OUP user menu

Examination of Pseudomonas aeruginosa lasI regulation and 3-oxo-C12-homoserine lactone production using a heterologous Escherichia coli system

Matthew J. Wargo, Deborah A. Hogan
DOI: http://dx.doi.org/10.1111/j.1574-6968.2007.00773.x 38-44 First published online: 1 August 2007


In Pseudomonas aeruginosa, the signaling molecule 3-oxo-C12-homoserine lactone (3OC12HSL) is synthesized by LasI, and lasI transcription is positively regulated by LasR. A heterologous model has been generated for the study of the LasRI/3OC12HSL regulatory network in Escherichia coli. Escherichia coli pAHL-BAC cultures produced LasI-synthesized acylhomoserine lactones (AHLs) at levels and with kinetics similar to what is observed in cultures of P. aeruginosa strain PAO1. Analysis of the lasI transcript also showed similar induction profiles in both the E. coli pAHL-BAC strain and P. aeruginosa. Transposon mutagenesis of pAHL-BAC confirmed that transcriptional regulation by LasR is necessary for 3OC12HSL production, and showed that artificially increasing lasI transcript levels leads to higher levels of 3OC12HSL. Previous studies have shown that P. aeruginosa 3OC12HSL inhibits hypha formation, but not growth, in Candida albicans, and the E. coli pAHL-BAC similarly inhibited filamentation when grown in coculture with the fungus. It is proposed that this system will be useful for the study of factors that impact lasI regulation and 3OC12HSL production, and for the examination of the role of LasI-produced AHLs in bacterial interactions with other organisms.

  • quorum sensing
  • acylhomoserine lactones
  • bacterial–fungal interactions


Quorum-sensing signals are used by microorganisms to regulate genes in a population density-dependent manner (reviewed in Fuqua et al., 1996; Visick & Fuqua, 2005; Waters & Bassler, 2005). In Pseudomonas aeruginosa, the 3-oxo-C12-homoserine lactone (3OC12HSL) autoinducer molecule is produced by the LasI acylhomoserine lactone (AHL) synthase. Transcription of the lasI gene is induced by the LasR regulator when bound to its ligand, 3OC12HSL, thereby creating a positive feedback loop. In addition to controlling lasI transcription, the LasR/3OC12HSL complex controls the transcription of a large number of P. aeruginosa genes, many of which have been implicated in virulence (Schuster et al., 2003; Wagner et al., 2003). The 3OC12HSL molecule has also been reported to play roles in cross-kingdom interactions with eukaryotes by altering eukaryotic gene expression in plants (Mathesius et al., 2003), causing inflammation in human cells (Smith et al., 2001, 2002), and altering fungal morphology (Hogan et al., 2004).

Transcription of the lasI gene is under control of LasR as well as additional P. aeruginosa transcription factors including RsaL (de Kievit et al., 1999) and VqsR (Juhas et al., 2004), among others (Juhas et al., 2005). These transcription factors, in addition to unidentified factors that regulate lasI transcription in response to a variety of environmental factors (Juhas et al., 2005), complicate the investigation of direct lasI regulation. Many studies of lasI transcriptional regulation have utilized lasI promoter fusions to reporter genes that were examined in Escherichia coli. These studies have identified many of the specific promoter elements that participate in the control of lasI transcription (Seed et al., 1995; Rampioni et al., 2006). However, they cannot evaluate the interplay of autoinducer production and synthase regulation on 3OC12HSL production. Here, a functional LasIR system in E. coli is presented that enables the study of LasI-synthesized AHL production kinetics in a heterologous system to study the regulators, compounds, and conditions that impact lasRI regulation and 3OC12HSL synthesis. This model can also be used to produce LasI-synthesized AHLs in the absence of other soluble P. aeruginosa virulence factors.

Materials and methods

Bacterial strains and culture conditions

Escherichia coli DH10B cells (Invitrogen) were grown in Luria–Bertani (LB) with the addition of 12.5 µg mL−1 chloramphenicol to maintain the pBAC constructs. Pseudomonas aeruginosa strain PAO1 was grown in LB without antibiotics. Cultures of both species were grown at 37°C with shaking. For all growth and kinetic assays described here, cells were grown overnight for 16–18 h, followed by 200-fold dilution to an OD600 nm of 0.05 into 5 mL of fresh LB.

BAC library construction and screening

Pseudomonas aeruginosa PA14 genomic DNA was isolated according to the Genomic DNA Wizard kit (Promega), partially digested with Sau3A, size selected for DNA >20 kb in size, and ligated into BamHI and CIP-treated pBeloBAC11 using T4 DNA ligase (Invitrogen). ElectroMax competent E. coli DH10B cells (Invitrogen) were transformed with dialyzed ligation reactions. White colonies on LB plates with chloramphenicol and X-gal (20 µg mL−1) were picked onto master plates, and later transferred by a 48-pin replicator to Agrobacterium tumefaciens bioassay plates.

The A. tumefaciens bioassay was conducted according to Shaw (1997). Briefly, A. tumefaciens NT1 was grown overnight in AB medium (Hwang et al., 1994) at 30°C, diluted 1/10 in fresh AB medium, and shaken at 30°C until an OD600 nm of 0.6. The A. tumefaciens cells were mixed with molten agar at 55°C to a final agar concentration of 0.75%, X-gal was added to 80 µg mL−1, poured into 100 mm Petri plates, and allowed to solidify. Escherichia coli clones containing P. aeruginosa DNA were transferred onto these plates and incubated at 30°C overnight. Blue color was visible after c. 12 h, and the color continued to develop for c. 18 h.

AHL extraction, thin-layer chromatography (TLC), and AHL kinetic assays

Total AHLs were recovered from 5 mL of culture via a 2 × 2.5 mL extraction with acidified ethyl acetate (0.001% glacial acetic acid). The extracts were dried down to c. 1 mL under nitrogen, and then evaporated to dryness in 1.5 mL microfuge tubes in a rotary evaporator. The residue left after drying was dissolved in 100 µL of acidified ethyl acetate and either used immediately or stored at −20°C until needed. TLC was used according to the methods of Shaw (1997). Determination of AHL identities was based on reported Rf values (Shaw et al., 1997) as well as standards run on the same TLC plate (data not shown).

For kinetic analyses, AHL samples were prepared in a different manner. At each time point, 50 µL of culture was removed from each tube and heat treated for 15 min at 65°C. These samples were frozen at −20°C for at least 2 h before being thawed at room temperature and applied to the bioassay plates. AHL concentrations were stable in the frozen samples for c. 2 weeks before noticeable degradation took place. This method produced similar results when compared with ethyl acetate extraction during the time points examined in this study. To allow for the comparison of 48 samples on a single A. tumefaciens bioassay plate, large format 20 × 20 cm plates covered with c. 60 mL of agar containing an A. tumefaciens AHL-reporter strain were poured. AHL-containing samples (1–3 µL) were applied with adequate spacing to resolve spot diameters. These plates were placed on top of saturated paper towels in plastic containers covered with plastic wrap. Color was allowed to develop overnight at 30°C. The relative AHL concentrations were determined by spotting serial dilutions from the 5.5 h PAO1 sample and plotting the resulting spot diameter against the dilution factor. The exponential curve fitting this relationship had an R2 value better than 0.94 for all experiments.

Samples for growth were conducted by taking 100 µL samples of culture and fixing with 1% formaldehyde to a final concentration of 0.2%. The OD600 nm of these samples were read in 96-well dishes compared with a blank but uncorrected for path length.

RNA extraction, reverse transcriptase-polymerase chain reaction (RT-PCR), and quantitative real-time PCR

RNA was collected from 0.4 mL of cells at each time point by first preserving with RNAProtect (Qiagen), followed by isolation and purification using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. cDNA was synthesized using Superscript III (Invitrogen) as per the manufacturer's instructions using 500 ng of total RNA as the template. PCR was conducted according to the Taq polymerase manufacturer (Invitrogen) with the addition of dimethyl sulfoxide to a total of 5%. Quantitative real-time RT-PCR (qRT-PCR) was conducted with SYBR Green and AmpliTaq Gold DNA polymerase according to the manufacturer's instructions (Applied Biosystems). Contaminating DNA was removed by on-column DNAse treatment (Qiagen), followed by treatment with the DNA-free kit (Ambion). The sequences for the lasI specific RT-PCR primers are lasI-F1 (5′-CACATCTGGGAACTCAGC-3′) and lasI-R1 (5′-ACGGATCATCATCTTCTCC-3′). Other primers used in this study were for P. aeruginosa clpP and rplU, as well as a universal 16S primer. Sequences for these primers are clpP-F1 (5′-TCTTTTATTCCGCACGTTCC-3′), clpP-R1 (5′-CAGGTTGGCCATGTAGTCCT-3′), rplU-F1 (5′-GCAGCACAAAGTCACCGAAGG-3′), rplU-R1 (5′-CCGTGGGAAACCACTTCAGC-3′), 16S-F1 (5′-GGTAGTCYAYGGMSTAAACG-3′), and 16S-R1 (5′-GACARCCATGCASCACCTG-3′) (Bach et al., 2002). The amplification conditions were 95°C for 10 min, followed by 40 cycles at 95°C for 30 s, 56°C for 30 s, and 72°C for 30 s using an Applied Biosystems 7500 instrument. Only one PCR product was obtained for all primers and all samples based on melting curve analysis.

pAHL-BAC transposon mutagenesis

Transposon mutagenesis of the AHL-BAC was performed with the Mariner transposon based on the methods of Rondon (2000). The pool of mutagenized plasmid was transformed back into E. coli strain DH10B and mutants were screened based on the protocols described above. Transposon insertion position was determined by direct sequencing of isolated BAC DNA using the transposon-specific primer (5′-CACCCAGCTTTCTTGTACAC-3′). Transposon strains 3E6 (DH 834) and 2A3 (DH 835) can be maintained on gentamycin (15 µg mL−1).

Coculture conditions

Wild-type Candida albicans SC5314 were grown overnight in yeast extract–peptone–dextrose (2% glucose) medium at 30°C, washed with fresh YNBNC medium [final concentrations: yeast nitrogen base (Difco 291940) 0.67% w/v, 2% glucose, 5 mM N-acetylglucosamine, 0.5% casamino acids], and mixed with E. coli cells. This mixture was spotted onto YNBNC plates and incubated at 37°C overnight.

Results and discussion

An E. coli BAC clone containing lasI and lasR produces LasI-derived AHLs typical of P. aeruginosa with a similar kinetic profile

To analyze the effect of P. aeruginosa factors on LasI-synthesized AHL production kinetics, a clone was analyzed containing a genomic DNA from P. aeruginosa that contains the lasR and lasI genes on a single copy BAC vector in E. coli. This clone was identified by a screen of a P. aeruginosa BAC library for clones that were positive for AHL production using an A. tumefaciens bioassay (Shaw et al., 1997). From a library of 864 random clones with an average size >20 kb, one strain was identified that produced AHLs. The ends of the insert were sequenced with M13 forward and reverse primers and restriction analysis was used to confirm the general internal sequence structure (data not shown). The AHL-producing clone spans a region from 2, 425, 958 to 2, 457, 012 in the PA14 genome. This region is 98.8% identical to the 1, 532, 524–1, 563, 590 region from PAO1. Within this region are the lasR and lasI genes that are encoded on the same strand of the chromosome and separated by 360 bp. Within this intergenic region is the rsaL gene, encoded on the opposite strand. The PA14 region from 300-bp upstream of the lasR ATG to the end of the lasI gene (1973 bp) is 99.7% identical to the PAO1 sequence, with six single base pair changes. One change is 215 bp upstream of the lasR start codon. The other five are single conservative base pair changes in the coding regions; three in lasR, one in rsaL, and one in lasI. This clone was referred to as pAHL-BAC.

LasI from P. aeruginosa can synthesize a number of AHLs with varying chain lengths (Shaw et al., 1997). To determine whether E. coli carrying the pAHL-BAC produced a comparable spectrum of AHLs when compared with P. aeruginosa, AHL profiles were determined by TLC analysis (Fig. 1a). The ratios of 3OC10HSL:3OC12HSL and 3OC8HSL:3OC12HSL were very similar between both organisms as measured by densitometry and diameter (Shaw et al., 1997). Over three independent experiments, the 3OC10HSL:3OC12HSL ratios were 1.54±0.23 and 1.77±0.35 for E. coli pAHL-BAC and PAO1, respectively. The 3OC8HSL:3OC12HSL ratios were 6.06±0.49 and 5.21±1.03, respectively. The differences in the ratios between strains were not significant (P>0.44) and the variability is presented here as SD. The ratios of 3OC10HSL:3OC12HSL and 3OC8HSL:3OC12HSL were not found to change over the course of growth (data not shown). While the ratios of LasI-synthesized AHLs with different chain lengths are different from those published previously (Gould et al., 2006), these differences are likely due, at least in part, to differences in detection methods (Shaw et al., 1997; Gould et al., 2006).

Figure 1

AHL production by Escherichia coli pAHL-BAC and Pseudomonas aeruginosa PAO1. (a) TLC analysis of total AHLs from cultures grown for 5.5 h in LB, followed by detection using the Agrobacterium tumefaciens reporter strain NT1 bioassay. Specific AHLs are marked based on Rf value comparison with Shaw (1997) and comparison with authentic standards (not shown). (b) Relative total AHL concentrations produced by AHL-BAC (○) and PAO1 (▪) over time. Data from independent experiments were combined by scaling all data to the peak in PAO1 AHL concentration. (c) Cell growth based on OD600 nm measurements for each strain. For both graphs, error bars represent SD based on at least four independent experiments.

To compare the kinetics of AHL production in E. coli and PAO1, the relative concentrations of AHLs were measured using the A. tumefaciens bioassay modified for the analysis of a larger number of samples. The E. coli pAHL-BAC and P. aeruginosa strain PAO1 cultures contained similar concentrations of total AHLs with similar kinetic profiles (Fig. 1b). The growth of the two strains was nearly identical under these conditions (Fig. 1c). Total LasI-derived AHL concentrations in PAO1 peaked at 6.4±0.82 h. The E. coli pAHL-BAC strain produced similar levels of AHLs peaking at 5.8±0.64 h. This difference in the timing was not significantly different (P=0.24). In separate experiments, it was determined that the AHL production kinetics in a P. aeruginosa rhlI mutant, which lacks the C4HSL synthase, were indistinguishable from those of the wild type, indicating that C4HSL does not contribute significantly to the signal detected using this assay (C. Cugini and D. Hogan, unpublished data). The finding that an E. coli strain containing the genomic DNA encoding lasR and lasI produces similar levels of AHLs compared with PAO1 suggests that P. aeruginosa synthesis of 3OC12HSL is not profoundly impacted by factors outside of the lasI-lasR genomic region under these conditions. Alternately, E. coli may possess homologous or distinct regulators that play the same role as regulators in P. aeruginosa.

While P. aeruginosa and E. coli pAHL-BAC have similar growth and AHL production rates while grown in LB, this is not the case under all medium conditions. In minimal medium with glucose and casamino acids as carbon sources, AHL production kinetics were different between these strains (data not shown). While differences in growth complicate a direct comparison of these differences, similar studies under various medium conditions may help to identify P. aeruginosa-specific factors that are important for AHL production in different environments.

Increases in lasI transcript levels are similar in P. aeruginosa and E. coli pAHL-BAC

To compare the production of LasI-derived AHLs with the transcription kinetics of the lasI gene encoding the 3OC12HSL synthase, qRT-PCR was used to assay lasI transcript levels at various time points in cultures of P. aeruginosa PAO1 and E. coli pAHL-BAC over the course of growth. While a number of transcripts (rplU and clpX for P. aeruginosa, and 16S rRNA gene for both species) were present at constant levels relative to total RNA throughout the exponential phase of growth, perhaps not surprisingly, none of these transcripts were unchanged relative to total RNA throughout all growth phases (data not shown). Because of the difficulties associated with identifying a control transcript useful for comparing transcript levels over the entire range of growth phases, the lasI transcript values are expressed relative to the amounts of RNA in the cDNA synthesis reaction. Each RT-PCR reaction was performed in triplicate and there were two independent biological replicates for each time point. Error bars represent SD. Consistent with the similarities in the production of LasI-derived AHLs, the E. coli pAHL-BAC showed a lasI induction profile that is similar to that of P. aeruginosa (Fig. 2). Both strains showed evidence of autoinduction in response to increasing levels of AHLs between 2 and 4 h, which corresponds to the exponential phase (see Fig. 1). There were some differences in both the timing of the peak of lasI induction and total fold induction of the lasI transcript. lasI transcript levels peaked at 3 h in PAO1, and then began to decline; in E. coli pAHL-BAC, the decrease in lasI transcript levels was delayed. The lasI transcript levels also showed a higher total fold increase in PAO1 (70.8-fold) than in E. coli pAHL-BAC (10.8-fold). There are a number of possibilities explaining the differences in timing and fold induction of lasI. The lasI promoter may be under the control of other regulators in P. aeruginosa, and not in E. coli, or there may be differences in transcript stability. While the differences in transcript levels do not translate into differences in total AHL levels under these conditions, the differences in lasI transcription may impact AHL production in different environments. These differences in lasI expression between P. aeruginosa and E. coli may be exploited to identify factors that alter the coupling of lasI transcript levels to final AHL concentrations.

Figure 2

Quantification of lasI transcript levels relative to total RNA (filled symbols) and relative AHL concentrations (open symbols). Error bars represent SD. The AHL profile for each strain is shown for the same experiment from which the RNA was isolated.

Transposon insertion analysis of pAHL-BAC validates the heterologous system as a model to understand lasRI regulation

A transposon insertion strategy were used to validate the roles of the lasI and lasR genes in the heterologous AHL-BAC system. As predicted, transposon insertions in the lasI coding sequence abolished AHL production (data not shown). Similarly, insertions into the lasR coding sequence reduced AHL levels below the limit of detection (data not shown). One of the E. coli pAHL-BAC transposon mutants (3E6) produced c. 4.1-fold more AHLs at 5.5 h (peak of production for all strains) when compared with the parent E. coli pAHL-BAC (Fig. 3a). The transposon in strain 3E6 inserted 18 bp upstream of the lasI start codon, which is proximal to the dominant transcriptional start site, but not the secondary start site (Seed et al., 1995). This transposon insertion separates the secondary transcriptional start site from the native −35 to −10 elements. It was predicted that the lasI transcript levels were increased in this mutant and this was confirmed by RT-PCR analysis of lasI in this strain (Fig. 3b). Induction of lasI transcription and greater AHL production also occurred when transposon inserted 160 bp upstream of the lasR start codon within the lasR promoter region (isolate 2A3, Fig. 3a and b). The increased expression of lasI or lasR, in this case due to transposon insertion events, led to higher overall AHL concentrations (Fuqua et al., 1996). These data indicate the potential use of E. coli pAHL-BAC system for the study of factors that alter the timing of lasI induction, thereby impacting the levels of LasI-synthesized AHLs produced.

Figure 3

(a) Relative AHL concentrations measured from cultures of Escherichia coli pAHL-BAC and two transposon insertion derivatives. Strain 2A3, which has an insertion upstream of lasR, and strain 3E6, which is an insertion upstream of lasI. (b) Quantification of lasI transcript levels at the peak of AHL concentration. Transcript levels are corrected for the 16S transcript.

While transposon mutants in the lasI and lasR promoter regions led to increased lasI transcription and higher AHL levels, these increases did not result in significant changes in the ratios of the AHL species (data not shown). Interestingly, the fold increase in lasI transcript did not lead to a 1 : 1 increase in AHL concentration. The relative ratio of total AHLs to lasI transcript for the pAHL-BAC was 4.2, while the ratios for 2A3 and 3E6 were 1.92 and 1.46, respectively. This suggests that while increases in lasI transcription resulted in increased AHL production, there was a diminishing return presumably due to insufficient supply of one or more substrates.

The pAHL-BAC system can be used to study eukaryotic responses to lasI products

The 3OC12HSL molecule has a variety of effects on eukaryotic cells including initiation of inflammation in mammals (Smith et al., 2001, 2002), alterations in gene expression in plants (Mathesius et al., 2003), and changes in fungal morphology (Hogan et al., 2004). Because pathogens like P. aeruginosa typically produce a number of virulence factors, the heterologous production of bacterial products in nonvirulent bacterial hosts has been used to study the effects of these factors in host–microorganism interaction systems (Ham et al., 1998; McClain & Cover, 2003). To validate the present pAHL-BAC system, its ability to alter C. albicans morphology, an alteration seen in coculture with P. aeruginosa that is known to be dependent on 3OC12HSL, was examined (Hogan et al., 2004).

When E. coli carrying only the empty BAC vector (pBAC) was mixed with C. albicans, before inoculation onto medium that induces hyphal growth of the fungus, C. albicans formed hyphae resulting in a wrinkled colony morphology (Fig. 4, left). When C. albicans was mixed with E. coli pAHL-BAC cells, hyphal induction was inhibited, resulting in smooth colony morphology (Fig. 4, right). Growth experiments confirmed that AHL production by E. coli does not negatively impact the growth of C. albicans (data not shown). These data indicate that the production of 3OC12HSL by the pAHL-BAC was sufficient to alter C. albicans morphology in a manner similar to P. aeruginosa (Hogan et al., 2004). Therefore, it is proposed that this model will be useful to study biological responses to 3OC12HSL, especially in systems where large quantities of AHL are required or in systems where direct delivery of the chemical compound in a solvent is difficult or results in toxic effects on the study organism.

Figure 4

The AHL-BAC strain is capable of altering fungal morphology. The empty vector (pBAC) results in rough colonies composed of hyphal Candida albicans cells. The pAHL-BAC results in smooth colonies composed of yeast form C. albicans cells.

There have been many reports analyzing the production of AHLs from luxI homologues in E. coli, most showing that much of the specificity of AHL production is due to the synthase itself. It was decided to couple the lasI to the corresponding inducer, lasR in its native genomic context. This arrangement preserves all promoter and cis control elements. This single-copy construct will likely facilitate the analysis of many aspects of lasI regulation and 3OC12HSL production. In addition, pAHL-BAC was subcloned to include only the lasR to lasI region (DH907). This construct shows a kinetic profile similar to the full-length construct and may be useful in situations where a large insert is a hindrance.

The pAHL-BAC system allows for examination of transcriptional regulation as well as regulation of transcription factor function (LasR) and enzyme activity (LasI). The ability to examine both transcriptional control and final enzyme activity makes the system useful for a number of studies. The pAHL-BAC clone could serve as a useful primary or secondary screening tool to identify compounds that directly target the LasR/LasI system. The system is also amenable to examining the role of LasI products in other model organisms that are incompatible with live P. aeruginosa. Finally, genes that regulate the las system, whether by transcriptional regulation, alteration of enzyme activity, or regulation at the level of AHL degradation, can be identified by using intracellular screens involving compatible plasmids.


This work was supported by the Pew Biomedical Scholars Program (D.A.H.) and the National Institutes of Health P20-RR018787 from the IDeA Program of the National Center for Research Resources (D.A.H.) and the Ruth Kirchstein NRSA institutional fellowship to the Department of Microbiology and Immunology, Dartmouth Medical School (T32 AI07519, supporting M.J.W.).


  • Editor: David Clarke


View Abstract