|Home | About | Journals | Submit | Contact Us | Français|
Quorum sensing (QS) is mediated by small molecules and involved in diverse cellular functions, such as virulence, biofilm formation, secondary metabolism, and cell differentiation. In this study, we developed a rapid and effective screening tool based on a cell-free Escherichia coli-based expression system to identify QS molecules of Streptomyces. The binding of QS molecules to γ-butyrolactone receptor ScbR was monitored by changes in the expression levels of the green fluorescent protein reporter in E. coli cell extract. Using this assay system, we could successfully confirm SCB1, a γ-butyrolactone molecule in Streptomyces coelicolor, binding to its known receptor, ScbR. In addition, we have shown that N-hexanoyl-dl-homoserine lactone, one of the QS molecules in many gram-negative bacteria, can regulate ScbR and trigger precocious antibiotic production in S. coelicolor. Our new method can be applied to other strains for which a screening tool for QS molecules has not yet been developed.
Quorum sensing (QS) is a mechanism for sensing population density, which allows cell-to-cell communication in bacteria (11, 40). Mediated by releasing and responding to small molecules, such as acyl homoserine lactones in gram-negative bacteria (10), small peptides (22), and butanolides in gram-positive bacteria (12), QS is involved in controlling a wide range of biological functions, including pathogenicity (41), biofilm formation (36), secondary metabolism (7) and cell differentiation (1). Screenings for QS molecules have drawn attention because QS molecules can be developed into drugs controlling cell-to-cell communication and potential antibacterial agents (32). Various QS molecules of gram-negative bacteria, such as Pseudomonas (35), Agrobacterium (25), and Vibrio (11), have been identified and characterized. Generally, to screen for novel QS molecules, biosensors capable of responding to a range of signal molecules are applied to culture supernatants. Based on this biosensor, the culture supernatant would be purified by several separation techniques, and QS molecules could be detected. In the next step, based on the structure of natural molecules, synthetic pools of QS analogues are prepared and screened for analogues (5, 6, 44). This type of screening was easily possible because some bacteria, such as Pseudomonas, Agrobacterium, Vibrio, and Staphylococcus, can grow quickly and allow tight and sensitive in vivo assays, which are not disturbed by their own cell color or complex metabolites (31).
In Streptomyces, γ-butyrolactones are well recognized as bacterial hormones and QS molecules working at concentrations of 10−6 to 10−9 molar (26, 38). They induce precocious antibiotic production and sporulation (38). After the structure (19, 42) and function (14, 39) of A-factor from Streptomyces griseus was elucidated, at least 17 kinds of similar molecules, such as SCB1 from Streptomyces coelicolor, factor I from Streptomyces viridochromogenes, IM-2 from Streptomyces lavendulae, and virginiae butanolides from Streptomyces virginiae, were identified (13, 17). In addition to γ-butyrolactones, several receptors of the butanolides, such as ArpA from S. griseus, ScbR from S. coelicolor, FarA from S. lavendulae, BarA from S. virginiae, and SpbR from Streptomyces pristinaespiralis, have been discovered (9, 21, 28, 33). In spite of increasing interest in γ-butyrolactones in Streptomyces, the complex regulatory pathways and poor compatibility of transcriptional and translational systems with gram-negative bacteria make it difficult to study Streptomyces QS (18). In addition, transporter systems in the cell wall are not well understood, and they appear to be different from other bacteria (2). Therefore, in vivo screening systems for QS molecules have not been well developed for Streptomyces.
To overcome these difficulties, we developed a novel assay system to detect Streptomyces coelicolor QS molecules based on cell-free Escherichia coli protein synthesis (27). Because cell-free translation systems using E. coli S30 can specifically synthesize proteins with speed and accuracy (15, 34), our method allows fast synthesis of green fluorescent protein (GFP) reporter proteins without cell growth in response to quorum signaling and can avoid many of the problems explained above. Contrary to a recently reported method with an in vitro system of Agrobacterium tumefaciens using its own cell extract (16), our system applied cell-free E. coli protein synthesis to screen QS molecules of Streptomyces. In this study, we showed successful applications of this novel system to detect interaction of various QS molecules to the γ-butyrolactone receptor of S. coelicolor.
Streptomyces coelicolor A3(2) M145 was obtained from the Korean Collection for Type Cultures (KCTC), and E. coli DH5α and BL21(DE3) were used as host strains. Streptomyces coelicolor A3(2) M145 was grown in R5− medium (18). Kanamycin, isopropyl 1-thio-β-d-galactopyranoside (IPTG), N-hexanoyl-dl-homoserine lactone (C6-HSL), N-octanoyl-dl-homoserine lactone (C8-HSL), N-3-oxo-octanoyl-homoserine lactone (3-oxo-C8-HSL), and N-z-l-homoserine lactone (HSL) were purchased from Sigma (Deisenhofen, Germany), and 2-hexanoyl-3-hydroxymethyl butanolide (VB-C) and 2-(6-methyl-1-hydroxyheptyl)-3-hydroxymethyl butanolide (SCB1) were purchased from Genechem (Daejun, South Korea).
cprA and cprB genes from S. coelicolor were amplified by PCR and cloned into pET24ma. ScbR was prepared as previously reported (43). His-tagged ScbR, CprA, and CprB proteins were grown with kanamycin (50 μg/ml) and expressed in E. coli BL21(DE3) by induction with 0.1 mM IPTG overnight and purified by Ni-nitrilotriacetic acid (NTA) affinity chromatography. To construct a fluorescent reporter plasmid, an enhanced GFP (eGFP) gene was amplified by PCR and cloned into a pET23b (Novagen) vector at NdeI/BamHI sites, and the binding site was amplified with primers (5′-CGTCGTTCTAGACCGGTGGACAAGCGCATC-3′/5′-CGTCGTTCTAGAGCCTGCCTCCTTGTTCAT-3′ [italic nucleotides indicate the restriction enzyme site]) and cloned into XbaI sites, upstream of the ribosome binding site. All the sequences also were confirmed by DNA sequencing.
Cell extract was prepared from E. coli strain BL21(DE3) (Novagen, Madison, WI) as described previously (20). Cells were grown at 37°C in 3 liters of 2× yeast extract-tryptone medium with vigorous agitation and aeration. When the cell density (optical density at 600 nm) reached 0.6, 1 mM of IPTG was added to the cell culture medium to express T7 RNA polymerase. The cells were harvested at an optical density at 600 nm of 4.5 and washed three times with 20 ml of buffer A, which contained 10 mM Tris-acetate buffer (pH 8.2), 14 mM magnesium acetate, 60 mM potassium glutamate, 1 mM dithiothreitol (DTT), and 0.05% (vol/vol) 2-mercaptoethanol, per gram of cells (wet weight). Before storing the pellets at −80°C, the wet cell pellets were weighed.
For the preparation of the standard S30 extract, the thawed cells (10 g) were suspended in 12.7 ml of buffer B (buffer A without 2-mercaptoethanol) and disrupted in a French press cell (Aminco) at a constant pressure of 20,000 lb/in2. The lysate was then centrifuged at 30,000 × g for 30 min at 4°C, and the top layer of the supernatant (lipid layer) and pellet was carefully removed and then centrifuged again. Three milliliters of the preincubation solution (293.3 mM Tris-acetate [pH 8.2], 2 mM magnesium acetate, 10.4 mM ATP, 200 mM creatine phosphate, 4.4 mM DTT, 0.04 mM amino acids, 26.7 μg/ml creatine kinase) was added gradually to 10 ml of the final supernatant while shaking at 100 rpm, and then incubated at 37°C for 80 min. The preincubated cell lysate with the preincubation mixture was dialyzed four times for 45 min each at 4°C against 50 volumes of buffer B, using a Pierce membrane (SnakeSkin pleated dialysis tubing; Rockford) with a molecular weight cutoff of 10,000. The retained extract was centrifuged at 4,000 × g for 10 min at 4°C to obtain the supernatant (the S30 extract).
The standard reaction mixture used for cell-free protein synthesis consists of the following components in a total volume of 30 μl: 57 mM HEPES-KOH (pH 8.2); 1.2 mM ATP; 0.85 mM (each) CTP, GTP, and UTP; 2 mM DTT; 0.17 mg/ml E. coli total tRNA mixture (from strain MRE600; Roche, Germany); 0.64 mM cyclic AMP; 90 mM potassium glutamate; 80 mM ammonium acetate; 12 mM magnesium acetate; 34 μg/ml l-5-formyl-5,6,7,8-tetrahydrofolic acid (folinic acid); 1.5 mM (each) 20 amino acids; 2% polyethylene glycol 8000; 67 mM creatine phosphate; 3.2 μg/ml creatine kinase; 10 μM l-leucine; 10 μg/ml DNA; 24% (vol/vol) of S30 extract.
The reaction mixtures were incubated with different amounts and kinds of QS receptors and QS molecules for 3 h. After 3 h, the reaction mixtures were diluted to 100 μl with deionized water and then transferred to 50-μl cuvettes. Fluorescence was measured by a fluorospectrometer (Jasco, Tokyo, Japan) with excitation and emission wavelengths of 496 nm and 516 nm (slit excitation size, 10 nm; slit emission size, 10 nm), respectively (4).
For the preparation of S. coelicolor cells, 0.5 μl of spore stock solution (5.4 × 105 colonies/μl) was cultured in 20 ml of R5− medium in a 250-ml baffled flask. To observe the effect of quorum molecules on cell growth and antibiotic production, 500 μl of cultivated solution was inoculated into 3 ml of fresh R5− medium for different concentrations and kinds of quorum molecules. To measure the amount of antibiotics such as actinorhodin (ACT) and undecylprodigiosin (RED), 1 ml of each sample was taken at specified times during cell growth. After centrifugation in a microcentrifuge for 10 min, 500 μl of the supernatant was sampled for further analysis following a previously described method (3, 18). Using 200 μl of treated samples in 96-well plates, the titers of ACT and RED were measured spectrophotometrically at 633 nm and 530 nm, respectively, with a multiscanner (Thermo Electron Corp., Finland).
The same PCR product of the scbR/A promoter used to construct the sensor was radiolabeled with T4 polynucleotide kinase in the presence of [γ-32P]dATP and separated from unreacted DNA by using ProbeQuant G-50 Micro columns (Amersham Biosciences). The labeled probes were incubated with purified ScbR in 20 μl reaction buffer containing 20 mM HEPES (pH 7.8), 10% (wt/vol) glycerol, 100 mM KCl, 0.05 mM EDTA, 5 mM MgCl2, 0.5 mM DTT, 0.01% Nonidet P-40, and 2 μg sheared salmon sperm DNA for 10 min on ice and for 30 min at 37°C. Protein-bound DNA and free DNA were resolved on 5% acrylamide gels in 0.5× Tris-borate EDTA buffer at room temperature. The gels were dried and the radioactive signals were analyzed by a Typhoon 8600 scanner (Molecular Dynamics).
In S. coelicolor, ScbR is a γ-butyrolactone SCB1 receptor regulating transcription by binding to the divergent promoter region between scbR and scbA. ScbA is involved in γ-butyrolactone synthesis. Without QS molecules, ScbR binds to the promoter, repressing the expression of ScbR. Binding of SCB1 releases ScbR from the promoter, preventing repression of the transcription of scbR (39). In order to screen QS molecules regulating ScbR, a novel assay system based on cell-free protein synthesis was developed. A reporter plasmid was generated containing the T7 promoter, the scbR/A promoter site where ScbR binds, a ribosome binding site, and the egfp gene encoding extended GFP in a consecutive manner. In addition, His-tagged ScbR was prepared as previously reported (29). Binding of ScbR to the scbR/A promoter will inhibit the transcription from the T7 promoter, inhibiting expression of the egfp gene in cell-free protein synthesis system. Fluorescence will eventually be less than in the control (Fig. 1A and B). However, when QS molecules are present, QS molecules will bind to ScbR, and ScbR dissociates from the promoter, leading to an increase in egfp expression (Fig. (Fig.1C1C).
To prove the feasibility of the new QS assay system, this assay system was compared with the gel mobility shift assay, which has been used previously to detect binding of ScbR to the scbR/A promoter (39). Zero to five picomoles of purified protein and 10 ng of reporter DNA were used for the new assay, and 0.5, 2.5, and 5 pmol protein and 1 ng of DNA were used for the gel mobility shift assay. The increase in the amount of ScbR protein reduced GFP fluorescence intensity (Fig. (Fig.2B),2B), reflecting the increased binding of ScbR to the promoter region detected by the gel mobility shift assay (Fig. (Fig.2A).2A). The control reaction using bovine serum albumin (BSA) showed negligible results in both fluorescence intensity and DNA binding compared to the dramatic changes exerted by ScbR. In the presence of increasing amounts of SCB1, ScbR was released from DNA, as expected (Fig. (Fig.2C).2C). In agreement with the changes in ScbR-DNA binding activity, SCB1 lessened the ScbR-mediated repression of GFP, showing an increase in fluorescence (Fig. (Fig.2C2C and andD).D). In contrast, the BSA control had a decrease in the fluorescence intensities proportional to the concentrations of SCB1. The nice correlation of the new assay system with gel mobility shift assay suggests that this cell-free assay system can work as a quorum sensor at proper concentrations.
To find other possible QS molecules, C6-HSL, C8-HSL, 3-oxo-C8-HSL, and VB-C were examined in comparison with SCB1. Additionally, HSL was also examined. To decrease basal fluorescence caused by incomplete binding to target DNA, a concentration of ScbR protein almost 1,000 times higher (5 nmol) was used, compared with the previous feasibility study. Because quorum molecules are not soluble in pure water, 50% methanol was added to the solution to make a 1 mM solution and then diluted serially, which seemed to affect the activity of ScbR protein. Using a high concentration of protein solved this problem and gave consistent results.
As expected, SCB1 showed the least repression of eGFP expression (Fig. (Fig.3.).3.). VB-C, which is a QS molecule of other Streptomyces species, such as Streptomyces virginiae (19), also had an increase in fluorescence. Most interestingly, C6-HSL also had high fluorescence values, indicating its binding to the ScbR receptor. C6-HSL is a QS molecule of gram-negative bacteria such as Pseudomonas aeruginosa, Serratia marcescens, and Aeromonas salmonicida (8, 37, 41).
To find out the effect of C6-HSL on antibiotic production of S. coelicolor, the cells were cultivated in the presence of C6-HSL, and the timing of antibiotic production was monitored. S. coelicolor produces at least four distinctive antibiotics, such as ACT (blue-pigmented aromatic polyketide), RED (red-pigmented tripyrolle), calcium-dependent antibiotic (nonribosomal peptide antibiotics), and methylenomycin (cyclopentanone antibiotics), and among them, some studies of RED and ACT have been performed with γ-butyrolactone (30, 38, 39, 43). The QS molecule did not seem to dramatically increase or decrease the amount of antibiotic production itself, but it affected the initiation of secondary metabolite production (43). As shown in Fig. Fig.4,4, nearly 100 times more C6-HSL than SCB1 was needed to trigger the precocious antibiotic production effect at 48 h. Although it did not exactly show a dose-dependent manner, it surely triggered precocious antibiotic production repeatedly, compared to the wild-type strain. As reported, a higher concentration of SCB1 showed an inhibitory effect on antibiotic production (39).
In S. coelicolor, there are several homologues of the QS regulator, including CprA and CprB (29). They showed a high sequence homology in the N-terminal region, which is known as a DNA binding region (38). From the results with the ArpA binding site from Streptomyces griseus (29), binding and regulation of CprA and CprB on the scbR/A binding site are highly expected. CprA and CprB share 92% amino acid identity, but their functions are totally different. CprA is known as an activator; however, CprB is known as a repressor for antibiotic production (29). The change in the fluorescence profile showed that CprA worked as a weak activator, especially at low concentrations, but CprB worked as a repressor, like ScbR (Fig. (Fig.5).5). In addition to in vivo studies (data not shown), these results agree with the previously suggested roles for CprA and CprB in S. coelicolor (29).
The purpose of this study was to develop a quorum screening system for strains of species such as S. coelicolor, which have complex regulatory pathways, produce various metabolites, and show different phenotypes depending on external factors such as nutrients and stresses. So far, in vivo quorum detection systems expressing Agrobacterium tumefaciens by using TraR, Sinorhizobium meliloti SinI, and Pseudomonas aeruginosa LasR in their own strain or E. coli have been used to characterize many QS molecules (24, 35, 44). However, the in vivo quorum detection system has limitations for general application to other bacteria, such as Streptomyces.
Here, we have developed a new cell-free E. coli-based assay system to detect QS molecules and applied this system to confirm SCB1 binding to its known receptor, ScbR. In addition, we demonstrated that C6-HSL, which is one of many QS molecules in gram-negative bacteria (23, 41), can also regulate ScbR and trigger precocious antibiotic production in S. coelicolor. This seemed to be the first evidence showing the possible cross talk between Streptomyces and gram-negative bacteria (40), and C6-HSL might function in cross communication between gram-negative bacteria and Streptomyces (31). The application to CprA shows that this system can be applied to decide whether a regulator is an activator or a repressor, depending on the amount of protein, suggesting that this system allows functional characterization of regulators in addition to the investigation of regulators in terms of DNA binding. The binding experiment of CprA and CprB with several different molecules with this system might be useful to find other binding molecules.
This cell-free sensor can be a good alternative method to screen QS. This system is free from the consideration of cell wall penetration and possible effects of other cellular factors, and it can be used for high-throughput screening because it can be done within 4 hours. Without the use of radioisotopes and acrylamide gels, the interaction of protein and ligand molecules can easily be monitored. This method can be adapted to any other QS system once we have the information about the DNA binding protein and its specific target promoter. Therefore, our new assay system would serve as a versatile and competent tool to study QS.
We acknowledge the Engineering Research Institute, Seoul National University.
This work was partially supported by a grant funded by the Korean government (MOST) (no. R0A-2007-000-10007-0) and a grant (code no. KRF-2005-005-J16002) from Korea Research Foundation, Republic of Korea.
Published ahead of print on 14 August 2009.