OUP user menu

Detection of quorum-sensing N-acyl homoserine lactone signal molecules by bacterial biosensors

Laura Steindler, Vittorio Venturi
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00501.x 1-9 First published online: 1 January 2007

Abstract

Many Gram-negative bacteria use N-acyl homoserine lactones (AHLs) as quorum-sensing (QS) signal molecules. AHL QS has been the subject of extensive investigation in the last decade and has become a paradigm for bacterial intercellular signaling. Research in AHL QS has been considerably aided by simple methods devised to detect AHLs using bacterial biosensors that phenotypically respond when exposed to exogenous AHLs. This article reviews and discusses the currently available bacterial biosensors which can be used in detecting and studying the different AHLs.

Keywords
  • N-acyl homoserine lactone
  • quorum sensing
  • signal molecules
  • bacterial communication

Introduction

In most bacteria, a major level of regulation exists involving intercellular communication via the production and response to signal molecules (Camilli & Bassler, 2006). This cell-density-dependent regulation has been termed quorum sensing (QS) (Fuqua et al., 1994) and confers a significant advantage to a community of bacteria [reviews: (Bassler, 2002; Camara et al., 2002; Fuqua & Greenberg, 2002; Waters & Bassler, 2005)].

In Gram-negative bacteria, a typical QS system usually involves the production and response to an acylated homoserine lactone (AHL; Fig. 1). An AHL-QS system is most commonly mediated by two proteins belonging to the LuxI-LuxR protein families, which originate from the LuxI-AHL synthase and LuxR-AHL-response regulator involved in regulating bioluminescence in Vibrio fischeri in relation to cell density. AHLs produced by different bacteria differ only in the length of the acyl-chain moiety and substitution at position C3, which can be either unmodified or carry an oxo- or hydroxyl group (Fig. 1). AHLs interact directly, at quorum concentration, with the cognate LuxR-type protein and this protein-AHL complex can then bind at specific gene promoter sequences called lux-boxes affecting expression of QS target genes. In most cases, the LuxR-AHL complex positively regulates the luxI AHL synthase family gene, creating a positive induction loop. LuxR-type proteins display preferential binding for the AHL produced by the cognate LuxI-family protein, guaranteeing a good degree of selectivity. To some extent, LuxR-family proteins can also respond to AHLs of different length and substitution of the acyl-chain moiety, raising important implications for the role of AHLs in interspecies communication. In some cases, bacteria possess multiple LuxI/R systems, producing multiple AHLs that can be hierarchically organized.

Figure 1

Construction and use of a bacterial AHL biosensor. At the top of the diagram, the structure of AHLs is shown. The letter ‘R’ refers to the moiety at position C3, which can be either unmodified or carries an oxo- or hydroxyl group. The letter ‘n’ refers to the length of the acyl chain, which is most commonly from 4 to 12 and in some cases bacteria can produce AHLs having chains of 14 to 18 carbons. The exogenous AHL interacts with a LuxR family protein inside the bacterial biosensor (non-AHL producer), which results in the transcription of a reporter gene(s) from a LuxR family-AHL regulated promoter as shown by the open triangle. The LuxR family gene is usually expressed from a constitutive promoter as shown with a filled triangle. The AHL biosensor can then be used in (i) a plate ‘T’ streak assay, resulting in the expression of the reporter phenotype in a gradient near the tested strain as indicated, (ii) a TLC overlay technique after separation of AHLs from a bacterial extract resulting in the detection of the reporter phenotype at the place where AHLs are found, (iii) a quantification assay by measuring levels of the reporter phenotype upon exposure to spent bacterial supernatants or bacterial organic extracts and (iv) an in vivo assay using gfp-based biosensors. See text for all details.

Bacterial AHL biosensors

The very large number of AHL QS systems identified has been rendered possible mainly via the use of bacterial biosensors that are able to detect the presence of AHLs. These biosensors do not produce AHLs and contain a functional LuxR-family protein cloned together with a cognate target promoter (usually the promoter of the cognate luxI synthase), which positively regulates the transcription of a reporter gene (e.g. bioluminescence, β-galactosidase, green-fluorescent protein and violacein pigment production; Fig. 1). The aim of this review is to summarize the various AHL bacterial biosensors currently available and their use in the identification and study of AHL QS systems.

The first step in determining whether a bacterial strain contains a LuxI/R QS system is to test for the production of AHL signal molecules. Each AHL biosensor relies on a particular LuxR family protein, thus displaying specificity towards the cognate AHL and in some cases to closely related AHLs. As many biosensors detect a narrow range of AHLs, it is essential, when testing a bacterium for AHL production, to use several biosensors, each responding to AHLs with different structural features. AHL biosensor strains can be used in different ways; firstly, the tester strain can be streaked and grown on solid media close to the biosensor to form a ‘T’ and the phenotypic change associated with the presence of exogenous AHLs will be observed as a gradient with most response observed at the meeting point of the two strains (Fig. 1). Secondly, AHLs can be extracted from spent supernatants of late exponential phase cultures (Shaw et al., 1997; Schaefer et al., 2000), and partial characterization can be carried out by TLC on C18 reversed-phase plates. This organic extraction increases many-fold the sensitivity of biosensors; AHLs generally partition into the organic phase, and the solvent is removed by drying. The TLC plates are loaded with the sample extracts and with different standards and, after chromatography, overlaid with a soft-agar suspension of the AHL biosensor strain [Fig. 1; (McClean et al., 1997; Shaw et al., 1997; Schaefer et al., 2000)]. Each AHL migrates with a characteristic mobility and results in a spot shape of response detected in a way depending on the reporter of the biosensor strain. The 3-oxo-AHLs produce tear-shaped spots, whereas alkanoyl-AHLs and 3-OH-AHLs migrate and form well-defined circles. It is sometimes advantageous that before TLC, the sample extracts are further purified by C18-reverse phase HPLC and the resulting fractions tested for activity against AHL-biosensors either directly or via TLC (Schaefer et al., 2000). Separation by TLC coupled with detection by AHL-biosensors gives a rapid and direct visual index of the AHL(s) produced by the tester bacteria. AHLs cannot be unambiguously identified using TLC. However, their chromatographic properties can be used to assign tentative structures, as Rfs calculated for the samples can be compared with Rfs of AHL standards. AHL structures are unequivocally determined on the basis of spectroscopic properties (Schaefer et al., 2000) including MS and nuclear magnetic resonance spectroscopy (NMR).

Several of the AHL biosensors can also be used for quantifying AHLs by measuring the activity of the reporter system present in the biosensor bacterial strain. This is useful for studying regulation of AHL synthesis and for identifying strain-level differences in AHL production. In order to quantify accurately one must determine, using the synthetic AHL, the minimal amount of AHL required for a response as well as the amount necessary for a saturated response in order to plot the linear dose response.

Biosensors for short and medium acyl chain AHLs

Some AHL biosensors have been constructed based on several LuxR-family proteins that are able to detect AHLs having acyl chains of C4 to C8 in length. A commonly used biosensor is based on Chromobacterium violaceum, a Gram-negative water and soil bacterium that produces the antibacterial purple pigment violacein. Chromobacterium violaceum regulates violacein production via the CviI/R AHL QS system, which produces and responds to C6-AHL (McClean et al., 1997). McClean (1997) constructed C. violaceum CV026, a violacein and AHL-negative double miniTn5 mutant. One transposon is inserted into the cviI AHL synthase gene, and the other is inserted into a putative violacein repressor locus. Exposure of strain CV026 to exogenous AHLs, which are able to interact with CviR, results in rapid production of a visually clear purple pigmentation. Unsurprisingly, the most active agonist AHL for CV026 is C6-AHL, the natural C. violaceum AHL; other AHLs that induce reasonably well include C6-3-oxo-AHL and C8-AHL (both sixfold less active than C6-AHL), C8-3-oxo-AHL (11-fold less active) and C4-AHL (30-fold less active). Strain CV026 responds very poorly to C4-3-oxo-AHL, and AHLs with acyl chains of C10 and longer have little or no agonist activity. In addition, all 3-hydroxy-AHLs are not detected by strain CV026. This strain is well suited for detection on solid media via a ‘T’ streak analysis as well as the TLC soft-agar overlay technique.

Several other biosensors rely on a plasmid construct harboring the luxCDABE operon of Photorhabdus luminescens resulting in bioluminescence as a reporter system (Winson et al., 1998b). These plasmids are usually harbored in Escherichia coli, which does not produce AHLs. Plasmid pSB401 (Winson et al., 1998a) and pHV200I (Pearson et al., 1994) are both based on LuxR of V. fischeri and cognate luxI promoter controlling luxCDABE expression. They are most sensitive to cognate C6-3-oxo-AHL and display good sensitivity towards C6-AHL, C8-3-oxo-AHL and C8-AHL. Little agonist activity was observed by C4-AHLs, C10- and longer acyl chain AHLs. The presence of AHLs therefore induces bioluminescence, which, in a TLC analysis, can be conveniently detected by exposing the TLC overlaid with the biosensor to autoradiographic paper. These biosensors can also be used for ‘T’ streak analysis; however, they do need photon camera equipment and therefore are not as handy as strain CV026 (see above). These two LuxR biosensors, E. coli (pSB401) and E. coli (pHV200I), can both be used for quantifications of AHLs being particularly sensitive to low AHL concentrations; however, a luminometer is necessary in order to measure bioluminescence (Winson et al., 1998a). Plasmid pSB403 contains the same arrangement of pSB401 (i.e. luxR and promoter of luxI controlling luxCDABE expression) cloned in a wide-host range mobilizable plasmid, conferring the advantage that it can be harbored in several other Gram-negative bacteria, thus not being only restricted to E. coli (Winson et al., 1998a). Finally, a biosensor sensitive for C4-AHL is E. coli (pSB536); the pSB536 plasmid was constructed using the ahyR of Aeromonas hydrophyla and the cognate ahyI gene promoter fused to luxCDABE (Swift et al., 1997). Escherichia coli (pSB536) is most sensitive to cognate C4-AHL and can be used in a TLC analysis, together with AHL quantification studies. Similarly, another E. coli plasmid-based sensor responding to C4-AHL is pAL101 composed of rhlR and cognate promoter rhlI fused to luxCDABE (Lindsay & Ahmer, 2005). The RhlI/R AHL QS system belongs to Pseudomonas aeruginosa and C4-AHL is the cognate AHL. Importantly, this plasmid sensor works best if harbored in an E. coli sdiA gene mutant. SdiA is an orphan LuxR family protein present in E. coli that can activate the rhlI promoter, thus interfering with C4-AHL detection (Lindsay & Ahmer, 2005). Escherichia coli does not have an SdiA cognate LuxI family synthase and does not synthesize AHLs.

Biosensors for long acyl chain AHLs

AHL biosensors for specific detection of C10-AHL, C12-AHL and their 3-oxo derivatives are based on the LasI/R system of P. aeruginosa, which produces and responds to C12-3-oxo-AHL. Plasmid sensor pSB1075 contains the lasR gene and cognate lasI gene promoter controlling luxCDABE expression (Winson et al., 1998a). This plasmid can only be harbored in E. coli and can be conveniently used in TLC analysis responding well to C12-3-oxo-AHL, C10-3-oxo-AHL and C12-AHL. Another E. coli plasmid sensor also based on the LasI/R system is pKDT17 (Pearson et al., 1994). This plasmid contains lasR under control of the lac promoter and a lasB::lacZ translational fusion; hence, the response to exogenous AHLs is detected via β-galactosidase activity. The lasB gene codes for an elastase, which is regulated by the LasI/R AHL QS system. The E. coli (pKDT17) AHL biosensor can be used in TLC analysis and responds strongly to C12-AHL, C10-AHL and their 3-oxo derivatives; it does not detect any of the shorter chain and 3-hydroxy AHLs (Cha et al., 1998). Both sensors can be used for quantification of AHLs, pSB1075 with a luminometer (Winson et al., 1998a), and pKDT17 through the standard β-galactosidase activity determination (Pearson et al., 1994). Quantification of long-chain AHLs, in particular C12-3-oxo-AHL, can also be performed with P. aeruginosa PAO1 M71LZ, a lasI genomic knock-out mutant and contains a transcriptional fusion of the promoter of rsaL and reporter gene lacZ (Dong et al., 2005). The rsaL gene is directly regulated by the LasI/R AHL QS system (de Kievit et al., 1999; Rampioni et al., 2006). Conferring C12-3-oxo-AHL to PAO1 M71LZ will, therefore, result in rsaL transcription via LasR quantifiable through determination of β-galactosidase activity. This sensor also works well for C10-3-oxo-AHL.

Biosensors with a broad range of AHL detection

AHL biosensors are limited to the specificity requirements of the LuxR family protein; hence, most biosensors are restricted in the range of AHLs to which they can respond. Agrobacterium tumefaciens AHL biosensors based on the TraI/R QS system detect a broad range of AHLs and also display the greatest sensitivity towards these compounds (Cha et al., 1998; Farrand et al., 2002; Zhu et al., 2003). The TraI/R AHL QS system is localized in the Ti-plasmid, and it is very well characterized, producing and responding to C8-3-oxo-AHL. Quorum sensing in A. tumefaciens is involved in the regulation of conjugal transfer of many plasmids (Farrand et al., 2002; Von Bodman et al., 2003). AHL biosensor A. tumefaciens NT1 (pZLR4) consists of strain NT1 cured of the Ti plasmid and thus unable to produce AHLs, and plasmid pZLR4. The plasmid contains the traR gene and one of the tra operons, responsible for Ti plasmid conjugal transfer, containing a traG::lacZ reporter fusion, the transcription of which is known to be regulated by the TraI/R AHL QS system (Cha et al., 1998; Farrand et al., 2002). Of all the AHL biosensors constructed so far, the A. tumefaciens displays the broadest sensitivity to AHLs at the lowest concentrations. This β-galactosidase-based biosensor is particularly well suited for TLC analysis. It is so sensitive to many AHLs that it requires only small volumes of AHL extracts from spent supernatants (Farrand et al., 2002). This sensor can also be used by spotting colonies, culture supernatants or sample extracts onto an overlay of the sensor grown in a suitable medium containing X-Gal. After overnight incubation, the presence of AHLs will result in a blue zone around the site of application (Farrand et al., 2002). It is also particularly suited to the study of AHL production profiles of a large number of bacterial strains. This biosensor detects 3-oxo-substituted AHL-derivatives with acyl chain lengths from 4 to 12 carbons and also 3-unsubstituted AHLs, with the exception of C4-AHL. Importantly, this sensor can detect 3-hydroxy derivatives, more precisely C6-3-hydroxy-AHL, C8-3-hydroxy-AHL and C10-3-hydroxy-AHL (Shaw et al., 1997; Cha et al., 1998).

A similar A. tumefaciens biosensor was constructed, namely WCF47 (pCF218)(pCF372) (Zhu et al., 1998); the strain WCF47 has a mutation in traI and consequently does not produce AHLs. The plasmid pCF218 contains traR expressed from the tetR vector promoter, and the plasmid pCF372 contains the traI promoter transcriptionally fused to lacZ. Similar to the other A. tumefaciens biosensor, it can also be used in TLC analysis and AHL quantification using β-galactosidase assays (Cha et al., 1998; Zhu et al., 1998). This last biosensor has recently been modified in order to make it even more sensitive by overexpressing the TraR protein as having larger amounts of the regulator inside the cell results in broader and greater sensitivity towards AHLs (Zhu et al., 2003). The biosensor consists of a three-plasmid system (pJZ384)(pJZ410)(pJZ372) into A. tumefaciens KYC55 that lacks the Ti plasmid and hence does not produce AHLs (Zhu et al., 2003). Plasmid pJZ384 contains the traR gene under the control of the phage T7 promoter, pJZ410 contains the phage T7 RNA polymerase gene and pJZ372 contains the traI-lacZ reporter fusion. This biosensor has great sensitivity to a very broad range of AHLs, making it particularly useful for detecting extremely low concentrations of AHLs. Like the other A. tumefaciens biosensors, it can be used for TLC analysis. Extracts or spent supernatants can also be assayed directly, and responses/quantifications can be determined using a culture of the A. tumefaciens biosensor (Zhu et al., 2003).

Biosensors for 3-hydroxy-AHLs

To detect 3-hydroxy-AHLs, in addition to the biosensor of A. tumefaciens (see above), another specific sensor has been developed recently based on the PhzI/R AHL QS system of Pseudomonas fluorescens 2-79 (Khan et al., 2005). The PhzI/R system in P. fluorescens 2-79 is genetically linked to and regulates the expression of the phzABCDEFG operon responsible for the biosynthesis of the antimicrobial compound phenazine-1-carboxylate (Khan et al., 2005). PhzI of P. fluorescens 2-79 is responsible for producing six different AHLs, among which are C6-3-hydroxy-AHL, C8-3-hydroxy-AHL and C10-3-hydroxy-AHL. The quantitative dominant and cognate signal for PhzR of strain 2-79 was determined to be C6-3-hydroxy-AHL (Khan et al., 2005). The strain 2-79 PhzI/R-based biosensor consists of a two-plasmid system harbored in the wild-type P. fluorescens 1855 strain, which does not produce AHLs. One plasmid, pSF105, harbors the phzR gene under the control of the trc promoter, and the other plasmid, pSF107, contains the phzR-phzA divergent PhzR-regulated dual promoter region fused between oppositely oriented uidA and lacZ reporters, which are easily detectable by β-glucuronidase and β-galactosidase activity, respectively. Using the phzA-lacZ reporter in pSF107, the AHL sensor responded best to C6-3-hydroxy-AHL with a 10-fold less sensitivity to C8-3-hydroxy-AHL (Khan et al., 2005). This sensor can be used in TLC analysis and for quantification purposes measuring either β-galactosidase or β-glucuronidase activities.

Biosensors for uncommon AHLs

A few bacterial species, mostly belonging to Alphaproteobacteria, have been reported to produce AHLs with acyl chains containing more than 12 carbons including Paracoccus denitrificans, Rhodobacter capsulatus, Rhizobium leguminosarum and Sinorhizobium meliloti (Schaefer et al., 2002; Llamas et al., 2004). Detection of these long-chain AHLs is not possible using the AHL biosensors described so far and a radiotracer technique (incorporation of 14C label from [14C]methionine into AHLs) was used in order to demonstrate that some of these bacteria produce long-chain AHLs (Marketon et al., 2002; Schaefer et al., 2002; Llamas et al., 2004). This led scientists to construct a specific AHL sensor for long-chain AHLs. Llamas (2004) constructed a biosensor based on the SinI/R system of S. meliloti; this system is specifically responsible for the production of long-chain AHLs including C12-AHL, C14-3-oxo-AHL, C16-AHL and C18-AHL, making it a logical candidate for the development of a specific long-chain AHL biosensor strain. The SinI/R AHL QS system plays a role in the synthesis of a low-molecular-weight exopolysaccharide and in nodulation efficiency (Marketon et al., 2003). The SinI/R-based biosensor was constructed by disrupting the sinI synthase and at the same time creating a sinI-lacZ transcriptional fusion in S. meliloti (Llamas et al., 2004). The resulting S. meliloti sinI::lacZ strain can be used in TLC (these long-chain AHLs, however, remain close to the TLC origin) and ‘T’ streak plate analysis for both detection and also quantification by β-galactosidase activity upon exposure to AHLs, extracts or spent supernatants. While this sensor is particularly sensitive to C14-3-oxo-AHL, C16:1-3-oxo-AHL, C16-AHL and C16:1-AHL, it is not very sensitive to C14-AHL and cannot detect C18-AHL. It was rendered more sensitive by overexpressing the SinR regulator from an arabinose-inducible promoter (Llamas et al., 2004). Plasmid pJNSinR contains sinR expressed from PBAD and in the presence of arabinose, S. meliloti sinI::lacZ (pJNSinR) showed an approximately twofold increase in the sensitivity and detection of long-chain AHLs when compared with S. meliloti sinI::lacZ. The specificity of the two sensors, however, did not change as both displayed a response to similar long-chain AHLs. Neither sensor could detect AHLs having acyl chains shorter than 14 carbons. Finally, the A. tumefaciens biosensor strains can also detect some of the long-chain AHLs produced by S. meliloti (Marketon et al., 2002; Llamas et al., 2004), and therefore can also be used if testing bacterial strains that produce unusually long-chain AHLs.

Biosensors able to detect AHLs at the single-cell level

Until 1999, the understanding of AHL QS systems in bacteria was mainly based on in vitro data. The ability to monitor AHL molecules in vivo was primarily determined by creation of AHL-sensor plasmids that utilize a nontoxic reporter gene that could be detected by epifluorescent microscopy: the gene encoding for the green fluorescent protein (gfp). The biosensors contained either a stable gfp-encoding gene (gfpmut3*) or an unstable version [gfp(ASV)] useful for detection of transient AHL expression. These sensors can be used in T-streak and TLC assays with the use of either epifluorescent microscopes or dark boxes equipped with blue illumination and filters to detect the gfp emission. The real advantage of these sensors, however, was their use in single-cell detection of AHL production. Andersen (2001) produced the plasmids pJBA130 and pJBA132, which contained components of the V. fischeri QS and either the stable or unstable version of gfp, respectively. These plasmids have broad host-range replicons and can thus replicate in most gram-negative bacteria. They were tested in E. coli, Serratia liquefaciens, Pseudomonas aureofaciens, and P. aeruginosa, strains in which AHL communication could be detected successfully. Plasmid pJBA130 was found to have some background fluorescence, while pJBA132 was completely nonfluorescent in non-AHL-producing strains. Strains containing pJBA132 are highly sensitive upshift responders (in the presence of AHLs, they respond with gfp production in 15–30 min), while the downshift of the response (disappearance of gfp in the absence of AHL) is much slower. They respond best to C6-3-oxo-AHL, but also respond to C6-AHL, C8-AHL and C12-3-oxo-AHL, at 10-fold lower sensitivity, and at least 500-fold less sensitively to C4-AHL. Such and similar sensors were utilized by Wu (2000) to visualize AHL production by P. aeruginosa in the lungs of infected mice.

Two other gfp-based AHL sensor plasmids that respond to a different spectrum of AHL molecules are plasmids pKR-C12 and pAS-C8 (Riedel et al., 2001). For both sensors, the unstable variant of gfp, gfp (ASV) (Andersen et al., 1998), was used, and they were inserted into the broad-host-range plasmid pBBR1MCS-5, enabling the transfer of the constructs to various AHL-negative strains. The sensor plasmid pKR-C12 is based on components of the P. aeruginosa PAO1 Las system. It contains a translational fusion of the lasB elastase gene of P. aeruginosa to gfp (ASV) together with, in the opposite direction, the lasR gene placed under the control of a lac-type promoter. Sensors containing this plasmid are most sensitive to cognate C12-3-oxo-AHL. The sensor plasmid pAS-C8 is based on components of the Burkholderia cepacia CepI/R system, and contains the cepI promoter translationally fused to gfp (ASV) together with, in the opposite direction, the cepR gene (coding for the cognate C8-AHL receptor protein) placed under the control of Plac. Sensors containing this plasmid are most sensitive to C8-AHL. Plasmids pKR-C12 and pAS-C8 were utilized to study the interspecies communication between P. aeruginosa and B. cepacia, which are two opportunistic pathogens that often coinfect the lungs of cystic fibrosis patients (Riedel et al., 2001).

Gfp-based AHL sensors were also useful to reveal the communication of different bacterial populations of the rhizosphere consortium of tomato plants (Steidle et al., 2001). The plasmids' strain background has a significant effect on the detection limits of the AHL. For example, pKR-C12 and pAS-C8 detected the lowest AHL concentrations in F117 (the AHL-negative derivative of rhizosphere Pseudomonas putida IsoF), while pJBA132 was more effective in MG44 (the AHL-negative derivative of S. liquefaciens MG1). In these backgrounds, it was found that pKR-C12 was most sensitive to C12-3-oxo-AHL, but also to C10-3-oxo-AHL, C10-AHL and C12-AHL, albeit with reduced sensitivity. For pAS-C8, the highest sensitivity was for C8-AHL, and at lower efficiency this sensor also detected C10-AHL, C10-3-oxo-AHL, C6-AHL, C12-AHL and C12-3-oxo-AHL. Finally, pJBA132 was mainly activated by C6-3-oxo-AHL, followed by C6-AHL, C10-3-oxo-AHL, C12-3-oxo-AHL and C8-AHL. Gfp-based sensors proved useful in studies analyzing the AHL-degradation activities of a large number of strains (Jafra & van der Wolf, 2004).

Reasons for biosensors not detecting AHLs

A negative result may indicate that the tested bacterial strain is not producing AHLs, but for several reasons it is not possible to conclude from the nonappearance of a signal that it does not produce AHL(s). Bacteria might produce structurally novel AHLs that are simply not detectable by the biosensors used; this is rather unlikely but cannot be excluded. The bacterial strain tested may produce AHLs at very low concentrations, which are below the threshold of sensitivity of the biosensor. In fact, AHL QS systems may themselves be regulated requiring specific environmental conditions in order to be switched on (Venturi, 2006). False-negative results using AHL biosensors can occasionally occur because of the bacteriostatic effect of compounds produced by the bacteria under investigation and/or by the bacterial sensor. In the latter case, this can sometimes be circumvented if the AHL sensor system is all contained within a plasmid that can be transformed or conjugated into the target bacteria. It is important to monitor the pH of the culture fluid or solid media as the AHLs will hydrolyze under basic conditions. Finally, long chain AHLs may exhibit low permeability through the cell membrane, which could affect purification and detection; in this case, AHL extraction using whole culture (i.e. cells and spent supernatant) should be performed (Llamas et al., 2004).

Concluding remarks

All AHL biosensors described in this review are listed in Table 1. AHL biosensor technology has been instrumental in the rapid isolation and study of AHL QS systems in Gram-negative bacteria. One of the reasons is that standard DNA hybridization technology as well as PCR strategies cannot be used in isolating AHL QS systems due to the surprising lack of nucleotide sequence similarity between luxI and luxR homologues.

View this table:
Table 1

AHL biosensors

Strain/Plasmid sensorHostBased on QS systemReporter systemBest responds toGood detectionCommonly used forReference
C. violaceum CV026C. violaceumCviI/R (C. violaceum)Violacein pigmentC6-AHLC6-3-oxo-AHL C8-AHL C8-3-oxo-AHL C4-AHLT.S., TLCMcClean et al. (1997)
pSB401E. coliLuxI/R (V. fisheri)luxCDABEC6-3-oxo-AHLC6-AHL C8-3-oxo-AHL C8-AHLTLC, Q.Winson et al. (1998a)
pHV200IE. coliLuxI/R (V. fisheri)luxCDABEC6-3-oxo-AHLC6-AHL C8-3-oxo-AHL C8-AHLTLC, Q.Pearson et al. (1994)
pSB403Broad host rangeLuxI/R (V. fisheri)luxCDABEC6-3-oxo-AHLC6-AHL C8-3-oxo-AHL C8-AHLTLC, Q.Winson et al. (1998a)
pSB536E. coliAhyI/R (A. hydrophyla)luxCDABEC4-AHLTLC, Q.Swift et al. (1997)
pAL101E. coli (sdiA mutant)RhlI/R (P. aeruginosa)luxCDABEC4-AHLTLC, Q.Lindsay & Ahmer (2005)
pSB1075E. coliLasI/R (P. aeruginosa)luxCDABEC12-3-oxo-AHLC10-3-oxo-AHL C12-AHLTLC, Q.Winson et al. (1998a)
pKDT17E. coliLasI/R (P. aeruginosa)β-galactosidaseC12-3-oxo-AHLC12-AHL C10-AHL C10-3-oxo-AHLTLC, Q.Pearson et al. (1994)
M71LZP. aeruginosalasILasI/R (P. aeruginosa)β-galactosidaseC12-3-oxo-AHLC10-3-oxo-AHLQ.Dong et al. (2005)
pZLR4A. tumefaciens NT1TraI/R (A. tumefaciens)β-galactosidaseC8-3-oxo-AHLAll 3-oxo-AHLs C6-AHL C8-AHL C10-AHL C12-AHL C14-AHL C6-3-hydroxy-AHL C8-3-hydroxy-AHL C10-3-hydroxy-AHLT.S., TLC, Q.Farrand et al. (2002)
pCF218 + pCF372A. tumefaciens WCF47TraI/R (A. tumefaciens)β-galactosidaseAs above with more sensitivityAs above with more sensitivityTLC, Q.Zhu et al. (1998)
pJZ384 + pJZ410 + pJZ372A. tumefaciens KYC55TraI/R (A. tumefaciens)β-galactosidaseAs above with more sensitivityAs above with more sensitivityTLC, Q.Zhu et al. (2003)
pSF105 + pSF107P. fluorescens 1855PhzI/R (P. fluorescens 2-79)β-glucuronidase β-galactosidaseC6-3-hydroxy-AHLC8-3-hydroxy-AHLTLC, Q.Khan et al. (2005)
S. melilotisinI::lacZS. melilotisinI::lacZSinI/R (S. meliloti)β-galactosidaseC14-3-oxo-AHLC16:1-3-oxo-AHL C16-AHL C16:1-AHL C14-AHLT.S., TLC, Q.Llamas et al. (2004)
pJNSinRS. melilotisinI::lacZSinI/R (S. meliloti)β-galactosidaseAs above with more sensitivityAs above with more sensitivityT.S., TLC, Q.Llamas et al. (2004)
pAS-C8Broad host rangeCepI/R (B. cepacia)gfpC8-AHLC10-AHLSingle cellRiedel et al. (2001)
pKR-C12Broad host rangeLasI/R (P. aeruginosa)gfpC12-3-oxo-AHLC10-3-oxo-AHLSingle cellRiedel et al. (2001)
pJBA-132Broad host rangeLuxI/R (V. fisheri)gfpC6-3-oxo-AHLC6-AHL C8-AHL C10-AHLSingle cellAndersen et al. (2001)
  • T.S. refers to ‘T’ streak analysis in solid media.

  • Note that TLC analysis of unusually long AHLs results in no migration.

  • Using epifluorescence microscopy.

  • Q refers to quantification.

There are limitations, and one must be cautious in the interpretation of data obtained with AHL biosensors. Spent medium extracts might contain non-AHL compounds that could potentially interfere or activate the response of biosensors (Holden et al., 1999; Bauer & Mathesius, 2004). In addition, from the negative response of AHL biosensors one cannot rigorously conclude that the screened bacterial strain does not produce AHLs. AHLs are usually active at very low concentrations and biosensor technology allows for the quantification of AHLs. As different LuxR homologues have diverse affinities with different AHLs, it is not accurate to compare the intensity of a response of one AHL with the response obtained with a different AHL.

The development and use of AHL biosensor technology will be instrumental in future studies that will involve interaction and communication with eukaryotes, detection of AHLs in environmental samples and of bacterial interspecies and intergeneric communication.

Acknowledgements

L.S. is benefiting from an ICGEB fellowship. We wish to thank coresearchers I. Bertani, G. Degrassi, G. Devescovi and S. Ferluga for stimulating discussions. V.V's laboratory is supported by ICGEB and by the Italian Cystic Fibrosis Research Foundation (Verona, Italy). V.V's laboratory in the ICGEB Biosafety Outstation of Ca'Tron di Roncade is also supported by the Fondazione Cassamarca (Treviso, Italy).

Footnotes

  • Editor: Richard Staples

References

View Abstract