Previous Article | Next Article ![]()
Applied and Environmental Microbiology, April 2005, p. 2113-2120, Vol. 71, No. 4
0099-2240/05/$08.00+0 doi:10.1128/AEM.71.4.2113-2120.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Center for Biomedical Microbiology,2 Center for Microbial Biotechnology, BioCentrum, Technical University of Denmark,3 Department of Seafood Research, Danish Institute for Fisheries Research, Kongens Lyngby, Denmark1
Received 13 June 2004/ Accepted 29 October 2004
|
|
|---|
|
|
|---|
P. phosphoreum also plays an important role in food microbiology, where it acts as the most important spoilage bacterium of packed, chilled fish fillets (11, 12, 52). Its ability to use trimethylamine oxide (TMAO) as a terminal electron acceptor facilitates its growth in packed fish products, where it reduces this compound to trimethylamine (TMA), imparting to the product a spoiled, fishy flavor. We are currently investigating the potential involvement of quorum sensing in food products in which spoilage is caused by growth of and metabolism by gram-negative bacteria.
It has recently been demonstrated that quorum-sensing signals, AHLs, can be extracted from several food products, including fish (smoked and fresh), vegetables, and meats (4, 26, 27), and are produced by several gram-negative food spoilage bacteria (4, 7, 8, 23, 26, 44, 45). In some food products, AHL production does not influence the decay in food quality (4, 8), yet in other cases, the production of degradative enzymes, which have a negative influence on the quality of milk and vegetable products, can be AHL regulated (7, 8, 44).
The present study investigates the production of AHLs in packed fish and identifies the AHL-producing bacteria. The nature of the AHL found in the fish and the AHL produced by P. phosphoreum is elucidated, and the AHL and light production kinetics in P. phosphoreum are investigated.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Strains and plasmids used in the study
|
Enumeration of bacteria from cod samples.
At regular intervals during storage, samples of cod fillets were withdrawn and aerobic plate counts were determined on L&H plates incubated at 15°C. Luminescent bacteria were counted on L&H plates after 2 to 3 days of incubation, and aerobic colonies on L&H plates were counted after 5 to 7 days. H2S-producing organisms were enumerated as black colonies in pour-plated IA incubated at 25°C for 3 to 5 days. A P. phosphoreum-specific count was obtained by using a conductometry-based method (13).
The level of AHL-producing bacteria was estimated by using a replica-plating procedure in which the L&H plates were replicated onto indicative agar plates containing A. tumefaciens NT1 (pZLR4) imbedded in ABT agar supplied with 50 µg of 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (X-Gal)/ml or C. violaceum CV026 embedded in LB5 medium (4, 27). The replica plates were incubated at 20°C for 1 to 5 days, and colonies causing a blue color (cleavages of X-Gal by A. tumefaciens NT1 [pZLR4]) or a purple pigment (produced by C. violaceum CV026) in the surrounding agar were interpreted as being AHL positive.
Identification of bacteria.
Strains isolated from the replica plating procedure as potential AHL producers were identified by standard biochemical tests. Strains were grown in a liquid medium, and shape and motility were determined by phase-contrast microscopy (x1,000 magnification). Colonies from agar plates (24 to 48 h) were tested for Gram reactions (with 3% KOH), catalase reactions (3% H2O2), oxidase reactions (BBL Dryslide Oxidase slides; Becton and Company, Sparks, Md.), the ability to ferment and oxidize glucose (Merck 10282 supplemented with a total of 1% NaCl), the ability to grow in the presence of vibriostaticum (2,4-diamino-6,7-di-isopropylteridine, 0/129 DD15, 150 µg; Oxoid 129150) and bioluminescence when grown aerobically on L&H plates with 1% NaCl at 15°C. Vibriostaticum-sensitive isolates were further tested for metabolism of arginine, lysine, and ornithine; growth without NaCl; growth at 0°C and 35°C; assimilation of D-mannitol; production of TMA from TMAO; and production of gas from glucose. All tests were carried out as previously described (2, 11, 14). The phenotypic tests selected for identification as P. phosphoreum are a subset of tests from several studies where numerous phenotypic characteristics and sequencing of 16S rDNA and luxA were used to identify type strains and environmental isolates of the photobacteria (14, 17). To verify the scheme, 16S rDNA sequences (1,333 to 1,497 bp) were determined for one luminous P. phosphoreum isolate (FR7) and for three nonluminous P. phosphoreum isolates (NCIMB 13478, NCIMB 13479, and P50) previously isolated from packed cod and identified by the selected phenotypic tests (14, 17). 16S rDNA sequences were determined by the BCCM/LMG Bacteria Collection, Laboratorium voor Microbiologie Univerrsitet Gent, Ghent, Belgium. The four 16S rDNA sequences have been deposited in the EMBL, GenBank, and DDBJ nucleotide databases. Gram-negative and catalase- and oxidase-positive mobile rods resistant to vibriostaticum were tentatively identified as Aeromonas, and these isolates all produced gas from glucose. Strains were identified as Shewanella spp., being gram-negative mobile rods positive for catalase and oxidase but unable to ferment or oxidize glucose and having the ability to form TMA from TMAO and to produce H2S on IA. These strains usually formed pink- to orange-pigmented colonies on IA (52).
Extraction of AHLs from cod fillets.
AHLs were extracted from approximately 10 g of the homogenized cod fillet by using three times as much acidified (0.5% formic acid) ethyl acetate (approximately 30 ml) as fish sample. The extracts were filtered through a Whatmann 4 filter and evaporated under a stream of nitrogen, dissolved in 1 ml of acidified ethyl acetate, and stored at 20°C.
Identification by TLC and quantification of AHLs.
AHL production in bacteria isolated as presumed AHL producers based on the replica plating procedure was verified in a well diffusion assay. Indicative plates were prepared as for the replica plating procedure, by using both A. tumefaciens NT1 (pZLR4) and C. violaceum CV026. Wells were cut into the agar in which 60 µl of sterile-filtered supernatant from an outgrown culture (24 to 48 h as described above) of the presumed AHL producer or 60 µl of ethyl acetate extract was added (45). AHLs were tentatively identified directly from sterile supernatants derived from outgrown bacterial cultures or from ethyl acetate extractions of the culture supernatants or cod fillets by separating them by thin-layer chromatography (TLC) and subsequently developing the TLC plates by using an indicative agar containing A. tumefaciens NT1 (pZLR4) (45, 49). A semiquantification of AHLs was performed by preparing a well diffusion assay based on A. tumefaciens NT1 (pZLR4). Increasing concentrations of N-(3-hydroxyoctanoyl)homoserine lactone (OH-C8-HL; provided by John Nielsen, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark) was added to the wells, and a standard curve correlating the size of the induced halo surrounding the well to the amount of AHL was constructed and used for estimating the concentration of AHL in a sample (23, 45).
Confirmation of AHLs by HPLC-ESI+-high-resolution mass spectrometry (HRMS).
Bacterial ethyl acetate extracts were evaporated to dryness in a vacuum, redissolved in 10 times less methanol-water (1:1), and analyzed on two different columns; 1 µl was injected in an Agilent 1100 high-pressure liquid chromatograph (HPLC) equipped with either a Phenomenex (Torrance, Calif.) Luna II C18, 3-µm particle, 50- by 2-mm column and a Phenomenex 2-mm C18 SecurityGuard precolumn or a Phenomenex Polar RP Phenyl, 3-µm particle, 150- by 2-mm column and a Phenomenex 2-mm Polar RP SecurityGuard precolumn. The HPLC was coupled to an LCT orthogonal time-of-flight mass spectrometer (MS; Micromass, Manchester, United Kingdom), and equipped with a Z-spray electrospray ionization (ESI) source and a LockSpray probe (using leucine encephalin as lock mass) (41). The MS was operated in the positive-ion ESI (ESI+) mode and tuned to a resolution of >6,500 (at half-peak height), and data were collected as centroid data from m/z 100 to 900 as previously described (41). Analyses were performed with water (MilliQ) containing 10 mM ammonium formate and 20 mM formic acid (both analytical grade) and acetonitrile (gradient grade) containing 20 mM formic acid. A flow rate of 0.3 ml/min was used, starting with 5% acetonitrile (Luna C18 column) or 0% acetonitrile (Polar RP column), going to 100% in 25 min, holding for 3 min, returning to 5% acetonitrile in 4 min, and then equilibrating by 7 min. To verify that no AHLs could be detected in a luminescent photobacterium (FR7), the following reference standards were coanalyzed on both columns: OH-C8-HL, N-butanoyl-homoserine lactone (C4-HL), N-hexanoyl-homoserine lactone (C6-HL), N-octanoyl-homoserine lactone (C8-HL), N-(3-oxohexanoyl)homoserine lactone (O-C6), N-(3-oxooctanoyl)homoserine lactone (O-C8), and N-(3-oxodecanoyl)homoserine lactone (O-C10). Besides checking for the extracts for the [M+H]+ and [M+Na]+ (±0.01 m/z) adducts of these reference standards, we also inspected the chromatograms for peaks resembling same adducts of N-decanoyl-homoserine lactone (C10-HL), N-dodecanoyl-homoserine lactone (C12-HL), N-tetradecanoyl-homoserine lactone (C14-HL), N-(3-oxododecanoyl)homoserine lactone (O-C12-HL), N-(3-oxotetradecanoyl)homoserine lactone (O-C14-HL), N-(3-hydroxyhexanoyl)homoserine lactone (OH-C6-HL), N-(3-hydroxydecanoyl)homoserine lactone (OH-C10-HL), and N-(3-hydroxydodecanoyl)homoserine lactone (OH-C12-HL).
AHL production kinetics.
Growth of P. phosphoreum strain P100 in GMB or in juice from cod (10, 11) was monitored by enumeration on L&H or (only for GMB) by absorbance measurements at 600 nm. Samples were withdrawn and analyzed for AHL content by using a well diffusion assay with OH-C8-HL as the standard.
Studies of bioluminescence.
The nonbioluminescent, AHL-producing P. phosphoreum strain P100 was cross-streaked (on L&H at 15°C) against the luminescent, non-AHL-producing P. phosphoreum strain FR7 to determine potential AHL inhibition or induction of bioluminescence. P. phosphoreum strain FR7 was grown in GMB supplemented with 10% sterile supernatant from an outgrown culture of P. phosphoreum strain FR7 or P100 to test the presence of either an inducing factor produced by the luminescent P. phosphoreum strain FR7 or an inhibiting factor from P. phosphoreum strain P100. Finally, P. phosphoreum strain FR7 was grown in the presence of 1 µM OH-C8-HL or 1 µM N-oxo-hexanoyl-homoserine lactone (O-C6-HL; Sigma-Aldrich 142537-62-6) to test any effect from AHLs on light production. Light production was measured by using a Wallac scintillation counter (MicroBeta 1450-405 511).
Genetic manipulation techniques used with P. phosphoreum.
Transposon mutagenesis was performed by using Tn5 and Tn10 transposons in an attempt to isolate, sequence, and knock out the AHL synthetase enzyme encoded by an luxI or ainS homologue from P. phosphoreum strain P100 (32). P. phosphoreum was mixed with a plasmid donor (pBSL199, pBSL180, or pRL27), and the mixture was plated on an L&H plate for 24 h. The cells were removed, resuspended in peptone saline, and plated on selective L&H plates supplied with appropriate antibiotics (typically 50 µg of gentamicin/ml) to remove the donor and another antibiotic to select for transposon mutants (see Table 1). Several incubation temperatures and incubation times (15, 20, 25, 30, and 37°C for 0 to 6 h) were used for the heat treatment tests of the P. phosphoreum before it was mixed with the donor. A gene library of P. phosphoreum strain P100 was created by cutting chromosomal DNA (purified by using the QIAGEN genomic DNA Tip 100/G System; Hilden, Germany) with BclI restriction enzyme for 1 h. The fragments (purified by using GFX PCR DNA purification kit; Amersham Biosciences, Little Chalfont, Buckinghamshire, England) were inserted into pNU121 (42), and clones containing inserts were selected on LB5 medium supplemented with 5 µg of tetracycline/ml. Outgrown colonies were screened for AHL production by replica plating on indicative plates containing A. tumefaciens NT1 (pZLR4) LRF6, which is a tetracycline-resistant variant of A. tumefaciens NT1 (pZLR4) created by using the Tn10-tet transposon delivery vector pBSL199 (1) encoding tetracycline resistance.
Nucleotide sequence accession numbers.
The four 16S rDNA sequences identified in this study (for nonluminous P. phosphoreum isolates NCIMB 13478, NCIMB 13479, and P50 and for luminous P. phosphoreum isolate FR7) have been deposited in the EMBL, GenBank, and DDBJ nucleotide databases under accession numbers AJ746357 to AJ746360, respectively.
|
|
|---|
Identification of AHL-producing bacteria among the spoilage microbiota.
Ninety-seven potential AHL-producing colonies were isolated from L&H plates based on their ability to elicit a response in the replica plating procedure. These isolates represented both package conditions and all sampling points from day 12 onward (approximately six isolates per sampling). Most colonies eliciting a positive response on the replica plates were detected by A. tumefaciens NT1 (pZLR4). However, these colonies were difficult to isolate from the master plates, as they constituted only approximately 1% of the microbial community. All isolates were therefore tested for AHL production as single cultures by using a well diffusion assay. Twenty-three percent (22 of 97) of the isolates were AHL positive in this screening, and these isolates were all identified as P. phosphoreum or Aeromonas spp., while the rest of the isolates were either non-AHL-producing P. phosphoreum or Shewanella spp. (Table 2). To further confirm that AHL production is common among spoilage isolates of P. phosphoreum but not among Shewanella spp., we also tested isolates from previous storage trials (Table 1) as well as 51 Shewanella spp. from strain collections (Table 2). Clearly, compounds evoking a response in the A. tumefaciens assay were detected with P. phosphoreum but not Shewanella strains. The 16S rDNA sequences from nonluminous and AHL-positive P. phosphoreum strains (NCIMB 13478, NCIMB 13479, and P50) and from the luminous and AHL-negative P. phosphoreum strain (FR7) showed 99.2 to 99.4% similarity with the type strain of P. phosphoreum (ATCC 11040T). The 16S rDNA sequence similarity between the nonluminous P. phosphoreum P100 and the luminous isolate FR7 was 99.9%.
|
View this table: [in a new window] |
TABLE 2. Production of AHLs and bioluminescence by P. phosphoreum, Aeromonas spp., and Shewanella spp. isolated from fish or obtained from culture collectionsa
|
0.2. In contrast, ethyl acetate extracts contained a compound with an Rf of
0.4 (similar to that of synthetic OH-C8-HL). However, when OH-C8-HL was added to a supernatant from the non-AHL-producing P. phosphoreum strain FR7, it changed the mobility to an Rf of
0.2, similar to that of supernatants derived from AHL-producing P. phosphoreum. Ethyl acetate extracts of cod fillets contained a compound similar to ethyl acetate extracts of pure cultures (Rf
0.4), but the Rf changed when the extract was added to P. phosphoreum culture supernatants (Fig. 1). More than 20 presumed AHL-producing P. phosphoreum isolates and all AHL-containing extracts from cod were separated on TLC, and the same AHL-inducing spot was found in all isolates (results not shown). The AHL produced by Aeromonas had a similar mobility and shape as OH-C8-HL, and a few Aeromonas strains produced, in addition, a molecule which migrated slower than the OH-C8-HL. We therefore conclude that pure cultures of P. phosphoreum and Aeromonas spp. did indeed produce the AHL compound detected in spoiling cod fillets.
![]() View larger version (80K): [in a new window] |
FIG. 1. TLC profiles of cell-to-cell communication signals found in spoiling cod fillets and produced by P. phosphoreum. Plates illustrate profiles of extracts (panel a, lanes 1, 4, and 7; panel b, lanes 1, 4, and 6) as well as supernatants or extracts added to supernatants. (a) Lane 1, N-(3-hydroxyoctanoyl)homoserine lactone; lane 2, extract of vacuum-packed cod fillets added to the supernatant from P. phosphoreum strain LRF1; lane 3, the supernatant from P. phosphoreum strain LRF1; lane 4, extract of vacuum-packed cod fillets; lane 5, extract of MAP cod fillets added to the supernatant from P. phosphoreum strain LRF2; lane 6, the supernatant from P. phosphoreum strain LRF2; lane 7, extract of vacuum-packed cod fillets; lane 8, the supernatant of P. phosphoreum strain P100. (b) Lane 1, N-(3-hydroxyoctanoyl)homoserine lactone; lane 2, N-(3-hydroxyoctanoyl)homoserine lactone added to the supernatant from P. phosphoreum strain FR7 (non-AHL producing); lane 3, N-(3-hydroxyoctanoyl)homoserine lactone added to the supernatant from P. phosphoreum strain P100 grown in GMB; lane 4, extract of P. phosphoreum strain P100 grown in GMB; lane 5, the supernatant from P. phosphoreum strain P100 grown in ABTG medium; lane 6, extract of P. phosphoreum strain P100 grown in ABTG medium.
|
![]() View larger version (22K): [in a new window] |
FIG. 2. HPLC-ESI+-MS chromatograms. The lower left chromatogram is the total ion current (TIC) chromatogram from the Photobacterium phosphoreum strain P100 sample. The middle and the upper chromatograms are the extracted ion chromatograms corresponding to the masses of the [M+H]+ adducts of N-(3-hydroxyoctanoyl)homoserine lactone (OH-C8-HL) for the P. phosphoreum strain P-100 sample and a OH-C8-HL reference standard, respectively. The electrospray mass spectrum for OH-C8-HL is shown (lower right) along with the molecular structure and the calculated masses (calc. mass) for the three adducts. TOF, time of flight; ES+, positive-ion ESI.
|
The amount of AHL, as estimated by the well diffusion assay, increased at the same rate as growth in the nonbioluminescent AHL-producing P. phosphoreum strain P100 in a standard laboratory growth medium (Fig. 3a) as well as in buffered juice from cod fillets (results not shown). The light production in the bioluminescent non-AHL-producing P. phosphoreum strain FR7 appeared to be strongly up-regulated in late exponential phase (Fig. 3b).
![]() View larger version (18K): [in a new window] |
FIG. 3. Production of cell-to-cell communication signals in P. phosphoreum strain P100 measured as OH-C8-HL equivalents (a) and bioluminescence (luminescence counts per second; LCPS) (b) in P. phosphoreum strain FR7 grown in GMB at 5°C.
|
Finally, attempts to generate AHL-negative mutants of P. phosphoreum strain P100 (to determine putative quorum-regulated phenotypes) by using Tn5 and Tn10 mini-transposons failed. Subsequently, a gene library of P. phosphoreum P100 was constructed in E. coli. The library (25,000 colonies) was screened for luminescent colonies by using decanal as substrate and was screened for AHL-producing clones by using a replica assay, but none were found.
|
|
|---|
The standard TLC procedure for identification of AHLs used in this and many other studies (6, 23, 33, 49) resulted in the identification of two different molecules when supernatants from P. phosphoreum were compared with extracts of the same supernatants or extracts from cod. The reason for this apparent difference was that the bacterial supernatant altered the mobility of the molecule and thereby dramatically changed the tentative identification. Hence, it is advisable to extract bacterial culture supernatants before TLC profiling and identification.
Only a single circular spot (Rf
0.4) was detected by TLC in extracts from cod fillets packed in a vacuum or modified atmosphere and from all strains of P. phosphoreum. The component did not tail, comigrated with the OH-C8-HL reference standard, and was unambiguously identified for P. phosphoreum strain P100 by HPLC-ESI+-HRMS (Fig. 2). It was necessary to lower the content of the organic phase significantly, compared to that used in the study of Morin et al. (40), in the HPLC analysis of both the C18 and the phenyl (with polar capping) columns. The molecule, found in cod fillets, also comigrated with OH-C8-HL, indicating that P. phosphoreum could be the producer of this compound in packed cod fillets. Furthermore, the concentration of the component reached measurable levels as P. phosphoreum reached 106 CFU/g at a time when the aerobic plate count level had been at that level or higher levels for several days, strengthening the hypothesis that P. phosphoreum is the producer of the cell-to-cell signaling compound. This specific AHL has, to our knowledge, been found previously only in Pseudomonas fluorescence (49) and hydroxyl-substituted AHLs are identified only in three other species: Xenorhabdus nematophilus (19), Vibrio anguillarum (39), and V. harveyi (5). Interestingly, the Aeromonas spp. isolated in this study produced a compound with similar properties in TLC (results not shown).
The AHL production kinetics of P. phosphoreum was investigated and the concentration of AHL in withdrawn samples was estimated based on standard curves generated by using OH-C8-HL (45). Production kinetics showed that the presumed AHL is produced constitutively by P. phosphoreum strain P100, although this species is closely related to V. fischeri, in which the production of O-C6-HL is regulated in an autoinducible loop. This finding adds P. phosphoreum to the growing list of bacterial species in which AHL production appears to be constitutive (26, 28, 36, 45, 51).
None of the luminescent strains of P. phosphoreum produced AHLs, and this finding allowed us to speculate that AHL negatively regulates the bioluminescent phenotype. For instance, some V. fischeri autoinducer analogs may reduce bioluminescence in P. phosphoreum strain 8265 (21). There seems to be an overlap in the biochemical pathways in the synthesis of AHL and the aldehyde (tetraaldehyde in the case of P. phosphoreum) used to fuel the bioluminescent reaction. Both compounds use acyl-acyl carrier protein from the fatty acid biosynthesis pathway, although the acyl used for OH-C8-HL synthesis is probably octanoyl, while it is presumably tetradecanoyl for tetraaldehyde (43). However, we could not show experimentally that AHLs influence bioluminescence, since the addition of synthetic O-C6-HL, OH-C8-HL, or supernatants derived from P. phosphoreum strain P100 had no effect on luminescence for P. phosphoreum strain FR7. Similarly, V. fischeri autoinducer analogs had no effect on bioluminescence for V. harveyi strain MAV or Photobacterium leiognathi strain 721 (21). The bioluminescence of P. phosphoreum strain FR7 was, however, strongly up-regulated in the late exponential growth phase, indicating the presence of a specific regulatory circuit. The presence of specific luminescence-inducing molecules, such as a new type of autoinducers, could not be proven in this study since we did not find an early up-regulation of the bioluminescent phenotype in P. phosphoreum strain FR7 supplemented with supernatants derived from outgrown cultures of P. phosphoreum strain FR7. Although we did not find any AHLs in bioluminescent P. phosphoreum, they may, in principle, produce types of AHLs that are not recognized by the monitors and the HPLC-MS method used in this study. A. tumefaciens NT1 pZLR4 responds primarily to oxo-substituted AHLs and nonsubstituted AHLs (except C4) (49), whereas C. violaceum CV026 produces predominantly violacein in response to nonsubstituted AHLs but also responds to oxo-substituted compounds (37). As is evident from our study, NT1 also reacts when exposed to OH-substituted AHLs; hence, when used in combination, they will cover most of the well-known AHLs.
We do not know, at present, if P. phosphoreum uses AHL in a quorum-sensing regulatory circuit or what the affected phenotype(s) is (if any). If AHLs are used as quorum-sensing molecules, they may play a role in the natural environment of the bacterium, e.g., in the gastrointestinal tract of fish. In this environment, they may up-regulate activities, e.g., chitinase activity, and hence facilitate the degradation of crustaceans. However, both luminescent and nonluminescent strains are chitinolytic (11), but AHL regulation may allow a more specific up-regulation in late exponential growth phase. Although we did find a slight up-regulation of chitinase activity in P. phosphoreum in late exponential phase (data not shown), we were unable to create any mutations in P. phosphoreum, which would have allowed us to investigate the regulatory function of AHLs.
In conclusion, our study adds the marine bacterium P. phosphoreum to the growing list of AHL-producing bacteria. In many of these bacteria, the regulatory function of the signals is unknown, and further studies are needed to determine the role of OH-C8-HL in quorum sensing in P. phosphoreum.
The study was supported by the Danish Research Agency (641-00-01970) and The Danish Technical Research Council. K. F. Nielsen acknowledges the Danish Technical Research Council (9901295) for financial support.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»