|Home | About | Journals | Submit | Contact Us | Français|
Many bacteria utilize acyl-homoserine lactones as cell to cell signals that can regulate the expression of numerous genes. Structural differences in acyl-homoserine lactones produced by different bacteria, such as acyl-side chain length and the presence or absence of an oxy group, make many of the commonly used detection bioassays impractical for broad range detection. Here we present a simple, broad range acyl-homoserine lactone detection bioassay that can be used to detect a wide range of these chemical signals. A plasmid (pEAL01) was constructed and transformed into Pseudomonas aeruginosa strain QSC105 to allow for detection of a broad range of acyl-homoserine lactones through induction of a lasB'-lacZ transcriptional fusion. Monitoring β-galactosidase activity from this bioassay showed that P. aeruginosa strain QSC105 (pEAL01) could detect the presence of eight acyl-homoserine lactones tested at physiological concentrations. This novel strain could also detect acyl-homoserine lactones from the extracts of four different bacteria that produce different acyl-homoserine lactones signals. These data indicate that strain QSC105 (pEAL01) can be used to detect a wide variety of acyl-homoserine lactones by a simple β-galactosidase assay and this bioassay could be a useful and inexpensive tool to quickly identify the presence of these signal molecules.
Greater than 50 species of gram-negative bacteria have been found to communicate intercellularly via acyl-homoserine lactone signals (AHLs) (Fuqua, 2001). These signals function as “quorum sensing” molecules in that their concentration is an indicator of the relative bacterial population density. When the signal concentration reaches a threshold level, it interacts with a transcriptional activator protein which will then induce specific genes. These cell to cell signals control a wide variety of bacterial functions such as symbiosis, virulence, competence, conjugation, antibiotic production, motility, sporulation, and biofilm formation (Miller and Bassler, 2001). The AHL signal systems are comprised of two genes, the first of which encodes a signal synthase protein that is a member of the LuxI-type family of proteins. The second gene encodes a member of the LuxR-type transcriptional activator protein family (see Fuqua et al., 1994 for a review). The activity of these transcriptional activators is dependent on their cognate AHL signal, which binds to the transcriptional activator protein and thereby leads to its activation.
AHL signals share a common base structure of a homoserine lactone ring but they are highly variable with regard to the attached acyl side chain. When comparing signals from different systems, the acyl side chain varies in length, oxidation state, and saturation (Fuqua et al., 1996). Such signal differences are what provides specificity for each LuxR-type transcriptional activator. The specificity of LuxR-type proteins is the limiting factor for detecting signals with the use of bioassays.
The ability to detect AHL signals has enabled quorum sensing systems to be identified in many gram-negative bacteria. Several bioassays have been designed for detecting AHLs (reviewed in Steindler and Venturi, 2007). However, most are limited in their detection range due the incorporation of only one quorum sensing system. Improved bioassays have been developed for detection of a wide range of AHLs (Farrand et al., 2002; Zhu et al., 2003). While these can detect a variety of AHLs, they are limited in their ability to detect combinations of several important AHLs (Shaw et al., 1997). One broad specificity assay utilizes thin layer chromatography in conjunction with a bioassay strain (Zhu et al., 2003). This is an excellent assay, but it requires the use of multiple organic solvents, something that some laboratories may find prohibitive. Other AHL detection bioassays have been developed that required multiple plasmids for functionality (Farrand et al., 2002), which may be prohibitive if the system needed to be transferred to another organism for a given application. We wanted to develop a simple and inexpensive broad-specificity bioassay for the detection of acyl-homoserine lactone signals that is based on measuring enzymatic activity from a lacZ reporter fusion. The bioassay only requires extracts and a spectrophotometer to measure light absorbance at 660 nm and 420 nm. This single plasmid bioassay is based on the two Pseudomonas aeruginosa quorum sensing systems (las and rhl) which use two AHLs that are structurally different. In P. aeruginosa, the las and rhl quorum sensing systems regulate numerous genes, including lasB, which encodes LasB elastase (Gambello and Iglewski, 1991; Pearson et al., 1997; Schuster et al., 2004; Wagner et al., 2004). Both the las and rhl quorum sensing systems positively control lasB transcription through their respective cell to cell signals, N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C12-HSL) and N-butyryl-L-homoserine lactone (C4-HSL) (Pearson et al., 1994; Pearson et al., 1995; Pearson et al., 1997). These signals associate with, and thereby activate, the respective transcriptional activators, LasR and RhlR. Both LasR and RhlR can also be activated by acyl-homoserine lactone signals that are similar to their cognate signals (Ochsner and Reiser, 1995; Passador et al., 1996). Because of the differences in their cognate signals, we felt that an assay that used both LasR and RhlR would allow for the detection of a broad range of AHL signals. In this manuscript, we report the construction of a plasmid that contains tacp-lasR, tacp-rhlR, and lasB'-lacZ, and its ability to confer a non-signaling P. aeruginosa strain the ability to respond to a wide-range of AHL signals.
Plasmid pEAL01 was constructed as follows (see Fig. 1). First, plasmid pECP16 was digested with HindIII (Invitrogen) and a 2120 bp fragment containing tacp-lasR was purified from an agarose gel using a GeneClean®III kit. Plasmid pECP16 was constructed by ligating the 1900 bp fragment from a ScaI and BamHI digest of pECP8 (Wade et al., 2005), which contained tacp-lasR, into the SalI digested plasmid pJPP8 (Pearson et al., 1997). The resulting 9.4 kb plasmid contains both tacp-rhlR and tacp-lasR. The plasmid pECP8 was constructed by ligating the PCR product of pKDT37 (Passador et al., 1996), containing the lasR coding region, into EcoRI-digested pEX1.8. The resulting plasmid contains a tacp-lasR fusion in which there is optimal spacing (9 bp) between the ribosome binding site from pEX1.8 and the start codon of lasR. The HindIII fragment from pECP16 was then ligated into the HindIII site on pECP64 (Pearson et al., 1997). Plasmid pECP64 contains tacp-lasR and lasB'-lacZ fusions and is also a pEX1.8-based vector. The resulting plasmid, pEAL01, contained tacp-lasR, tacp-rhlR, lasB'-lacZ, bla, lacI, and ColE1 origin of replication for enteric bacteria, such as Escherichia coli, and the origin of replication for Pseudomonads. The plasmid was transformed by electroporation (Choi et al., 2006) into P. aeruginosa strain QSC105 (Chugani et al., 2001), which is a lasI, rhlI, pqsH mutant. This strain was chosen because it cannot synthesize AHL signals or the Pseudomonas quinolone signal, which eliminates any background from endogenously produced signals. P. aeruginosa strain QSC105 (pEAL01) was stored at -80°C in 10% skim milk (Fisher Scientific) and plated before each experiment.
β-Galactosidase (β-Gal) assays were performed as follows. Cultures of P. aeruginosa strain QSC105 (pEAL01) were inoculated from freshly grown plates and grown overnight at 37°C with vigorous shaking (≥ 180 rpm) in 10 ml Peptone Tryptic Soy Broth (PTSB) (Ohman et al., 1980) supplemented with 200 μg/ml carbenicillin (Sigma). The following day, a subculture was inoculated to a final OD660 of 0.08 and allowed to grow for eight hours. The culture was then diluted into fresh media to a final OD660 of 0.1 and 1 ml aliquots were transferred to tubes containing dried signals. The 1 ml cultures containing varying concentrations of signals were grown for 18 hours at 37°C with vigorous shaking and β-gal assays were performed as described by Miller (1972). All β-gal assays were performed in duplicate and data presented are the mean from at least three independent experiments. All signals were tested at 0.01, 0.1, 1, 10 and 50 μM. The tested concentration does not exceed 50 μM in order to maintain a concentration within the physiological concentration range found in nature.
The following cell to cell signals or signal-like compounds were tested at various concentrations (see Fig. 2): N-3-oxo-dodecanoyl-homoserine lactone (3-oxo-C12-HSL) (Quorum Sciences, Inc.), N-dodecanoyl-homoserine lactone (C12-HSL) (Sigma/Fluka), N-decanoyl-homoserine lactone (C10-HSL) (Sigma/Fluka), N-octanoyl-homoserine lactone (C8-HSL) (Sigma/Fluka), N-heptanoyl-homoserine lactone (C7-HSL) (Sigma/Fluka), N-3-oxo-hexanoyl-homoserine lactone (3-oxo-C6-HSL) (gift from E. P. Greenberg), N-hexanoyl-homoserine lactone (C6-HSL) (Sigma/Fluka), and N-butanoyl-homoserine lactone (C4-HSL) (Sigma/Fluka). All signals were maintained in acidified ethyl acetate (Pesci et al., 1999) at -20°C until used.
Extractions were performed on P. aeruginosa wild-type strain PAO1 grown in PTSB at 37°C, Burkholderia cepacia strain J2549 grown on Luria Bertani media (LB) at 37°C, Erwinia carotovora strain Ecc71 grown in LB at 30°C, and Agrobacterium tumefaciens strain R10 (pCF218) (Zhu et al., 1998) grown in LB supplemented with 2 μg/ml tetracycline (this is a wild type strain that overproduces the 3-oxo-C8-HSL). Overnight cultures from freshly streaked plates were sub-cultured to a final OD660 of 0.08 and allowed to grow for eight hours. A 50 ml subculture was then inoculated to a final OD660 of 0.1. This was grown for 18 hours at the appropriate temperature, with shaking, before the extraction with acidified ethyl acetate was performed. The cultures were centrifuged at 16,000 × g for 10 minutes and the supernatant was extracted twice with an equal volume of acidified ethyl acetate as described elsewhere (Pesci et al., 1999). After the addition of 10% sodium sulfate, the extract was dried and resuspended in 1 ml acidified ethyl acetate. For the B. cepacia, E. carotovora, and A. tumefaciens strains, we tested an amount of extract derived from 10 ml, 1 ml, or 0.1 ml of culture supernatant. For P. aeruginosa strain PAO1 we tested an amount derived from 1 ml or 0.1 ml of culture supernatant, as this was more than enough to activate the bioassay. As in the previous bioassays, each extraction was repeated three separate times and the β-gal activity was measured.
The purpose of this study was to develop a simple bioassay that could detect a large range of AHL signals. We constructed a plasmid (pEAL01) that would constitutively express lasR and rhlR, and contained a lasB'-lacZ transcriptional fusion. We tested the ability of P. aeruginosa strain QSC105 (pEAL01) to detect eight structurally diverse AHLs that had acyl chains ranging from four to twelve carbons. We found that this bioassay strain was able to detect the presence of each AHL signal tested (Fig. 3). The amount of signal required to activate the lasB'-lacZ fusion ranged from 0.1 μM 3-oxo-C12-HSL to 50 μM C12-HSL (Fig. 3). Not surprisingly, the signals that strain QSC105 (pEAL01) were most sensitive to, 3-oxo-C12-HSL and C4-HSL, are produced by P. aeruginosa, and are the cognate signals for the LasR and RhlR transcriptional activator proteins encoded on plasmid pEAL01. Although the signal concentration required to activate the lasB'-lacZ fusion differed greatly depending on the signal, strain QSC105 (pEAL01) was able to detect each signal at a physiologically relevant concentration. The ability of strain QSC105 (pEAL01) to detect the AHL signals with acyl chains ranging from four to twelve carbons makes this strain an efficient and simple AHL detection bioassay.
Because the plasmid can replicate in Pseudomonads and enteric bacteria, we also tested the ability of the plasmid to confer the ability to detect 3-oxo-C12-HSL and C4-HSL to E. coli (data not shown). As expected, E. coli strain DH5α (pEAL01) was sensitive to both AHLs tested and could detect 3-oxo-C12-HSL at concentrations as low as 10 nM, indicating the assay functions well in an E. coli background. These data also confirm that the detection of AHLs by this bioassay is a result of the plasmid and not due to our host P. aeruginosa strain.
We then set out to learn whether our bioassay could be used to detect signals from culture supernatants. To test this, β-gal activity was determined after P. aeruginosa strain QSC105 (pEAL01) was exposed to a culture supernatant extract from four different species of AHL-producing bacteria. Supernatants taken from P. aeruginosa strain PAO1, B. cepacia strain J2549, E. carotovora strain Ecc71, and A. tumefaciens strain R10 (pCF218) were extracted with acidified ethyl acetate as described above. Fig. 4 shows that strain QSC105 (pEAL01) detected the presence of AHL signals in the extracts from all four bacteria. As would be expected, the addition of strain PAO1 extract resulted in the highest β-gal activity (Fig. 4a). An extract derived from 1 ml of strain PAO1 culture caused a very large activation of the assay (46,063 ± 4,590 Miller Units), which is most likely caused by a combination of the two AHL signals, 3-oxo-C12-HSL and C4-HSL, produced by P. aeruginosa (Brint and Ohman, 1995). The bioassay also detected the presence of high concentrations of AHLs or AHL-like compounds in 1 ml extracts of B. cepacia strain J2540 (2,360 ± 1,100 Miller units) and 10 ml extracts of A. tumefaciens strain R10 (pCF218) (20, 546 ± 4,808 Miller units; Fig. 4c). Interestingly, the bioassay had decreased responsiveness to the 10 ml extracts of B. cepacia strain J2540 as compared to the 1.0 ml extracts (Fig. 4b). This could be attributed to the 10 ml extract having a high concentration of a compound that inhibits P. aeruginosa responsiveness to AHL signals, and indicates the necessity for testing a range of extract concentrations. The bioassay was also able to detect an AHL or AHL-like molecules in extracts derived from 10 ml of supernatant from E. carotovora strain Ecc71 (1,888 ± 710 Miller units; Fig. 4d). This strain utilizes the CarI and ExpI synthases (Andersson et al., 2000) to produce 3-oxo-C6-HSL as its major AHL signal. This AHL signal is the most structurally divergent AHL tested as compared to those recognized by RhlR and LasR. Although the bioassay did not readily detect 3-oxo-C6-HSL in the 1 or 0.1 ml strain Ecc71 extracts, the fact that it was able to detect the signal in the extract derived from 10 ml of culture supernatant shows that the bioassay would be of practical use for detection of AHLs unrelated to those produced by P. aeruginosa.
The AHLs tested are produced by many different bacteria and differ in their structure, from 3-oxo-C12-HSL to C4-HSL. A bioassay based on one quorum sensing system would have a limited detection range and may not respond to AHLs that greatly differ by acyl side chain length or in the presence or absence of an oxy group. Previous work (Zhu et al, 1997) has shown that ovevexpression of the LuxR homolog, TraR, allowed for the detection of numerous AHLs besides its cognate signal. Our data are consistent with these findings and the overexpression of both LasR and RhlR in plasmid pEAL01 further broadens the potential detection of AHLs by our bioassay. The expression of both LasR and RhlR in plasmid pEAL01 gives this bioassay the ability to recognize a range of structurally diverse AHLs.
Taken together, the data support the notion that strain QSC105 (pEAL01) can be used to detect a broad range of AHL signals. Our data also shows that the plasmid could be used to transform E. coli and potentially any Pseudomonad or enteric bacteria into an AHL bioassay strain if so desired. This method provides a simple alternative for previously described bioassays that have a limited detection range or require more complicated experimentation. To our knowledge, this is the first AHL bioassay that has been shown to recognize all AHLs produced by P. aeruginosa with high sensitivity as well as the other AHL signals tested.
We thank Daniel W. Martin, Arun K. Chatterjee, and Stephen C. Winans for providing B. cepacia strain J2450, E. carotovora strain Ecc71, and A. tumefaciens strain R10 (pCF218), respectively. We also thank E. Peter Greenberg for supplying N-(3-oxo-hexanoyl)-L-homoserine lactone, and John M. Farrow, 3rd, Claire A. Lindsey, Jennifer M. Gaines, and James P. Coleman for critically reviewing the manuscript. This work was supported by a research grant from the National Institute of Allergy and Infectious Diseases (grant R01-AI076272).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.