SprI/SprR Quorum Sensing System of Serratia proteamaculans 94.

In this study, we investigated the quorum sensing (QS) regulatory system of the psychrotrophic strain Serratia proteamaculans 94 isolated from spoiled refrigerated meat. The strain produced several N-acyl-L-homoserine-lactone (AHL) QS signal molecules, with N-(3-oxo-hexanoyl)-L-homoserine lactone and N-(3-hydroxy-hexanoyl)-L-homoserine lactone as two main types. The sprI and sprR genes encoding an AHL synthase and a receptor regulatory protein, respectively, were cloned and sequenced. Analysis of their nucleotide sequence showed that these genes were transcribed convergently and that their reading frames partly overlapped by 23 bp in the terminal regions. The genes were highly similar to the luxI/luxR-type QS genes of other Gram-negative bacteria. An spr-box (analog of the lux-box) was identified upstream of the sprR gene and found to be overlapped with the sequence of -10 sequence site in the promoter region of this gene. Inactivation of the sprI gene led to the absence of AHL synthesis, chitinolytic activity, and swimming motility; decrease of extracellular proteolytic activity; affected the cellular fatty acid composition; and reduced suppression of the fungal plant pathogen mycelium growth by volatile compounds emitted by strain S. proteamaculans 94. The data obtained demonstrated the important role of the QS system in the regulation of cellular processes in S. proteamaculans 94.

1. Introduction

In recent years, it has become evident that bacteria regulate various physiological processes depending on their population density. This specific type of regulation is called quorum sensing (QS). QS signaling systems include low-molecular-weight molecules of different chemical nature and regulatory proteins that interact with the signaling molecules. With an increase in the population density, the concentration of signaling molecules increases to a definite threshold value, which may lead to an abrupt activation of the transcription of sets of genes in the entire bacterial population. It has been shown that the QS systems play a key role in the regulation of metabolism in bacteria living in different ecological niches and belonging to different taxonomic groups. Transcriptome and proteome analyses have shown that QS systems function as global regulators of gene expression in bacteria. QS signaling molecules (autoinducers) participate in intercellular transmission of information between bacteria belonging to the same and different species, genera, and even families. Due to QS regulation, bacteria are able to coordinately control gene expression in the population [1-3].

The best studied QS systems are the LuxI/LuxR-type systems first detected in Vibrio fischeri that function in Gram-negative bacteria with AHLs as the signal molecules. The QS regulation in V. fischeri occurs with the participation of two major components, the AHL synthase LuxI (which catalyzes AHL synthesis) and the LuxR protein. The latter forms a complex with AHL that binds to the promoter of the lux operon to activate its transcription, which leads to luciferase production and light emission. Most proteins of the LuxR type in different bacteria function as transcriptional activators, e.g., R proteins of Pseudomonas, Agrobacterium, Burkholderia, and Rhizobium [1-4]. In other bacteria of the Enterobacteriaceae family (e.g., Pantoea, Erwinia, Serratia, and Yersinia), another type of QS regulation is observed. In these bacteria, the LuxR-type proteins act as negative regulators, repressing their transcription and the transcription of target genes. Interaction of AHL with the R protein leads to conformational changes in the protein structure, which inhibit its binding to DNA and cause derepression of transcription [5-9].

The present work is devoted to study the QS regulation of Serratia proteamaculans 94 and its role in the control of cellular processes. This strain was isolated from meat spoiled in a refrigerator and identified by an analysis of nucleotide sequences of 16S rRNA genes (accession number EU327084.1) [10]. S. proteamaculans bacteria are adapted to various ecological niches in nature. They live in soil and in the rhizosphere of plants and are involved in spoilage of milk, dairy products, and meat [10-12]. Since QS systems play an important role in the adaptation of bacteria to different environmental conditions, including food-related species [13-15], one would expect the presence of actively functioning QS systems in S. proteamaculans 94. QS systems of this species are poorly studied. The S. proteamaculans B5a strain isolated from cold-smoked salmon was shown to cause QS-controlled spoilage of milk and to produce OHHL as the main AHL. The QS genes of the strain, sprI and sprR, were sequenced. According to proteome analysis, expression of many proteins in the B5a strain was affected by the QS system [11]. We recently investigated the functional role of the S. proteamaculans 94 gene luxS responsible for the synthesis of AI-2, a signaling molecule that participates in QS regulation in a large number of bacteria. It was shown that knockout of this gene led to an absence of AI-2 synthesis, chitinolytic activity, and swimming motility, suppression of the growth of the fungal plant pathogens Rhizoctonia solani and Helminthosporium sativum by volatile compounds emitted by S. proteamaculans 94 strain, and a decrease in extracellular proteolytic activity but did not affect synthesis of AHLs and lipolytic and hemolytic activities [16].

In this work, we have shown that S. proteamaculans 94 produces several types of AHLs, including OHHL and HHHL as the prevalent types. Genes of its QS system, namely, the sprI gene encoding an AHL synthase and the sprR gene encoding a receptor regulatory protein, were cloned and sequenced, and the organization of these genes was investigated. A S. proteamaculans 94 mutant with an inactivated AHL synthase (sprI) gene was constructed, and the role of QS in the regulation of several metabolic and physiological processes and properties in S. proteamaculans 94, including extracellular proteo- and chitinolytic activities, swimming motility, antifungal activity related to emission of volatile compounds, and cellular fatty acid composition, was demonstrated.

2. Materials and Methods

2.1. Organisms, Media, and Growth Conditions

The strains and plasmids used in the work are listed in Table 1. Bacterial strains were grown in Luria-Bertani broth (LB) or on solidified (1.5% w/v agar) Luria-Bertani (LA) agar (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) at 30°C. The fungi R. solani and H. sativum (from the collection of the Institute of Molecular Genetics, Moscow) were grown at 25°C on PDA (Difco) or Czapek agar medium (in g/l of distilled water): saccharose, 30; NaNO3, 2; KH2PO4, 1; MgSO4, 0.5; KCl, 0.5; FeSO4, 0.01; and agar-agar, 15.

Table 1

Strains and plasmids used in this study.

Strains or plasmids	Characteristics	Source or reference
Strains
Serratia proteamaculans 94	Isolated from spoiled refrigerated meat	Demidyuk et al. [10]
S. proteamaculans 94 rif-r	rif-r spontaneous mutant	Zaitseva et al. [16]
Escherichia coli S17-1 (λ-pir)	thi pro hsdR hsdM recA rpsL RP4-2 (Tcr::Mu) (Kmr::Tn7) (λ-pir)	Simon et al. [17]
E. coli TG-1	K-12 supE thi-1 Δ(lac-proAB) Δ(mcrB-hsdSM)5, (rK−mK−)	StrataGene
	F′ [tra D36 proAB+lacIqlac ZΔM15]
E. coli XL-1 blue	recA1 endA1 gyrA96 thi1 hsdR17 supE44 relA1 lac	Stratagene
Chromobacterium violaceum CV026	Violacein production-based AHL bioreporter, Km-r	McClean et al. [18]
Agrobacterium tumefaciens NT1/pZLR4	pZLR4 traG::lacZ β-galactosidase induction-based AHL bioreporter, Gmr Cbr	Shaw et al. [19]
S. proteamaculans 94 rif-r sprI::Gm	sprI::Gm gene replacement mutant of strain 94 rif-r	This study
Plasmids
p34S-Gm	Source of Gmr-cassette	Dennis and Zylstra [20]
pEX18Tc	Gene replacement vector, Tetr oriT sacB	Hoang et al. [21]
pAL-TA	Vector for cloning of PCR-products, Apr ori pUC Plac	Eurogen
pACYC184	Vector for cloning, ori p15 A, Tetr Camr	Sambrook and Russel [22]
pUCP26-splIR	splI and splR genes cloned in pUCP26-host-range vector, Tcr	Liu et al. [23]

Antibiotic supplements were used at the following concentrations: ampicillin (Ap, 200 μg/ml, “Biochemist,” Russia), kanamycin (Km, 100–200 μg/ml, “Synthesis,” Russia), gentamicin (Gm, 25 or 40 μg/ml, KRKA, Slovenia); rifampicin (Rif, 50 μg/ml), tetracycline (Tc, 20 μg/ml), and chloramphenicol (Cm, 20 μg/ml), from Sigma-Aldrich (USA).

2.2. Assay of AHL Production

An assay of AHLs was performed on plates with the violacein-producing reporter strain Chromobacterium violaceum CV026 [18] and the β-galactosidase-producing reporter strain Agrobacterium tumefaciens NTL4/pZLR4 [19].

2.3. Quantitative Determination of AHL Production by Liquid Chromatography and Mass Spectrometry (LC-MS/MS)

An overnight S. proteamaculans 94 culture was diluted to a starting OD600 of 0.15–0.25 (1 − 3 × 108 CFU·ml−1), and the cells were grown for 24 h. AHL was extracted by ethyl acetate (EtAc) as described [24]. In brief, cell-free culture supernatants (50 ml) prepared from the liquid cultures of the AHL-producing strains were acidified to pH 2.0 to prevent additional hydrolysis of AHLs and extracted twice with the same volume of EtAc; the extracted organic phases were then pooled. The solvent was evaporated under vacuum, and the resultant EtAc extracts were dissolved in 50 μl of HPLC-grade acetonitrile and stored at −20°C. An LC-MS/MS analysis was performed using an Agilent 120 Rapid Resolution LC system (Agilent Technologies Inc., Santa Clara, CA, USA) at a 0.25 ml/min flow rate. Mobile phase A (aqueous) was based on 0.05% acetic acid in H2O (MS grade), and mobile phase B was based on 0.05% acetic acid in acetonitrile (HPLC grade). The percentage of solvents A and B varied from 5 to 75 and 25 to 95%, respectively. The temperature of the HPLC column was maintained at 30°C, and the injection volume was 5 μl. HPLC separations were carried out using a Phenomenex Gemini C18 HPLC column (2 × 150 mm, particle size 3.5 μm). The liquid chromatography system was coupled with an Agilent 6410 triple quad mass selective detector equipped with an electrospray ionization ion source. The mass spectrometer was operated in the positive ionization mode at a capillary voltage of 4000 V, drying gas (nitrogen) temperature of 350°C, flow rate of 10 l/min, and nebulizer pressure of 35 psi, with nitrogen (99.999%) as the collision gas. The LC-MS system was controlled, and data were analyzed using MassHunter software (Agilent Technologies Inc.). Quantitative analysis of homoserine lactones was performed in the multiple reaction monitoring (MRM) mode by the method of standard addition. A set of synthetic AHLs with or without 3-oxo or 3-hydroxy substitutions and with acyl side chain lengths ranging from C4 to C12 at either 1 or 5 μM concentration was used as standards. AHLs were identified and confirmed by comparison of the elution time and MS spectra of their peaks and those of the standards.

For quantification, the mass spectrometer was configured for selected ion monitoring. The ions selected and the respective AHLs were as follows: m/z 172, BHL; m/z 200, HHL; m/z 214, OHHL; m/z 216, HHHL; m/z 228, OHL; m/z 226, HOHL; m/z 269, ODHL; and m/z 297, OdDHL. The concentrations of AHLs in EtAc extracts from supernatants of strain S. proteamaculans 94 applied to the column are presented in ng/ml and in μM. All AHL standards were purchased from Sigma-Aldrich Israel Ltd (Rehovot, Israel).

2.4. DNA Manipulations

Total genomic and plasmid DNA isolation, restriction enzyme digestion of DNA, ligation, agarose gel electrophoresis, PCR, Southern blot hybridization, electroporation, and transformation of E. coli cells with plasmid DNA were generally performed according to standard procedures [22]. Biotin-labeled specific probes were obtained with a Biotin DecaLabel DNA Labeling Kit (Fermentas, Life Sciences, Lithuania). Restriction enzymes and T4 DNA ligase were purchased from MBI Fermentas (Vilnius, Lithuania) and used according to the manufacturer's instructions.

PCR reactions were performed in a total volume of 20 μl containing a buffer for Taq DNA polymerase (“Sylex,” Russia), 200 μM concentrations of each of the four deoxynucleoside triphosphates, 10 рМ of each PCR primer (“Syntol,” Russia), and 1 U of Taq DNA polymerase (Institute of Molecular Genetics, Moscow, Russia). Plasmid DNA or boiled cells obtained from freshly grown colonies were used as templates for PCR. The primers used are listed in Table 2. DNA sequencing was performed with an Applied Biosystems 3730 DNA Analyzer using an ABI PRISM® BigDye™ Terminator v. 3.1 set. BLASTN and BLASTX homology searches for DNA sequences were performed using the NCBI server (http://www.ncbi.nlm.nih.gov/BLAST/).

Table 2

Primers used in this study.

Primer	Sequence (5′ ⟶ 3′)
deg SprI-F	5′-ATGCTTGAA(C,T)T(A,G)TTTGA(C,T)GT(C,T)AG
deg SprI-R	5'-(G,C)GGCCAGGT(A,G)ATAAC(T,G,C)GA
SprI-F	5′-ATGCTTGAATTATTTGACGTCAG+
SprI-R	5′-GCTGGAACTTATTACACCGG+
SprR-F	5′-GAGCCTGTATGTTTTCCATC+
SprR-R	5′- CAACTTCCGCCATCACCTG+
pACYC-F	5′-CAAATGTAGCACCTGAAGTC+
pACYC-R	5′- CGATGCGTCCGGCGTAGAG+
GM-F	5′-GGCTCAAGTATGGGCATCATT+
GM-R	5′-GGCGGTACTTGGGTCGATA+
М13-F	5′-GTAAAACGACGGCCAGT+
М13-R	5′-CAGGAAACAGCTATGAC+
TET-F	5′-CGAACGCCAGCAAGACGTAG
TET-R	5′-CTGCTCGCTTCGCTACTTGG

2.5. Cloning and Sequencing of the sprI and sprR Genes

Identification and cloning of the sprI and sprR genes were performed using PCR amplification with the degenerate primers deg SprI-F and deg SprI-R. The following program for PCR was used: 94°С for 3 min, followed by 30 cycles at 94°С for 20 s, 50°С for 40 s, and 72°С for 40 s. Boiled cells obtained from freshly grown colonies were used as templates for PCR. PCR products were isolated from gels using a Wizard SV Gel and PCR Clean-Up System (Promega) and were sequenced using the deg SprI-F and deg SprI-R primers. The obtained nucleotide sequences were compared with those in the GenBank database. A fragment that contained the complete sprI gene and a fragment of the sprR gene (83% identity to the respective gene sequences in S. proteamaculans B5a, AY040209.1) was obtained.

DNA fragments containing the whole sprI and sprR genes were determined by Southern blot hybridization with biotin-labeled specific probes and amplified by PCR. Chromosomal DNA was digested with BamHI and HindIII. A DNA fragment containing the sprI and sprR genes of approximately 1700–1750 bp was detected. A mixture of DNA fragments of 1500–2000 bp in size obtained after restriction digestion was ligated into pACYC184 vector DNA digested with BamHI and HindIII and then transformed into E. coli TG1. The cells were plated on the LA medium supplemented with chloramphenicol (20 μg/ml). Selected clones were tested for the presence of the sprI and sprR genes by PCR using the primers deg SprI-F and deg SprI-R (the sprR gene was located next to the sprI gene) and primers for the pACYC184 vector, pACYC-F, and pACYC-R (Table 2). DNA of the recombinant plasmid pACYC184-sprIR was isolated using a GeneJET Plasmid Miniprep Kit (Fermentas (Vilnius, Lithuania)). Sequencing of the cloned DNA sequence containing the sprI and sprR genes was performed using the primers pACYC-F and pACYC-R, which recognized the vector, and the deg SprI-F and deg SprI-R primers. The nucleotide sequence of the sprI and sprR genes was determined in a fragment of 1660 bp and deposited in GenBank (JX901285).

2.6. Gene Replacement Mutagenesis of the sprI Gene

For obtaining an sprI mutant, the sprI gene was cloned using PCR with specific primers to this gene: SprI-F and SprI-R (Table 2). PCR was conducted at 94°С for 2 min followed by 30 cycles at 94°С for 20 s, 58°С for 20 s, and 72°С for 40 s, and the last step was 72°С for 4 min. PCR products were isolated from gels using a Wizard SV Gel kit and PCR Clean-Up System (Promega, USA) and ligated to pAL-TA (Evrogen, Russia) vector DNA. The ligation mixtures were transformed into E. coli TG1. The cells were plated on the LA medium supplemented with 200 μg/ml ampicillin, and the presence of the sprI gene in selected clones was tested by PCR using the universal M13 primers. The plasmids obtained were named pAL-TA-sprI.

The sprI gene was inactivated by inserting a suicide vector into the S. proteamaculans 94 chromosome by homologous recombination using the sacB-based strategy [21, 25]. To inactivate the sprI gene, an 847-bp KpnI-fragment of the p34S-Gm plasmid bearing the gentamicin resistance gene [20] was cloned into the KpnI site of the sprI gene in the pAL-TA-sprI plasmid. Then, a 1.6-kb EcoRI fragment of the obtained plasmid pAL-TA-sprI::Gm was cloned into the EcoRI site of the pEX18Tc vector. The resulting plasmid pEX18Tc-sprI::Gm was used to transform the E. coli strain S17-1 (λ-pir). The transformed cells were in turn used as a donor for plasmid transfer into S. proteamaculans 94 by conjugation with selection for Gm-r, Rif-r, and Tc-r transconjugants (the S. proteamaculans 94 rif-r mutant used in this work was obtained previously [16], as a spontaneous mutant selected on the LA medium supplemented with 100 μg/ml rifampicin). The obtained clones were grown overnight in LB with Gm and Rif and then plated on the LA medium supplemented with Gm, Rif, and 5% (w/v) sucrose. The gentamicin-resistant, rifampicin-resistant, sucrose-resistant, and tetracycline-sensitive clones were tested by PCR with the Gm-F and Gm-R primers as well as with the SprI-F and SprI-R primers. As a result, the mutant strain S. proteamaculans sprI::Gm was obtained.

2.7. Assays of Proteolytic, Lipolytic, Chitinolytic, and Hemolytic Activities

For enzymatic activity assays, 3 μl of a 300 times diluted overnight culture of bacteria was plated on the surface of the solidified medium. The experiments for all enzyme assays were performed as described [26] and repeated in triplicate.

The activity of extracellular proteases was assayed after incubation at 30°C for 48 h on plates with the LA medium supplemented with 0.5% milk (1/3 vol.) by the appearance of casein degradation halos around bacterial colonies. To determine the lipase activity, strains were cultured for 4 days on LA plates supplemented with 0.01% CaCl2 and 1% Tween 20. Lipases hydrolyze Tween 20 to produce lauric acid that is converted into its insoluble calcium salt, which leads to an appearance of turbid zones around bacterial colonies. Enzyme activity was estimated by the radius of the turbid zones. The hemolytic activity of bacteria was determined on the LA medium supplemented with 5% sterile blood (human blood from Blood Transfusion Center, Moscow). Hemolytic activity was estimated by the radius of clear zones of hemolysis around bacterial colonies after growth for 6 days at 30°С. To determine the chitinolytic activity, cells were plated on the solidified (1.5% agar) medium containing 0.1% (NH4)2SO4, 0.03% MgSO4.7H2O, 0.08% KH2PO4, 0.04% KNO3, and 0.05% of yeast extract (Difco) supplemented with 0.2% colloid chitin. Bacteria were incubated for 96 h at 30°С. Chitinolytic activity was detected by the appearance of clear zones around bacterial colonies.

2.8. Action of Volatile Compounds Synthesized by S. proteamaculans 94 on Phytopathogenic Fungi

The antifungal activity experiments were performed in two-compartment Petri plates (92 × 16 mm) filled with the LA medium on the one side and solid Czapek's medium on the other side as described [27]. LA was seeded with the S. proteamaculans 94 strain or sprI mutant (20 μl of overnight culture, ∼3–5 × 107 CFU, was distributed on the surface of LA) and incubated for 2 h at 28°C. Then, a block of solid Czapek's medium (8 mm in diameter) covered with 5-day-old fungal mycelium was excised and placed onto the Czapek-filled part of the plate. All plates were 4-times sealed with «Parafilm M» Sealing Film (Pechiney Plastic Packaging Company, Chicago, IL, USA) and incubated for 6 days at 25°C.

2.9. Swimming Assay

The swimming motility of S. proteamaculans 94 cells was measured as described [28], with slight modifications. Overnight cultures grown in LB in tubes at 30°C with shaking were 300-fold diluted with LB and grown for 2.5 h. Then, 2 μl of the culture was plated on the solidified (0.3% Difco agar) M9 medium containing 0.4% glucose and 0.5% casamino acids (Difco), and the diameters of swimming zones were measured after 36 h of growth at 30°C.

2.10. Fatty Acid Assay

An overnight S. proteamaculans 94 culture was diluted to a starting OD600 of ∼0.10, and the cells were grown for 20 h at 30°C. Extraction of fatty acids from dried samples (5 mg) of the tested strains was carried out by acid methanolysis of the whole biomass in 0.4 ml of 1.2 N HCl in methanol by heating to 80°C for 1 h. The resulting fatty acid methyl esters were extracted twice with 0.2 ml of hexane and processed on a computer-assisted Microbial Identification System (MIS) (Microbial ID Inc., Newark, Del.) with a Hewlett Packard 5890A gas chromatograph and Hewlett Packard 7673A automatic sampler. The parameters of chromatography were chosen as recommended in the operational manual of the MIS [29, 30]. Gas chromatography-mass spectrometry (GC-MS) analysis was performed using an Agilent Technologies AT-5975 GC-MS system equipped with a cross-linked methyl silicone capillary column HP-5 ms. The oven temperature was 135°C for 2 min and then programmed up to 320°C at 7 degrees/min. Next, 1–2 μl of the derivatized sample was injected into the gas chromatograph at 280°C. Fatty acids and other lipid components were ionized by electron impact at 70 eV after separation in the GC column and analyzed in the scan mode. The quadrupole mass spectrometer had a resolution of 0.5 mass units over the whole mass range of 2–950 amu. The sensitivity of the GC-MS system was 0.01 ng of methyl stearate. Each substance was confirmed by its mass spectrum and a NIST mass spectral database library search.

3. Results

3.1. Synthesis of AHL Signaling Molecules by the S. proteamaculans 94 Strain

For detailed determination of AHL production, LC-MS/MS analysis was used. It was shown that strain 94 synthesized several types of AHLs, the amount of which significantly differed (Table 3). S. proteamaculans 94 cells synthesized and secreted OHHL and HHHL in the largest quantities, whereas the amounts of produced HHL, OHL, and BHL were much smaller. Trace amounts of ODHL, DDHL, and OOHL were also found in cell-free supernatants.

Table 3

Amount of AHLs detected in S. proteamaculans 94 culture supernatants by LC-MS/MS analysis.

AHL	BHL	HHL	OHHL	HHHL	HOHL	OHL	ODHL	OdDHL	DDHL	OOHL
ng/ml∗	12.5	41	>2500	>2000	nd	5.5	2	nd	2	Traces
μM∗∗	<0.1	0.2	>11.7	>9.3	nd	<0.1	<0.01	nd	<0.01	<0.001

nd, not detected. Concentration of AHLs in EtAc extracts, which were prepared from supernatants of strain S. proteamaculans 94 strain, was measured (mass and molar concentrations). Analysis was performed after 24 hours of bacterial growth.

3.2. Nucleotide Sequence and Organization of the sprI and sprR Genes in S. proteamaculans 94

The nucleotide sequences of the sprI and sprR genes were determined and annotated (GenBank, JX901285). The sprI and sprR genes are 633 and 750 bp in size, respectively. A comparison of the nucleotide sequences of the S. proteamaculans 94 genes sprI and sprR to those of the corresponding genes of other bacteria revealed high similarity of the QS genes in Serratia species: the S. proteamaculans strain B5a (83 and 88% identity of the sprI and sprR genes, respectively, AY040209.1), the Serratia plymuthica RVH1 strain (83 and 87% identity, AY394723.2), HRO-C48 strain (83 and 87% identity, AY841161.1), and G3 strain (83 and 87% identity, FJ919305.1); the Serratia liquefaciens HUMV-21 strain (79 and 85% identity, CP011303.1) and FDAARGOS_125 strain (79 and 85% identity, CP014017.1); and the Serratia marcescens H30 strain (77 and 81% identity, EU570248.1 and EU570249.1). High similarity of QS genes was also found with the corresponding genes of other genera of Enterobacteriaceae: the Yersinia enterocolitica LC20 (76 and 81% identity, CP007450.1); the Erwinia sp. EM595 (72 and 75%, LN907828.1); Pantoea ananatis strain LMG2665 (72 and 75%, KM249358.1); and Hafnia sp. CBA7124 (73 and 73%, AP017469.1).

The sprI and sprR genes of S. proteamaculans 94 partially overlapped at their 3′-end regions and were transcribed convergently from opposite DNA strands. The area of overlap is 23 bp long and rich in GC-bases (18 GC out of 23 bp), and it contains repeats (underlined) (Figure 1).

Figure 1

Organization of S. proteamaculans 94 QS genes and location of the Gmr cassette insertion in the sprI gene. The arrows indicate the direction of transcription. The bottom part of the figure shows the nucleotide sequence of the region of overlapping of sprI and sprR genes. The repeats presumable are underlined.

We have analyzed the sequence of a DNA fragment carrying the QS system genes sprI and sprR and adjacent regions with the help of promoter analysis programs (http://molbiol-tools.ca, PPP—Prokaryotic Promoter Prediction). As a result, we found the most likely sequences of −10 and −35 sites of the promoters of the sprI and sprR genes in S. proteamaculans 94 (Figure 2).

Figure 2

Promoter regions of the sprI (a) and sprR (b) genes. The sequences underlined are presumable −10 and −35 sites, and the initiation codon is indicated by an asterisk. The selected sequence is a presumable spr-box. The arrows indicate the direction of transcription.

A site similar to a lux-box of V. fischeri [31] was detected upstream of the sprR gene (an spr-box) (Figure 2(b)). The sequence of the spr-box contains a probable −10 site in the promoter of the SprR protein gene, binding of this protein to the spr-box could presumably inhibit initiation of transcription of the gene encoding the SprR protein.

3.3. Construction of a Mutant with Knocked-Out sprI Gene

To determine how the absence of synthesis of AHL signaling molecules affects the properties of S. proteamaculans 94 cells, we constructed a mutant bearing an inactivated sprI gene. Figure 1 shows the location of the Gm-r cassette insertion in this gene. Using the biosensors C. violaceum CV026 and A. tumefaciens NT1/pZLR4, we showed that the S. proteamaculans sprI::Gm mutant with the inactivated AHL synthase gene did not produce AHL.

To test whether AHL synthesis in sprI mutant cells would be restored by introduction of cloned QS genes into the cells, we used a pUCP26-splIR plasmid containing the QS system genes of the closely related bacterium S. plymuthica HRO-C48 for a complementation experiment [23]. As indicated above, the sprI and splI genes of the S. proteamaculans 94 and S. plymuthica HRO-C48 strains are highly similar (83 and 87% identity, respectively). The pUCP26-splIR plasmid was transferred by electroporation into S. proteamaculans 94 and the sprI::Gm mutant. The obtained transformants were selected for resistance to tetracycline (40 µg/ml). The presence of the pUCP26-splIR plasmid in transformants was tested by PCR using the TET-F/TET-R primers to show the presence of the tetracycline resistance gene of the vector plasmid and the SplI–F/SplI–R primers to confirm the presence of the splI/splR genes [23]. Introduction of the pUCP26-splIR plasmid into sprI::Gm mutant cells restored the synthesis of AHL (Table 4).

Table 4

Effect of mutation sprI:Gm and introduction of pUCP26-splIR plasmid into the mutant on the properties of cells.

Properties	S. proteamaculans 94	sprI:Gm	sprI:Gm/pUCP26-splIR
AHL production∗∗	+	−	+
Extracellular proteolytic activitya	5.1 ± 2.0	1.2 ± 0.0	4.8 ± 1.5
Chitinolytic activityb	4.6 ± 1.5	0	4.2 ± 1.0
Lipolytic activityc	5.4 ± 1.3	4.6 ± 1.5	4.1 ± 2.0
Hemolytic activityd	2.0 ± 1.0	1.6 ± 0.6	1.9 ± 0.4
Effect of VOCs on R. solani growthe	7.7 ± 1.5	19.6 ± 3.0	8.4 ± 1.0
Effect of VOCs on H. sativum growthe	8.6 ± 2.5	21.4 ± 2.0	7.5 ± 1.5
Swimmingf	33.0 ± 2.5	0	24.0 ± 2.0

The results of four independent experiments are expressed as the mean ± standard deviation (SD). Assay of AHL production was performed with the C. violaceum CV026 and A. tumefaciens NT1/pZLR4 biosensors. aRadius of the zone of casein hydrolysis, mm. bRadius of the zone of hydrolysis of chitin, mm. cRadius of the turbid zones around the colonies, mm. dRadius of the zone of hemolysis, mm. eGrowth of mycelium measured as distance in mm between the block of fungus and the border of its mycelium (action of S. proteamaculans 94 volatile compounds). In the absence of S. proteamaculans 94, the growth of fungal mycelium was 20–24 mm. fDiameter of the bacterial growth of the strains around a 2 μl drop of culture grown for 36 h at 30°C.

3.4. Effect of Inactivation of the sprI Gene on the Properties of S. proteamaculans 94

We studied the influence of inactivation of the sprI gene on the activities of some extracellular enzymes synthesized by strain 94. It was shown that the proteolytic activity in the mutant strain sprI::Gm was reduced (Table 4). The mutant strain lacked the chitinolytic activity of the parent strain (Table 4). The original and mutant strains grown on blood agar plates showed almost identical transparent small zones of hemolysis with radii of 1–2 mm; lipase activity (turbid zones around colonies of the parent and mutant strains) was similar (Table 4). Earlier, we showed that S. proteamaculans 94 produced volatile compounds, including VOCs (with dimethyl disulfide as the main VOC) that suppressed the growth of bacteria and fungi [27]. In this work, we found that inactivation of the sprI gene reduced or completely inhibited the growth suppression of two fungi, R. solani and H. sativum, induced by volatile substances (Table 4). We also tested the effect of the sprI gene inactivation on the ability of cells to migrate by swimming. It was shown that swimming zones were absent in the case of the sprI::Gm mutant (Table 4).

Introduction of the pUCP26-splIR plasmid into sprI::Gm mutant cells restored the characteristics of the parent strain for properties studied (Table 4).

3.5. Analysis of Fatty Acid Composition in S. proteamaculans 94 and the sprI::Gm Mutant

Saturated, unsaturated, and hydroxy acids were found in samples of S. proteamaculans 94 and the sprI::Gm mutant (Table 5). In total, 22 fatty acids with a chain of 12 to 20 carbon atoms were identified. The main saturated acids in S. proteamaculans 94 were hexadecanoic, octadecanoic and tetradecanoic acids, constituting approximately 37%, 11%, and 2.7% of the total amount of fatty acids in this strain, respectively. The dominant unsaturated acids were 9-octadecenoic, 11-octadecenoic, and 9-hexadecenoic (33%, 4.6%, and 3.4%, respectively). The unsaturation coefficient of fatty acids (the ratio of total unsaturated fatty acids to total saturated fatty acids) in the strain S. proteamaculans 94 was 0.83.

Table 5

Fatty acid composition in the strains S. proteamaculans 94 (A) and S. proteamaculans sprI::Gm (B).

No.	RT∗	Fatty acid	Amount of fatty acids in percent of the sum of areas of all chromatographic peaks
			A	B
1	6.983	Dodecanoic	0.5 ± 0.1	0.9 ± 0.1
2	9.495	9-Tetradecenoic	0	0.2 ± 0.1
3	9.578	11-Tetradecenoic	0	0.3 ± 0.1
4	9.873	Tetradecanoic	2.7 ± 0.3	4.9 ± 0.4
5	10.120	2-Hydroxy-dodecanoic	0	0.2 ± 0.1
6	11.305	Pentadecanoic	0.6 ± 0.1	0.6 ± 0.1
7	12.379	7-Hexadecenoic	0.4 ± 0.1	0
8	12.449	9-Hexadecenoic	3.4 ± 1.2	27.6 ± 2.4
9	12.797	Hexadecanoic	37.2 ± 1.7	34.6 ± 0.8
10	12.839	3-Hydroxy-tetradecanoic	0.6 ± 0.1	6.4 ± 0.5
11	13.570	Isoheptadecanoic	0.2 ± 0.1	0
12	13.694	Anteiso heptadecanoic	0.3 ± 0.1	0
13	13.764	Heptadecenoic	0.2 ± 0.0	0
14	13.888	Cyclopropane heptadecanoic	0.2 ± 0.1	5.5 ± 0.3
15	14.071	Heptadecanoic	0.8 ± 0.1	3.3 ± 0.2
16	15.144	9-Octadecenoic	33.2 ± 1.3	9.4 ± 2.1
17	15.186	11-Octadecenoic	4.6 ± 0.2	0.3 ± 0.1
18	15.427	Octadecanoic	11.0 ± 0.3	5.1 ± 0.2
19	16.170	Octadecadienoic, conjugated	1.3 ± 0.1	0.5 ± 0.1
20	17.556	9-Eicosenoic	1.5 ± 0.2	0.3 ± 0.0
21	17.597	11-Eicosenoic	0.9 ± 0.2	0
22	17.845	Eicosanoic	0.5 ± 0.1	0.2 ± 0.1
Coefficient of unsaturation of fatty acids			0.83	0.63

The numbers in the table are mean ± standard deviation (SD) of the analysis of the two independently grown cultures of each variant. RT: chromatographic retention time of fatty acids.

The sprI::Gm mutant of strain 94 was characterized by significant changes in the amount of fatty acids compared to that of the wild-type strain (for example, nos. 8, 10, 14, and 16) (Table 5). Fatty acids in the mutant strain were characterized by lower unsaturation compared with those in the parent S. proteamaculans 94 strain. The unsaturation coefficient of fatty acids of the sprI::Gm mutant was 0.63. The composition of fatty acids of the mutant strain differed from those of the wild-type strain by a sharply increased content of cyclopropane-heptadecanoic and 3-hydroxy-tetradecanoic acids (∼23- and 10-fold for the sprI::Gm mutant, respectively). Several fatty acids (nos. 7, 11, 12, 13, and 21) were absent in the mutant strain in contrast to S. proteamaculans 94. Three fatty acids (nos. 2, 3, and 5) absent in strain 94 were found in the sprI::Gm mutant strain in an amount less than 1%. Thus, the data above show that inactivation of the sprI gene in S. proteamaculans 94 significantly affected the composition of fatty acids.

4. Discussion

Intensive studies of QS regulation systems carried out in the recent years have shown that this type of regulation is widespread in bacteria of different taxonomic groups. AHL signals synthesized by different Serratia species are represented by mainly BHL, HHL, OHL, and OHHL and rarely also by N-hydroxy-acyl-L-homoserine lactones [8, 25, 32]. Our work is devoted to the study of the QS system in one of the representatives of the genus Serratia, namely, in the S. proteamaculans 94 strain isolated from spoiled refrigerated meat. QS systems in S. proteamaculans have been little studied. As far as we know, the only exception is the B5a strain, which produces OHHL as the main AHL [11]. A preliminary analysis of AHL production in S. proteamaculans 94 performed with the C. violaceum CV026 and A. tumefaciens NT1/pZLR4 biosensors demonstrated AHL synthesis in the S. proteamaculans 94 strain [16]. Quantitative mass spectrometry analysis showed that the S. proteamaculans strain synthesized five types of AHLs after 24 h of growth, among which OHHL and HHHL were predominant. The amounts of synthesized AHLs were on the same order as, for example, those in Pseudomonas aeruginosa. At least two concentration levels were observed in culture supernatants of this bacterium: AHLs at a relatively high content (5–15 μM) and those occurring at a lower content (<0.2 μM) [33].

Cloning and sequencing of the S. proteamaculans 94 AHL synthase sprI gene and the sprR gene encoding a regulatory receptor protein revealed that these genes are convergently transcribed and overlap in their terminal areas. Such QS system gene expression is different from that of the classical LuxI-LuxR QS model of V. fischeri, where the luxI and luxR genes are transcribed divergently [34]. The same difference was found in other members of the Enterobacteriaceae family, including species of Serratia, Erwinia, and Pantoea [5, 11, 35–37]. The area of overlap of the sprI and sprR genes is highly similar to the corresponding regions in other Serratia species: there is up to 100% identity with the area of overlap of these genes in S. proteamaculans B5a [11], S. plymuthica HRO-C48 [23], G3 [32], and RVH1 [38], and somewhat less similarity in other representatives of Serratia, for example, 70% sequence identity in S. marcescens [39]. However, the functional role of such overlapping QS genes is currently unclear.

A characteristic feature of the S. proteamaculans 94 QS system is the presence of the spr-box, which is similar to the lux-box of V. fischeri, in the promoter region of the sprR gene. A lux-box of V. fischeri is a 20 bp inverted repeat sequence of DNA that binds the QS regulatory receptor LuxR protein [31]. The spr-box in S. proteamaculans 94 overlaps with a tentative −10 site in the sprR gene promoter. Therefore, the binding of the SprR protein to the spr-box could apparently inhibit transcription of the sprR gene, preventing the binding of RNA polymerase. This assumption is in good agreement with available published data on transcription regulation of the esaR gene of P. stewartii, which encodes the EsaR protein, an analog of LuxR [5, 6]. The ExpR protein of E. chrysanthemi also represses transcription of its gene through interaction with the lux-box in the absence of AHLs, the transcription being derepressed in the presence of micromolar amounts of AHLs [37]. Lux-like-boxes were also found in other species of Serratia, for example, in S. marcescens SS-1 and S. plymuthica RVH1, upstream of the spnR and splR genes, respectively [37, 38].

The important role of AHL-dependent QS systems in regulation of cellular processes has been demonstrated for various species of Serratia. These QS systems participate in the control of the production of extracellular enzymes and some antibiotics, as well as in biofilm formation and motility [11, 23, 38, 40]. A proteomic analysis of the S. proteamaculans B5a strain showed QS-controlled expression of 39 cellular proteins, including the lipB-encoded secretion system and several lytic enzymes [11].

To study the role of QS in the regulation of cellular processes related to various aspects of metabolism in S. proteamaculans 94, an sprI mutant was constructed using gene-replacement mutagenesis. Inactivation of the sprI gene resulted in a decrease in extracellular proteolytic and chitinolytic activity but did not affect lipase and hemolytic activity. A mutation in the sprI gene also affected swimming motility of cells and antifungal activity related to the action of volatiles emitted by the parental strain. It is known that many strains of the genus Serratia have antifungal activity based on production of antibiotics and other antimicrobial substances, VOCs among them [8, 23, 25, 41, 42]. We showed previously that the volatiles of the strain S. proteamaculans 94 inhibited the mycelium growth of several phytopathogenic fungi [27]. Here, it was found that inactivation of the sprI gene reduced the inhibitory effect of S. proteamaculans 94 volatile compounds on the phytopathogenic fungi R. solani and H. sativum, suggesting that this gene is involved in the regulation of the synthesis of these substances in strain 94. Volatile compound activity may be involved in the competition of S. proteamaculans with other microorganisms in the rhizosphere and soil.

It is worth noticing that previously we have inactivated luxS gene, that allows the synthesis of another type of signaling molecules, AI-2, in strain 94, and investigated in this mutant almost the same phenotypes as those investigated here [16], with the exception of fatty acid synthesis. The comparison of these previously published data and our current data indicates that the deficiency in AHLs and AI-2 signaling molecules (or luxS gene inactivation) could lead to the very similar mutant phenotype. The results that are presented here may indicate that both AHLs and AI-2 systems in strain S. proteamaculans 94 could be involved (separately or in coordination) in QS control of the same physiological and metabolic processes.

The composition of cellular fatty acids is a stable genetic trait that is maintained in many generations of bacteria and used for their identification [43]. The ratio of saturated and unsaturated fatty acids affects physical properties of bacterial membranes and the degree of their fluidity and has a very significant effect on the activity of membrane-bound enzymes, permeability of membranes, and ability of cells to grow at low temperatures [44, 45]. Recently, it was shown that monounsaturated chain fatty acids and palmitoleic and myristoleic acids prevented QS-dependent biofilm development and drastically reduced motility in the nosocomial pathogen Acinetobacter baumannii [46]. The authors highlighted that these fatty acids decrease the expression of the regulator gene abaR belonging to the LuxIR-type QS communication system AbaIR and consequently reduce AHL synthesis. Essentially, this effect can be countered by the addition of exogenous AHLs. Moreover, oxygenated unsaturated fatty acids, known as oxylipins, produced by P. aeruginosa, function as autoinducers of a novel QS system [47]. The discovery of this oxylipin-dependent QS system reveals that prokaryote-derived oxylipins also mediate cell-to-cell communication in bacteria. Here, we have shown that the composition of fatty acids in S. proteamaculans 94 mutant cells with an inactivated QS gene sprI was significantly different from fatty acid composition of the original strain. This allows us to assume that the deficiency in AHL synthesis bidirectionally affects the synthesis of various fatty acids.

5. Conclusions

In this work, we have studied an AHL-mediated QS system in the psychrotrophic bacterium Serratia proteamaculans strain 94 isolated from spoiled meat. We have shown that this strain produced several types of AHLs, with OHHL and HHHL as the predominant signals. luxIR-like genes of the strain 94 QS system (sprI and sprR) were cloned and sequenced, and the organization of these genes was investigated. The S. proteamaculans 94 mutant with an inactivated sprI gene was found to be deficient in several physiological and metabolic traits, including AHL production, extracellular proteolytic and chitinolytic activities, swimming motility, and fungal mycelium growth suppression related to production of volatile compounds. All these phenotypes were restored to the parental strain level by complementation with similar wild-type luxIR genes cloned from another strain of Serratia. Moreover, inactivation of the sprI gene led to changes in the fatty acid composition detected in the wild-type strain, indicating involvement of QS regulation also in this basic metabolic process in strain 94. The obtained data provide new information about bacteria that are able to spoil various dairy, meat, or fish products, which may help to control food spoilage, for instance, by disrupting QS networks related with this undesirable activity of S. proteamaculans.