosa quorum-sensing molecule N-3-(oxododecanoyl)-L-homoserine lactone inhibits T-cell differentiation and cytokine production by a mechanism involving an early step in T-cell activation. Infect. Immun. 73:1648-1655.
Schaefer, A. L., B. L. Hanzelka, M. R. Parsek, and E. P. Greenberg. 2000. Detection, purification, and structural elucidation of the acylhomoserine lactone inducer of Vibrio fischeri luminescence and other related molecules. Methods Enzymol. 305:288-301.
Schuster, M., C. P. Lostroh, T. Ogi, and E. P. Greenberg. 2003. Identification, timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled genes: a transcriptome analysis. J. Bacteriol. 185:2066-2079.
Schweizer, H. P., T. T. Hoang, K. L. Propst, H. R. Ornelas, and R. R. Karkhoff-Schweizer. 2001. Vector design and development of host systems for Pseudomonas, p. 69-81. In J. K. Setlow (ed.), Genetic engineering. Kluwer-Academic/Plenum, New York, N.Y.
Shaw, P. D., G. Ping, S. L. Daly, C. Cha, J. E. J. Cronan, K. L. Rinehart, and S. K. Farrand. 1997. Detecting and characterizing N-acyl-homoserine lactone signal molecules by thin-layer chromatography. Proc. Natl. Acad. Sci. USA 94:6036-6041.
Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff. 1990. Use of the T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185:60-89.
Taga, M. E., and B. L. Bassler. 2003. Chemical communication among bacteria. Proc. Natl. Acad. Sci. USA 100:14549-14554.
Tateda, K., Y. Ishii, M. Horikawa, T. Matsumoto, S. Miyairi, J. C. Pechere, T. J. Standiford, M. Ishiguro, and K. Yamaguchi. 2003. The Pseudomonas aeruginosa autoinducer N-3-oxododecanoyl homoserine lactone accelerates apoptosis in macrophages and neutrophils. Infect. Immun. 71:5785-5793.
Teplitski, M., A. Eberhard, M. R. Gronquist, M. Gao, J. B. Robinson, and W. D. Bauer. 2003. Chemical identification of N-acyl homoserine lactone quorum-sensing signals produced by Sinorhizobium meliloti strains in defined medium. Arch. Microbiol. 180:494-497.
Wagner, V. E., D. Bushnell, L. Passador, A. I. Brooks, and B. H. Iglewski. 2003. Microarray analysis of Pseudomonas aeruginosa quorum-sensing regulons: effects of growth phase and environment. J. Bacteriol. 185:2080-2095.
Watson, W. T., F. V. Murphy, T. A. Gould, P. Jambeck, D. L. Val, J. E. Cronan, S. Beck von Bodman, and M. E. A. Churchill. 2001. Crystallization and rhenium MAD phasing of the acyl-homoserine lactone synthase EsaI. Acta Crystallogr. D 57:1945-1949.
Watson, W. T., T. D. Minogue, D. L. Val, S. Beck von Bodman, and M. E. A. Churchill. 2002. Structural basis and specificity of acyl-homoserine lactone signal production in bacterial quorum sensing. Mol. Cell 9:685-694.
White, S. W., J. Zheng, Y. M. Zhang, and K. L. Rock. 2005. The structural biology of type ii fatty acid biosynthesis. Annu. Rev. Biochem. 74:791-831.
Whitehead, N. A., A. M. Barnard, H. Slater, N. J. Simpson, and G. P. Salmond. 2001. Quorum sensing in Gram-negative bacteria. FEMS Microbiol. Rev. 25:365-404.
Yates, E. A., B. Philipp, C. Buckley, S. Atkinson, S. R. Chhabra, R. E. Sockett, M. Goldner, Y. Dessaux, M. Camara, H. Smith, and P. Williams. 2002. N-Acylhomoserine lactones undergo lactonolysis in a pH-, temperature-, and acyl chain length-dependent manner during growth of Yersinia pseudotuberculosis and Pseudomonas aeruginosa. Infect. Immun. 70:5635-5646.
Zhu, J., Y. Chai, Z. Zhong, S. Li, and S. C. Winans. 2003. Agrobacterium bioassay strain for ultrasensitive detection of N-acylhomoserine lactone-type quorum-sensing molecules: detection of autoinducers in Mesorhizobium huakuii. Appl. Environ. Microbiol. 69:6949-6953.(Ty A. Gould, Jake Herman,)
ABSTRACT
Many gram-negative bacteria produce a specific set of N-acyl-L-homoserine-lactone (AHL) signaling molecules for the purpose of quorum sensing, which is a means of regulating coordinated gene expression in a cell-density-dependent manner. AHLs are produced from acylated acyl-carrier protein (acyl-ACP) and S-adenosyl-L-methionine by the AHL synthase enzyme. The appearance of specific AHLs is due in large part to the intrinsic specificity of the enzyme for subsets of acyl-ACP substrates. Structural studies of the Pantoea stewartii enzyme EsaI and AHL-sensitive bioassays revealed that threonine 140 in the acyl chain binding pocket directs the enzyme toward production of 3-oxo-homoserine lactones. Mass spectrometry was used to examine the range of AHL molecular species produced by AHL synthases under a variety of conditions. An AHL selective normal-phase chromatographic purification with addition of a deuterated AHL internal standard was followed by reverse-phase liquid chromatography-tandem mass spectrometry in order to obtain estimates of the relative amounts of different AHLs from biological samples. The AHLs produced by wild-type and engineered EsaI and LasI AHL synthases show that intrinsic specificity and different cellular conditions influence the production of AHLs. The threonine at position 140 in EsaI is important for the preference for 3-oxo-acyl-ACPs, but the role of the equivalent threonine in LasI is less clear. In addition, LasI expressed in Escherichia coli produces a high proportion of unusual AHLs with acyl chains consisting of an odd number of carbons. Furthermore, these studies offer additional methods that will be useful for surveying and quantitating AHLs from different sources.
INTRODUCTION
Many gram-negative bacteria produce acyl-homoserine lactones (AHLs) for a type of intercellular signaling known as quorum sensing, which is commonly associated with their pathogenic or communal behavior (reviewed in references 39 and 64). At least 70 different bacterial species produce AHLs, which can vary in acyl chain length from C4 to C18. In addition, AHLs can have different degrees of unsaturation at the C-7 or C-8 position, as well as oxidation at the 3 position (Fig. 1) (15, 36). Most bacterial species produce a restricted range of AHLs that is strongly coupled to the AHL selectivity of their own detection machinery, which uses transcriptional regulators, known as R-proteins, that selectively bind the AHL to either activate or repress many downstream target genes (53, 55, 60). This apparent coupling has led to the notion that AHL production is relatively restricted so that signaling within the species can be distinguished from interspecies signaling (21).
The intrinsic specificity of the AHL synthase for subsets of acyl-acyl carrier protein (acyl-ACP) substrates is an important factor in the production of particular AHLs. The AHLs are synthesized in the cell by the AHL synthase enzyme from substrates that are common components of cellular metabolism, acyl-ACP and S-adenosyl-L-methionine (40, 46). The Pantoea stewartii enzyme EsaI and Pseudomonas aeruginosa enzyme LasI preferentially produce 3-oxo-C6-homoserine lactone (HSL) and 3-oxo-C12-HSL, respectively (1, 48). Structural studies of these two enzymes revealed a common binding site for the acyl-ACP phosphopantetheine prosthetic group, and modeling studies implicated a particular threonine in the acyl-chain binding site as being important for hydrogen bonding to the 3-oxo position of acyl-ACP (17, 62). Furthermore, comparison of AHL synthase sequences and the AHLs they preferentially produce suggested that a threonine at that position correlates with the production of 3-oxo-HSLs, whereas alanine and glycine correlate with unsubstituted AHLs and serine correlates with the production of 3-hydroxy-HSLs. The AHLs produced by EsaI and a threonine 140-to-alanine substitution mutant, identified using two sensitive reporter bioassays, show a dramatic shift in AHL production from 3-oxo-C6-HSL in the wild type (WT) to C6-HSL in the mutant (62). Although this assay confirmed the role of this residue in AHL synthase specificity for 3-oxo-acyl-ACP, the true specificity could not be assessed using this AHL detection method.
AHLs can be detected effectively using AHL-sensitive bioassays in solution or in a thin-layer-chromatography (TLC) overlay format. The TLC overlay assay is widely used to determine the types of AHLs produced by a particular bacterium, or more specifically by a particular AHL synthase (37, 55). These bioassays are highly sensitive, with the ability to detect sub-picomole amounts of particular AHLs. However, there is an inherent detection bias, because they rely on the particular specificity of the reporter strain that is used for AHL detection, such as Chromobacterium violaceum (37), Agrobacterium tumefaciens (TraR) (55), or Pseudomonas aeruginosa (LasR) (48). Recent developments in reporter strains have provided some relatively promiscuous detection systems that recognize a much wider variety of AHLs (49, 66). However, for complete coverage of known AHLs, multiple reporter strains must be used with a full complement of standards to assess which spots correspond to each particular signal (7).
Mass spectrometry (MS) and gas chromatography (GC)-MS offer alternative methods of AHL detection, which are based on the physical/chemical properties of the compounds, such as the mass/charge ratio of molecular ions, collisionally induced product ions, and chromatographic retention properties. Derivatization of AHLs has been used for the examination of constellations of 3-oxo-HSLs by GC-MS (8). Direct analysis of AHLs has been reported using GC-MS (6) and high-performance liquid chromatography (HPLC) coupled to ion trap mass spectrometry (LC-MS) (36, 41, 44, 48, 59). The addition of a preconcentration step has been reported to result in highly sensitive methods of AHL detection (13, 14).
Here we describe a combination of approaches that show how distributions of AHLs change with mutations introduced into two AHL synthases, EsaI and LasI. The distribution of AHLs from experimental samples was determined by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC/MS/MS) using a triple-quadrupole mass spectrometer. We examined the influence of the active site threonines of EsaI and LasI in the specificity of AHL synthesis, as well as the contributions of different bacterial backgrounds to the variety of AHLs produced by these AHL synthases in vivo.
MATERIALS AND METHODS
AHL nomenclature. For brevity, an acyl chain of a specific length is denoted by "Cn" for the number of carbons in the chain, rather than the chemical names (i.e., hexanoyl is C6), and the substitution type and position as 3-oxo or 3-hydroxy (eg., 3-oxo-C6-HSL). D3 indicates that the AHL contains three deuterium atoms at the terminal position of the acyl chain.
Synthesis of D3-C6-homoserine lactone. Solid-phase carbodiimide chemistry was used to synthesize D3-C6-HSL, using N-cyclohexylcarbodiimide-N'-propyloxymethyl polystyrene resin (PS-carbodiimide; Argonaut Technologies) (12, 45). The fatty acid D3-hexanoic acid (34.5 μmol [4.11 mg]) (Cambridge Isotope) was added to 400 μl of methylene chloride (CH2Cl2) with 10% N,N-dimethylformamide in a separate glass reaction vial. To this reaction vial, 40 mg of PS-carbodiimide was added and the mixture was stirred for 5 min at room temperature. After the fatty acid had been allowed to equilibrate with the resin, 4.2 mg (23 μmol) of -amino--butyrolactone hydrobromide (HSL-HBr) (Aldrich) dissolved in 200 μl of CH2Cl2 with 10% N,N-dimethylformamide and 4.6 μmol (6.4 μl) of triethylamine (subsequent analysis of products revealed no significant lactonolysis when triethylamine was used as the base [data not shown]) was added and stirred for 30 h at room temperature. The reaction workup consisted of filtering out the resin-bound substrate/intermediate, coupled to solid-phase extraction as described below for AHL extracts. The D3-C6-HSL product was resuspended in methanol to a final concentration of 200 μM. This dilution was shown by LC/MS to be equivalent to 0.2 nmol of C6-HSL and was used for all remaining experiments (see Fig. 4B).
Mutagenesis of the LasIG plasmid. LasIG codon substitutions were made from pLasIG using the QuikChange mutagenesis method (Stratagene) with the complementary primers for each mutation shown in Table 1. The mutations were confirmed by restriction digestion analysis, where possible, and by DNA sequencing of the entire gene.
Extraction and purification of acyl-homoserine lactones for mass spectrometry analysis. E. coli strains (Table 1), harboring expression plasmids for the AHL synthases EsaI from P. stewartii and pLasI and pLasIG from P. aeruginosa (18, 61), were grown in 5-ml cultures in Luria-Bertani (LB) broth with 100 μg/ml of ampicillin at 37°C with shaking for 12 to 16 h. Each culture was diluted 1:50 into 10 ml of fresh LB broth with 100 μg/ml ampicillin and incubated until the cell density reached an optical density at 600 nm (OD600) of 0.6 to 0.8. The temperature was lowered to 25°C for 15 min, and after equilibration, each culture was induced by the addition of IPTG (isopropyl--D-thiogalactopyranoside) to 0.5 mM. Cultures were incubated at 25°C for 6 to 8 h with shaking. The P. aeruginosa cultures were grown in the same way, except that strain PAO1 required no antibiotic and strain PAO214 with pEX30-las required 500 μg/ml carbenicillin. Culture supernatants were processed by centrifugation at 3,200 x g for 10 min, followed by decanting into a 20-ml syringe and passing through a 0.2-μm filter. For samples being prepared for quantitative analysis, 0.4 nmol of the synthetic D3-C6-HSL was added per 10 ml of culture after harvesting the supernatant but prior to filtering it. The culture supernatants were extracted two times, with 10 ml of acidified ethyl acetate (0.1 ml/liter acetic acid), and 9 ml from each extraction was pooled and taken to complete dryness in 2-ml glass sample vials. This extraction procedure was expected to extract at least 75% of the 3-oxo-C6-HSLs. AHLs were also obtained from commercial sources (Sigma, Quorum Sciences Inc.).
Initial purification of the AHL molecular species was accomplished by solid-phase extraction (SPE). Each sample, which had been redissolved in 100 μl of methanol (Optima, Sigma), was applied to activated Sep Pak Plus silica cartridges (Waters), which were fitted with a 6-ml glass reaction vial and attached to the vacuum manifold. Cartridges had been activated by successive washings of 6 ml of each solvent in the following sequence: equal volumes of isooctane and ethyl ether (ethyl ether must not have the preservative butylated hydroxytoluene), acidified ethyl acetate, and isooctane and ethyl ether. The sample in methanol was added to 5 ml isooctane-ethyl ether, and was pipetted into the reaction vial and then loaded onto the SPE cartridge. The SPE cartridge was washed twice with 6 ml of isooctane-ether and then eluted into a glass fraction tube with 5 to 8 ml of acidified ethyl acetate. The purified samples were taken to near dryness by vacuum evaporation and then transferred to a glass autosampler vial to be evaporated completely. The samples were redissolved in 50 or 100 μl of methanol and capped.
Hexane can be used in place of isooctane as a solvent in the purification procedure, but impurities in the reagent-grade hexane yielded electrospray ionization ions of m/z 284 and 228 in the AHL analysis. Therefore, hexane was prepurified by passage over the SPE cartridge prior to being used in the purification of AHLs from cell culture extracts. The true AHLs were identified by LC/MS/MS in a multiple reaction monitoring (MRM) experiment, where the transition from the parent ion to both the acyl and lactone moiety peaks overlapped by comparison to known standards and by evaluation of chromatographic retention times. Retention time analysis was used to distinguish between the C(n)-HSL and the 3-oxo-C(n-1)-HSL molecular species, which have similar masses (data not shown).
LasR reporter assay. The LasR reporter strain E. coli MG4/pKDT17a (48, 52) was grown overnight in A medium (52), supplemented with 1 mM MgSO4 and 100 μg/ml ampicillin at 30°C. Each extract was incubated with 1 ml of diluted culture (diluted to an OD600 of 0.1) for 5 to 6 h at 30°C. The samples were then subjected to analysis of -galactosidase activity using the Miller assay (16, 38).
High-performance liquid chromatography. Samples for LC/MS and LC/MS/MS analysis were resuspended in 100 μl of methanol, and 10 μl was injected onto a 2.0-mm by 150.0-mm Columbus C18 reverse-phase column (Phenomenex) operated at a flow rate of 200 μl/min with the effluent flowing directly into the mass spectrometer. Solvent A consisted of H2O containing 0.1% glacial acetic acid, and solvent B consisted of methanol containing 0.1% glacial acetic acid. A gradient elution method was utilized which started at 5% solvent B for 5 min, went to 95% solvent B over 30 min, and remained isocratic at 95% solvent B for 15 min. The column was reequilibrated for 5 min, and a blank run was performed between each analysis. This gradient was optimized for broad-range detection by lowering the initial organic concentration in the system, but it could easily be adjusted to better separate both shorter- and longer-acyl-chain AHLs.
Mass spectrometry. Mass spectrometric analyses were performed on a PE Sciex API-3000 triple-quadrupole tandem mass spectrometer (Perkin-Elmer Life Sciences, Thornhill, Ontario, Canada). Precursor ion-scanning experiments were performed in positive-ion mode with Q3 set to monitor for m/z 102.2 and Q1 set to scan a mass range of m/z 50 to m/z 400 over 5 s. The instrument parameters were as follows: ion spray voltage of 4,200 V, declustering potential of 50 V, focusing potential of 200 V, and collision energy of 25 V. Nitrogen was used as the collision gas.
Multiple reaction monitoring experiments were conducted using the same HPLC conditions and MS parameters as previously described. The ions monitored in Q1 and Q3 are shown in Table 2 for each AHL. These ions correspond to the transition from the parent ion of each AHL to both the acyl moiety [M + H-101]+, as well as the lactone moiety at m/z 102 (Fig. 1B) for each of the AHLs identified in the precursor ion-scanning experiments, as well as other AHLs predicted to be present.
Quantitative analysis of AHLs. Standard curves were prepared by making 1-mg/ml stock solutions of each of the following AHLs in methanol: 3-oxo-C6-HSL (Sigma), C6-HSL (Fluka), and C12-HSL (Fluka). The stock solutions (5 mM in methanol) were serially diluted to yield concentrations of 1,000, 200, 40, 8, and 1.6 μM. For each of the points in the standard curve, 10 μl of each dilution was added to an autosampler vial containing 0.4 nmol of the D3-C6-HSL internal standard in methanol. The standard curve samples were analyzed by MRM using the same HPLC conditions and instrument parameters as previously described. The analyte peak areas were integrated using a quantitation software package (Analyst 1.2; Perkin-Elmer). The standard curves generated compared the ratio of the areas of the analyte and internal standard peaks to the ratio of the amount of analyte and internal standard in each sample (see Results). Each integrated peak was inspected manually, and the data were normalized to the internal standard peak areas.
RESULTS
To understand how AHL synthases achieve such high specificity in AHL signal production, we examined the role of threonine 140 of EsaI and threonines 142 and 144 of LasI by comparing the AHL products of the wild-type and mutant enzymes. Determination of the AHL synthase intrinsic specificity with respect to utilization of different acyl-ACPs as substrates requires measurements of affinity and Km values for a wide range of acyl-ACPs. Not only is it difficult to make sufficient quantities of acyl-ACPs for such studies (26, 32), but it would also be necessary to know the identities and be able to make all of the relevant AHLs. Although AHLs can be surveyed using commonly used reporter strain biosensors, a complete analysis requires multiple reporter strains and numerous standards to detect and confirm many diverse AHLs. Furthermore, the reporter strains are biased for a specific range of acyl chain lengths as well as substitution at the 3 position. In order to achieve a more rapid and less-biased analysis of AHL synthase specificity than can be achieved using purified acyl-ACPs or AHL reporter bioassays, we used mass spectrometry to examine the range of AHLs produced by each enzyme expressed in E. coli under a variety of conditions.
Purification of AHLs from extracts. To analyze the AHLs produced by either native or mutated AHL synthases, it was necessary to first isolate the signals made by these enzymes. AHLs produced by gram-negative bacteria share a homoserine lactone ring and readily diffuse or are exported by the bacteria into the cell culture supernatant (30). Due to their inherent lipophilicity, AHLs were extracted from the cell culture supernatant by organic solvents such as ethyl acetate (37). To reduce the amount of lactone hydrolysis during workup and subsequent storage, the ethyl acetate was slightly acidified by the addition of 0.1 ml/liter acetic acid prior to the extraction (55, 65). Alternatively, AHLs can be extracted with other solvents, such as methylene chloride, which can extract 3-oxo and 3-hydroxy-HSLs more efficiently than does ethyl acetate, or AHLs can be extracted from whole cells using the Bligh and Dyer lipid extraction procedure (2). The mixtures of compounds of relevant molecular weight in either case were quite complex, and it was helpful to purify the AHLs from contaminating lipids and small molecules.
SPE under normal-phase conditions was effective in separating AHLs from contaminants based on a common chemical feature, the polar nature of the homoserine lactone ring. Development of the purification protocol used commercial AHLs and ethyl acetate extracts of Escherichia coli culture supernatants from cells expressing AHL synthase enzymes. Normal-phase TLC analysis was used to optimize the solvent system for SPE with purified 3-oxo-C6-HSL, 3-oxo-C12-HSL, and the ethyl acetate extracts of a 100-ml LasIG culture. The LasR reporter bioassay was used to track the elution of the AHL signal, as shown in Fig. 2. Ethyl acetate elution was sufficient to recover the input signal, which was satisfactory for this analysis, since an internal standard would correct for any loss from extraction and no further significant losses of AHL were seen in the SPE washing steps. This process eliminated the need to perform TLC analysis before mass spectrometric analysis and reduced the complexity of the samples to be examined while concentrating the AHLs.
Mass spectrometry of AHLs. Previous reports described that, when collisionally activated, the [M + H]+ ion derived from AHLs decomposed into specific ions, including an ion corresponding to the lactone moiety at m/z 102 and an ion derived from the acyl chain moiety [M + H-101]+ (see scheme in Fig. 1B) (41, 55). The ions generated by collisional activation provided precursor and product ion pairs, which were used to identify AHLs in precursor ion-scanning and MRM experiments. Using a triple-quadrupole mass spectrometer in precursor ion-scanning mode, AHLs were detected from a cell culture extract of the LasIG protein expressed in E. coli by scanning for parent ions that decompose to m/z 102, which correspond to the amino-butyrolactone (Fig. 3A). This analysis revealed that a significant proportion of odd-chain-length AHLs were produced and necessitated the use of a stable-isotope-labeled internal standard rather than an odd-chain-length AHL. No stable-isotope-labeled AHLs were available for any of the 10 to 15 AHLs expected to be present in the biological samples. Therefore, a suitable internal standard, D3-C6-HSL, was synthesized using a new method with the resin-bound coupling reagent PS-carbodiimide and D3-hexanoic acid (Fig. 4A). This product was tested by mass spectrometry (Fig. 4B).
Mass spectrometry can be used as a quantitative tool to determine accurate amounts of lipids present in a biological extract. The general approach has been to use a stable isotope dilution strategy to convert ion-abundant signals into ratio measurements that can be directly converted into absolute quantities of lipids using a calibration curve of isotope dilution against reference standard quantity (42). The use of the isotope-labeled internal standard corrects for changes in instrument sensitivity as well as ion chemistry effects, which determine the exact yield of product ions from collisional activation of a precursor ion. In the present study, an isotope-labeled internal standard was available for the C6-HSL, D3-C6-HSL, which was used to generate calibration curves for analysis of commercially available AHL standards, 3-oxo-C6-HSL, C6-HSL, and C12-HSL, and for subsequent analysis of biological samples. The calibration curves used to assess relationships between ion signal ratios and the quantities of AHLs (Fig. 5A) were linear, but for 3-oxo-C6-HSL and C12-HSL, the slopes were quite different (Fig. 5A). This difference in slope was likely caused by an alteration in the basic ion chemistry between the 3-oxo-C6-HSL and C12-HSL compared to the D3-C6-HSL internal standard, such as the presence of the keto group in 3-oxo species (41). Therefore, the ion currents for the two major product ions observed following collisional activation of each of the individual AHLs [M + H]+, i.e., [M + H-101]+ and m/z 102, were summed. When compared to that of the summed ion abundance of the internal standard [m/z 203102], the relationships between total ion response ratio to quantity of AHL were observed to be quite similar for all three AHLs investigated (Fig. 5B). Considering the similarity in the summed ion responses to the concentration of each AHL, a reasonably accurate assessment of individual AHL species quantity entering the mass spectrometer could be determined using this approach and the internal standard calibration curve (Fig. 5B).
The measurements of the AHLs presented here were comparative. The internal standard, D3-C6-HSL, was an isotopimer of only one AHL, namely C6-HSL, so that only this AHL was measured in an accurate and precise manner (Fig. 5). D3-C6-HSL was an AHL analog and, as such, did not correct for extraction and purification efficiencies of each individual AHL as does a stable-isotope-labeled internal standard. Thus, the values of AHLs other than C6-HSL do not represent exact amounts, and in particular the amounts of 3-oxo species may be underrepresented because the unsubstituted AHLs and AHLs with long chains (C12, for example) are extracted from the sample with higher efficiency than the 3-oxo-AHLs (52, 55). Although the use of a chemical analog as an internal standard cannot correct for all of the potential problems in extraction and purification, it does correct to some extent for extraction efficiency, and thus while the exact quantity of the AHLs (except for C6-HSL) cannot be precisely stated, it is highly likely they are in the rank order presented.
Specificity of AHL synthases. Based on structural analysis of the AHL synthase EsaI, a threonine (Thr140) was identified as a potential contributor to the specificity of the AHL synthase family for 3-oxo-acyl-ACPs (62). The TLC overlay bioassays showed a shift toward greater production of C6-HSL with concomitant lower production of 3-oxo-C6-HSL. However, the reporter assay did not reveal whether this was due to a shift in specificity or merely a loss of specificity. The AHLs produced by E. coli expressing the wild-type EsaI and the threonine-to-alanine mutant (T140A) were examined using the methods described above. There was a dramatic difference in the production of AHLs by the wild-type EsaI compared to the T140A EsaI mutant (in Fig. 6A). The level of 3-oxo-C6-HSL remained approximately the same, but the amount of C6-HSL increased dramatically. In addition, a number of unsubstituted AHLs of lengths C4, C5, C7, and C8 that were not seen before for either the wild-type or mutant EsaI were observed. Therefore, it appeared that mutation of threonine 140 to alanine led to a loss of specificity with a strong bias toward acyl-ACP substrates of length C6 without any loss of enzyme activity. This indicates that in EsaI, the role of threonine 140 is in restricting the acyl-ACPs that can bind to the enzyme, rather than enhancing the affinity of the enzyme for 3-oxo-acyl-ACPs.
In LasI from P. aeruginosa, the residue equivalent to EsaI Thr140, Thr142, was replaced with alanine, glycine, and serine. The T142G mutant produced smaller amounts of AHLs than the other mutants and WT LasI (Fig. 6B), but all of the T142 mutants did produce slightly more C12-HSL relative to 3-oxo-C12 than the WT LasI. However, the shift in specificity was very small compared to that observed for EsaI. This surprising result prompted our analysis of the only other amino acid side chain in a position to interact with the 3-oxo position of the acyl chain in the LasI acyl chain binding tunnel, threonine 144. The T144V mutant produced lower levels of AHLs in general than the WT LasI, but did not have any shift in the predominance of 3-oxo-AHLs compared to unsubstituted AHLs. Notably all of the LasI proteins produced AHLs with acyl chain lengths of odd and even numbers of carbons and unsaturated AHLs (Table 3) in addition to the AHLs, which are normally observed to have even acyl chain lengths. In addition, unsaturated AHLs were observed as abundant products. The unit of unsaturation was determined to be a single double bond, because the species were labile to catalytic reduction (data not shown). The unsaturated AHLs present at approximately 10% or above of the amount of the 3-oxo-C12 were 3-oxo-C14-HSL, C14-HSL, 3-oxo-C12-HSL, and C12-HSL. However, no precise quantitation was performed, because of the lack of a suitable reference standard.
The bacterial host of the AHL synthase altered the types of AHLs produced. AHL extracts from P. aeruginosa strain PAO214 expressing LasIG and the wild-type strain PAO1 grown for different times were compared to extracts of the same proteins expressed in E. coli, using this mass spectrometric method. It is well known that LasI makes predominantly 3-oxo-C12-HSL and, to a lesser degree, 3-oxo-C10-HSL, and that RhlI produces C4-HSL in vivo (Fig. 3B to D) (48). The extracts from E. coli (in Fig. 3A) showed predominantly the following ions, based on the observed transitions to m/z 102 (see also Table 2): m/z 242 for 3-oxo-C8-HSL, m/z 270 for 3-oxo-C10-HSL, m/z 284 for 3-oxo-C11-HSL, m/z 298 for 3-oxo-C12-HSL, m/z 312 for 3-oxo-C13-HSL, and m/z 326 for 3-oxo-C14-HSL. The retention time analysis with reference standards confirmed the identification of the odd-chain-length AHLs (data not shown). In contrast, the AHLs detected from extracts of LasIG expressed in P. aeruginosa (in Fig. 3B) included significant amounts only of 3-oxo-C10-HSL and 3-oxo-C12-HSL. The AHLs produced by the WT strain of P. aeruginosa PAO1 are also predominantly 3-oxo-C12-HSL and 3-oxo-C10-HSL at early times during bacterial growth (Fig. 3C) and show a clear shift to greater production of C4-HSL later in the progression of the culture, due to the increased activity of RhlI (Fig. 3D). Therefore, despite comparable levels of AHL production between species, the AHL extracts from LasI proteins expressed in E. coli gave a much broader range of AHLs, including the unusual appearance of odd-length AHLs than wild-type P. aeruginosa or LasIG expressed in an AHL synthase-null background of P. aeruginosa.
DISCUSSION
The specificities of two AHL synthases, EsaI and LasI, were examined using techniques that would define structural characteristics of the homoserine lactone products. EsaI had an apparent loss of enzyme specificity with mutation to a key threonine residue in the acyl chain binding pocket of EsaI. Among AHL synthases, there is a strong correlation between the identity of the residue at the position equivalent to Thr140 in EsaI and the AHL signal produced. In EsaI, this threonine appears to block the activity of the enzyme against unsubstituted acyl-ACPs, in contrast to what was previously believed to be a positive selection of 3-oxo-acyl-ACPs through a hydrogen bond between the ThrO1 and the 3-oxo position of acyl-ACP (62). Although this effect was much more subtle for the LasI mutants, there was also a slight loss of specificity, when the codon for the threonine residue was mutated to specify different amino acids that have been observed at that position in other AHL synthases. The neighboring T144, when mutated to valine, did not have any effect on the ratio of 3-oxo- to unsubstituted AHLs. This analysis confirms the importance of a specific threonine identified as a potential contributor to the specificity of the AHL synthase family for 3-oxo-acyl-ACPs. For EsaI, a number of other AHLs were produced, which suggests that this mutation also caused the enzyme to become less specific not only with respect to the 3 position of the acyl chain, but also the length of the acyl chain.
The mass spectrometric analysis of AHLs extracted from bacteria expressing recombinant or endogenous AHL synthases yielded unexpected results for the LasI enzyme. The AHLs produced by LasI expressed in P. aeruginosa and E. coli were vastly different (Fig. 3). In contrast, WT Yersinia pestis grown in culture and the YspI AHL synthase expressed in E. coli (31), produced similar AHLs, and EsaI expressed in E. coli produced only the AHLs it is known to produce in the WT strain (not shown). LasI may deplete its favored acyl-ACP but still be able to synthesize AHLs from other acyl-ACPs that have sufficiently high concentrations relative to their Km values for the enzyme. This is possible because the tunnel shape of the acyl chain binding site of the enzyme places no steric restriction on the length, long or short, of acyl chains that can be recognized by the enzyme. In contrast, the acyl binding pocket of EsaI restricts access of acyl-ACPs with acyl chain lengths significantly longer than C6 to the active site. Therefore, even though the preferred acyl donor for LasI is 3-oxo-C12-ACP, it is still able to synthesize a variety of shorter- and longer-acyl-chain AHLs in the E. coli environment.
In addition to differences in intrinsic AHL synthase specificity, heterologous expression of LasI could lead to changes in AHL distribution through differences in the pools of acyl-ACP that exist in a cell at a given time. P. aeruginosa has at least three ACP genes, and it is not known which of these three ACPs LasI uses under different conditions in vivo. In contrast, RhlI has a preference for AcpP Acp1 > Acp3, based on Km values of 5.9 μM, 7.4 μM, and 283 μM, respectively (50). AcpP is most similar to E. coli ACP and is thought to be essential for fatty acid biosynthesis, but Acp1 and Acp3 are upregulated by quorum sensing (22, 53, 60), which suggests that they may have specific functions in Pseudomonas pathogenicity that are still not understood. Therefore, the E. coli acyl-ACPs may not provide the specificity in the interactions with LasI that is seen with P. aeruginosa ACPs.
The available acyl-ACP pools in bacteria may be susceptible to metabolic changes. The fatty acid biosynthetic pathways in bacteria mostly produce even-chain fatty acids by extending the chain length of acyl-ACP through the addition of two carbons from malonyl-coenzyme A (reviewed in references 10 and 63). It has been recognized that the specific AHLs produced by a bacterium could be altered through modulation of the fatty acid biosynthetic pathway. In these studies (26), the amount of FabG (-ketoacyl-ACP reductase) was decreased in the cell and this impaired the ability of the bacterium to elongate acyl-ACPs. As a result, the mutant P. aeruginosa produced AHLs that had shorter acyl chains than the WT strain because the pool of acyl-ACPs most likely had shorter acyl chain lengths. Recent analyses of mutants in the P. aeruginosa VqsR (a novel quorum-sensing response regulator) gene also show the importance of the metabolic state and fatty acid biosynthesis cycle in AHL production. VqsR-null mutants fail to produce and/or secrete AHLs, and their ACP genes are significantly downregulated (28, 29).
Cellular metabolism may also be dramatically altered by AHLs. In addition to the lower apparent specificity of LasI in E. coli, a significant proportion of the AHLs have acyl chains with an odd number of carbon atoms and unsaturation. The pools of acyl-ACP in E. coli typically do not even contain acyl-ACPs of odd chain lengths (reviewed in references 10 and 63), which suggests that there must be another explanation for the appearance of the odd-chain-length AHLs. The most likely reason for the appearance of the unusual AHLs is that propionyl-ACP is being used as the precursor in the acyl-ACP elongation cycles in addition to the normal substrate acetyl-ACP. Propionyl-ACP is formed from propionyl-CoA by the acetyl-transacylase-catalyzed reaction, which normally produces acetyl-ACP but also has the ability to produce other ACPs. Propionate is formed and metabolized predominantly by the threonine degradation and methylcitrate cycles (4, 19, 20). In E. coli, anaerobic pathways could increase the levels of propionate in the cell (23). Therefore, the presence of the 3-oxo-C12-HSL or a receptor such as SdiA, or some other consequence of LasI expression in E. coli, either enhances the amount or activity of enzymes that produce propionate or alternatively inhibits the enzymes that deplete it from the cell. AHLs with odd acyl chain lengths have been reported previously, which suggests that these modulations of the cell's metabolic state may be more common than expected (3, 34).
AHLs have their main role in intraspecies communication, where the consequence of their synthesis and detection is relatively well understood (reviewed in reference 57). However, AHL-mediated signaling also functions in mixed bacterial populations (33), as well as in bacterial interactions with eukaryotic hosts such as the bobtail squid (11, 35, 43), plants, yeast, and also in patients infected with P. aeruginosa (9, 27, 51, 58). Our understanding of the importance and complexity of interspecies and interkingdom signaling by AHLs is still in its infancy. This work supports the idea that metabolic and environmental changes may alter the "language" spoken by a particular bacterium, which presents the potential to alter the consequences of that AHL signaling in vivo.
In conclusion, LC/MS/MS using a triple-quadrupole tandem mass spectrometer was an effective and sensitive approach for analysis of AHLs when used with an internal standard. A convenient method was developed to synthesize unsubstituted AHLs using carbodiimide chemistry for synthesis of deuterium-labeled internal standard. A method of AHL purification from cell culture supernatants, based on the lactone moiety, was effective at producing the appropriate samples for mass spectrometric analysis. LC/MS/MS analysis using a triple-quadrupole tandem mass spectrometer permitted precursor ion-scanning analysis, based on the fragment ion at m/z 102, to identify the AHLs in the sample. The summed MRM approach provided a semiquantitative method to compare the amounts of AHLs obtained from biological samples. The reference standards representing both the unsubstituted AHLs and the unsubstituted C6-HSL compared to the 3-oxo-C6-AHLs behave similarly to one another. Therefore, it can be expected that, with respect to chain length, unsubstituted AHLs between C6 and C12 and the 3-oxo-HSLs in this range will behave similarly enough to the D3-C6-HSL standard, in terms of ionization efficiency and of CID, that this standard will be sufficient for their quantitative analysis. This method revealed changes in AHL production that were useful in developing a better mechanistic understanding of the intrinsic specificity of AHL synthases and revealed underlying metabolic changes that also critically influence AHL synthesis in vivo.
ACKNOWLEDGMENTS
We thank Adam Barker, Kee-Yung Choi, and Herbert Schweizer for the P. aeruginosa strains; Don Zapien for assistance with plasmid mutagenesis; and Joe Hankin, Chris Johnson, and Wesley Martin for mass spectrometry assistance.
REFERENCES
Beck von Bodman, S., and S. K. Farrand. 1995. Capsular polysaccharide biosynthesis and pathogenicity in Erwinia stewartii require induction by an N-acylhomoserine lactone autoinducer. J. Bacteriol. 177:5000-5008.
Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911-917.
Brader, G., S. Sjoblom, H. Hyytiainen, K. Sims-Huopaniemi, and E. T. Palva. 2005. Altering substrate chain length specificity of an acylhomoserine lactone synthase in bacterial communication. J. Biol. Chem. 280:10403-10409.
Brock, M., C. Maerker, A. Schutz, U. Volker, and W. Buckel. 2002. Oxidation of propionate to pyruvate in Escherichia coli. Involvement of methylcitrate dehydratase and aconitase. Eur. J. Biochem. 269:6184-6194.
Buckman, B. O., M. M. Morrissey, and R. Mohan. 1998. Solution-phase parallel synthesis of benzoxazines using a polymer-supported carbodiimide. Tetrahedron Lett. 39:1487-1488.
Cataldi, T. R., G. Bianco, M. Frommberger, and P. Schmitt-Kopplin. 2004. Direct analysis of selected N-acyl-l-homoserine lactones by gas chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 18:1341-1344.
Cha, C., P. Gao, Y. C. Chen, P. D. Shaw, and S. K. Farrand. 1998. Production of acyl-homoserine lactone quorum-sensing signals by gram-negative plant-associated bacteria. Mol. Plant-Microbe Interact. 11:1119-1129.
Charlton, T. S., R. de Nys, A. Netting, N. Kumar, M. Hentzer, M. Givskov, and S. Kjelleberg. 2000. A novel and sensitive method for the quantification of N-3-oxoacyl homoserine lactones using gas chromatography-mass spectrometry: application to a model bacterial biofilm. Environ. Microbiol. 2:530-541.
Chhabra, S. R., C. Harty, D. S. Hooi, M. Daykin, P. Williams, G. Telford, D. I. Pritchard, and B. W. Bycroft. 2003. Synthetic analogues of the bacterial signal (quorum sensing) molecule N-(3-oxododecanoyl)-l-homoserine lactone as immune modulators. J. Med. Chem. 46:97-104.
Cronan, J. E. 2003. Bacterial membrane lipids: where do we stand Annu. Rev. Microbiol. 57:203-224.
Eberhard, A., A. L. Burlingame, C. Eberhard, G. L. Kenyon, K. H. Nealson, and N. J. Oppenheimer. 1981. Structural identification of autoinducer of Photobacterium fischeri. Biochemistry 20:2444-2449.
Flynn, D. L., R. V. Devraj, W. Naing, J. J. Parlow, J. J. Weidner, and S. L. Yang. 1998. Polymer-assisted solution phase (PASP) chemical library synthesis. Med. Chem. Res. 8:219-243.
Frommberger, M., P. Schmitt-Kopplin, F. Menzinger, V. Albrecht, M. Schmid, L. Eberl, A. Hartmann, and A. Kettrup. 2003. Analysis of N-acyl-l-homoserine lactones produced by Burkholderia cepacia with partial filling micellar electrokinetic chromatography-electrospray ionization-ion trap mass spectrometry. Electrophoresis 24:3067-3074.
Frommberger, M., P. Schmitt-Kopplin, G. Ping, H. Frisch, M. Schmid, Y. Zhang, A. Hartmann, and A. Kettrup. 2004. A simple and robust set-up for on-column sample preconcentration-nano-liquid chromatography-electrospray ionization mass spectrometry for the analysis of N-acylhomoserine lactones. Anal. Bioanal. Chem. 378:1014-1020.
Fuqua, C., and A. Eberhard. 1999. Signal generation in autoinduction systems: synthesis of acylated homoserine lactones by LuxI-type proteins, p.211-230. In G. Dunny and S. C. Winans (ed.), Cell-cell signaling in bacteria. ASM Press, Washington, D.C.
Giacomini, A., V. Corich, F. J. Ollero, A. Squartini, and M. P. Nuti. 1992. Experimental conditions may affect reproducibility of the beta-galactosidase assay. FEMS Microbiol. Lett. 79:87-90.
Gould, T. A., H. P. Schweizer, and M. E. A. Churchill. 2004. Structure of the Pseudomonas aeruginosa acyl-homoserinelactone synthase LasI. Mol. Microbiol. 53:1135-1146.
Gould, T. A., W. T. Watson, S. H. Choi, H. P. Schweizer, and M. E. A. Churchill. 2004. Crystallization of Pseudomonas aeruginosa AHL synthase LasI using beta-turn crystal engineering. Acta Crystallogr. D 60:518-520.
Haller, T., T. Buckel, J. Retey, and J. A. Gerlt. 2000. Discovering new enzymes and metabolic pathways: conversion of succinate to propionate by Escherichia coli. Biochemistry 39:4622-4629.
Heath, R. J., and C. O. Rock. 1996. Inhibition of beta-ketoacyl-acyl carrier protein synthase III (FabH) by acyl-acyl carrier protein in Escherichia coli. J. Biol. Chem. 271:10996-11000.
Henke, J. M., and B. L. Bassler. 2004. Bacterial social engagements. Trends Cell Biol. 14:648-656.
Hentzer, M., H. Wu, J. B. Andersen, K. Riedel, T. B. Rasmussen, N. Bagge, N. Kumar, M. A. Schembri, Z. Song, P. Kristoffersen, M. Manefield, J. W. Costerton, S. Molin, L. Eberl, P. Steinberg, S. Kjelleberg, N. Hoiby, and M. Givskov. 2003. Attenuation of Pseudomonas aeruginosa virulence by quorum sensing inhibitors. EMBO J. 22:3803-3815.
Hesslinger, C., S. A. Fairhurst, and G. Sawers. 1998. Novel keto acid formate-lyase and propionate kinase enzymes are components of an anaerobic pathway in Escherichia coli that degrades l-threonine to propionate. Mol. Microbiol. 27:477-492.
Hoang, T. T., Y. Ma, R. J. Stern, M. R. McNeil, and H. P. Schweizer. 1999. Construction and use of low-copy number T7 expression vectors for purification of problem proteins: purification of Mycobacterium tuberculosis RmlD and Pseudomonas aeruginosa LasI and RhlI proteins, and functional analysis of purified RhlI. Gene 237:361-371.
Hoang, T. T., A. J. Kutchma, A. Becher, and H. P. Schweizer. 2000. Integration-proficient plasmids for Pseudomonas aeruginosa: site-specific integration and use for engineering of reporter and expression strains. Plasmid 43:59-72.
Hoang, T. T., S. A. Sullivan, J. K. Cusick, and H. P. Schweizer. 2002. Beta-ketoacyl acyl carrier protein reductase (FabG) activity of the fatty acid biosynthetic pathway is a determining factor of 3-oxo-homoserine lactone acyl chain lengths. Microbiology 148:3849-3856.
Hooi, D. S., B. W. Bycroft, S. R. Chhabra, P. Williams, and D. I. Pritchard. 2004. Differential immune modulatory activity of Pseudomonas aeruginosa quorum-sensing signal molecules. Infect. Immun. 72:6463-6470.
Juhas, M., L. Eberl, and B. Tummler. 2005. Quorum sensing: the power of cooperation in the world of Pseudomonas. Environ. Microbiol. 7:459-471.
Juhas, M., L. Wiehlmann, P. Salunkhe, J. Lauber, J. Buer, and B. Tummler. 2005. GeneChip expression analysis of the VqsR regulon of Pseudomonas aeruginosa TB. FEMS Microbiol. Lett. 242:287-295.
Kaplan, H. B., and E. P. Greenberg. 1985. Diffusion of autoinducer is involved in the regulation of the Vibrio fischeri luminescence system. J.Bacteriol. 163:1210-1214.
Kirwan, J. P., T. A. Gould, H. P. Schweizer, S. W. Bearden, R. C. Murphy, and M. E. A. Churchill. 2006. Quorum-sensing signal synthesis by the Yersinia pestis acyl-homoserine lactone synthase YspI. J. Bacteriol. 188:784-788.
Kutchma, A. J., T. T. Hoang, and H. P. Schweizer. 1999. Characterization of a Pseudomonas aeruginosa fatty acid biosynthetic gene cluster: purification of acyl carrier protein (ACP) and malonyl-coenzyme A:ACP transacylase (FabD). J. Bacteriol. 181:5498-5504.
Lewenza, S., M. B. Visser, and P. A. Sokol. 2002. Interspecies communication between Burkholderia cepacia and Pseudomonas aeruginosa. Can. J. Microbiol. 48:707-716.
Lithgow, J. K., A. Wilkinson, A. Hardman, B. Rodelas, F. Wisniewski-Dye, P. Williams, and J. A. Downie. 2000. The regulatory locus cinRI in Rhizobium leguminosarum controls a network of quorum-sensing loci. Mol. Microbiol. 37:81-97.
Lupp, C., and E. G. Ruby. 2004. Vibrio fischeri LuxS and AinS: comparative study of two signal synthases. J. Bacteriol. 186:3873-3881.
Marketon, M. M., M. R. Gronquist, A. Eberhard, and J. E. Gonzalez. 2002. Characterization of the Sinorhizobium meliloti sinR/sinI locus and the production of novel N-acyl homoserine lactones. J. Bacteriol. 184:5686-5695.
McClean, K. H., M. K. Winson, L. Fish, A. Taylor, S. R. Chhabra, M. Camara, M. Daykin, J. H. Lamb, S. Swift, B. W. Bycroft, G. Stewart, and P. Williams. 1997. Quorum sensing and Chromobacterium violaceum: exploitation of violacein production and inhibition for the detection of N-acylhomoserine lactones. Microbiology 143:3703-3711.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
Miller, M. B., and B. L. Bassler. 2001. Quorum sensing in bacteria. Annu. Rev. Microbiol. 55:165-199.
More, M. I., L. D. Finger, J. L. Stryker, C. Fuqua, A. Eberhard, and S. C. Winans. 1996. Enzymatic synthesis of a quorum-sensing autoinducer through use of defined substrates. Science 272:1655-1658.
Morin, D., B. Grasland, K. Vallee-Rehel, C. Dufau, and D. Haras. 2003. On-line high-performance liquid chromatography-mass spectrometric detection and quantification of N-acylhomoserine lactones, quorum sensing signal molecules, in the presence of biological matrices. J. Chromatogr. A 1002:79-92.
Murphy, R. C., R. M. Barkley, K. Zemski Berry, J. A. Hankin, K. Harrison, C. Johnson, J. Krank, A. McAnoy, C. Uhlson, and S. Zarini. 2005. Electrospray ionization and tandem mass spectrometry of eicosanoids. Anal. Biochem. 346:1-42.
Nyholm, S. V., and M. J. McFall-Ngai. 2004. The winnowing: establishing the squid-vibrio symbiosis. Nat. Rev. Microbiol. 2:632-642.
Orsatti, L., S. Di Marco, C. Volpari, A. Vannini, P. Neddermann, and F. Bonelli. 2002. Determination of the stoichiometry of noncovalent complexes using reverse-phase high-performance liquid chromatography coupled with electrospray ion trap mass spectrometry. Anal. Biochem. 309:11-18.
Parlow, J. J., D. A. Mischke, and S. S. Woodard. 1997. Utility of complementary molecular reactivity and molecular recognition (CMR/R) technology and polymer-supported reagents in the solution-phase synthesis of heterocyclic carboxamides. J. Org. Chem. 62:5908-5919.
Parsek, M. R., D. L. Val, B. L. Hanzelka, J. E. J. Cronan, and E. P. Greenberg. 1999. Acyl homoserine-lactone quorum-sensing signal generation. Proc. Natl. Acad. Sci. USA 96:4360-4365.
Passador, L., K. D. Tucker, K. R. Guertin, M. P. Journet, A. S. Kende, and B. H. Iglewski. 1996. Functional analysis of the Pseudomonas aeruginosa autoinducer PAI. J. Bacteriol. 178:5995-6000.
Pearson, J. P., K. M. Gray, L. Passador, K. D. Tucker, A. Eberhard, B. H. Iglewski, and E. P. Greenberg. 1994. Structure of the autoinducer required for expression of Pseudomonas aeruginosa virulence genes. Proc. Natl. Acad. Sci. USA 91:197-201.
Ravn, L., A. B. Christensen, S. Molin, M. Givskov, and L. Gram. 2001. Methods for detecting acylated homoserine lactones produced by Gram-negative bacteria and their application in studies of AHL-production kinetics. J. Microbiol. Methods 44:239-251.
Raychaudhuri, A., A. Jerga, and P. A. Tipton. 2005. Chemical mechanism and substrate specificity of RhlI, an acylhomoserine lactone synthase from Pseudomonas aeruginosa. Biochemistry 44:2974-2981.
Ritchie, A. J., A. Jansson, J. Stallberg, P. Nilsson, P. Lysaght, and M. A. Cooley. 2005. The Pseudomonas aeruginosa quorum-sensing molecule N-3-(oxododecanoyl)-L-homoserine lactone inhibits T-cell differentiation and cytokine production by a mechanism involving an early step in T-cell activation. Infect. Immun. 73:1648-1655.
Schaefer, A. L., B. L. Hanzelka, M. R. Parsek, and E. P. Greenberg. 2000. Detection, purification, and structural elucidation of the acylhomoserine lactone inducer of Vibrio fischeri luminescence and other related molecules. Methods Enzymol. 305:288-301.
Schuster, M., C. P. Lostroh, T. Ogi, and E. P. Greenberg. 2003. Identification, timing, and signal specificity of Pseudomonas aeruginosa quorum-controlled genes: a transcriptome analysis. J. Bacteriol. 185:2066-2079.
Schweizer, H. P., T. T. Hoang, K. L. Propst, H. R. Ornelas, and R. R. Karkhoff-Schweizer. 2001. Vector design and development of host systems for Pseudomonas, p. 69-81. In J. K. Setlow (ed.), Genetic engineering. Kluwer-Academic/Plenum, New York, N.Y.
Shaw, P. D., G. Ping, S. L. Daly, C. Cha, J. E. J. Cronan, K. L. Rinehart, and S. K. Farrand. 1997. Detecting and characterizing N-acyl-homoserine lactone signal molecules by thin-layer chromatography. Proc. Natl. Acad. Sci. USA 94:6036-6041.
Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff. 1990. Use of the T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185:60-89.
Taga, M. E., and B. L. Bassler. 2003. Chemical communication among bacteria. Proc. Natl. Acad. Sci. USA 100:14549-14554.
Tateda, K., Y. Ishii, M. Horikawa, T. Matsumoto, S. Miyairi, J. C. Pechere, T. J. Standiford, M. Ishiguro, and K. Yamaguchi. 2003. The Pseudomonas aeruginosa autoinducer N-3-oxododecanoyl homoserine lactone accelerates apoptosis in macrophages and neutrophils. Infect. Immun. 71:5785-5793.
Teplitski, M., A. Eberhard, M. R. Gronquist, M. Gao, J. B. Robinson, and W. D. Bauer. 2003. Chemical identification of N-acyl homoserine lactone quorum-sensing signals produced by Sinorhizobium meliloti strains in defined medium. Arch. Microbiol. 180:494-497.
Wagner, V. E., D. Bushnell, L. Passador, A. I. Brooks, and B. H. Iglewski. 2003. Microarray analysis of Pseudomonas aeruginosa quorum-sensing regulons: effects of growth phase and environment. J. Bacteriol. 185:2080-2095.
Watson, W. T., F. V. Murphy, T. A. Gould, P. Jambeck, D. L. Val, J. E. Cronan, S. Beck von Bodman, and M. E. A. Churchill. 2001. Crystallization and rhenium MAD phasing of the acyl-homoserine lactone synthase EsaI. Acta Crystallogr. D 57:1945-1949.
Watson, W. T., T. D. Minogue, D. L. Val, S. Beck von Bodman, and M. E. A. Churchill. 2002. Structural basis and specificity of acyl-homoserine lactone signal production in bacterial quorum sensing. Mol. Cell 9:685-694.
White, S. W., J. Zheng, Y. M. Zhang, and K. L. Rock. 2005. The structural biology of type ii fatty acid biosynthesis. Annu. Rev. Biochem. 74:791-831.
Whitehead, N. A., A. M. Barnard, H. Slater, N. J. Simpson, and G. P. Salmond. 2001. Quorum sensing in Gram-negative bacteria. FEMS Microbiol. Rev. 25:365-404.
Yates, E. A., B. Philipp, C. Buckley, S. Atkinson, S. R. Chhabra, R. E. Sockett, M. Goldner, Y. Dessaux, M. Camara, H. Smith, and P. Williams. 2002. N-Acylhomoserine lactones undergo lactonolysis in a pH-, temperature-, and acyl chain length-dependent manner during growth of Yersinia pseudotuberculosis and Pseudomonas aeruginosa. Infect. Immun. 70:5635-5646.
Zhu, J., Y. Chai, Z. Zhong, S. Li, and S. C. Winans. 2003. Agrobacterium bioassay strain for ultrasensitive detection of N-acylhomoserine lactone-type quorum-sensing molecules: detection of autoinducers in Mesorhizobium huakuii. Appl. Environ. Microbiol. 69:6949-6953.(Ty A. Gould, Jake Herman,)