Thermoregulation of N-Acyl Homoserine Lactone-Based Quorum Sensing in the Soft Rot Bacterium Pectobacterium atrosepticum

The psychrotolerant bacterium Pectobacterium atrosepticum producesfour N-acyl homoserine lactones under a wide range of temperatures.Their thermoregulation differs from that of the exoenzyme production,described as being under quorum-sensing control. A mechanisminvolved in this thermoregulation consists of controlling N-acylhomoserine lactones synthase production at a transcriptionallevel.

INTRODUCTION

Top

INTRODUCTION

Several gram-negative bacteria synthesize N-acyl homoserinelactone (HSL) signal molecules that serve for cell-to-cell communicationcalled for the first time quorum sensing (QS) by Fuqua et al.(7). Such regulatory systems operate to allow bacteria to sensecell density and to synchronize the functions of the entirepopulation (reviewed in references 15, 23, and 29). In the Pectobacteriumatrosepticum species (formerly Erwinia carotovora subsp. atroseptica)(8), QS regulation is involved in pectinolytic, cellulase, andprotease activities, tuber maceration, and harpin synthesisleading to a hypersensitive response in nonhost plant (18, 25).In contrast, QS via HSL production does not regulate growthand mobility during initial infection events or during liquidculture in a synthetic medium (25).

Pectobacterium atrosepticum is a psychrotrophic bacterium involvedin the soft rot of Solanum tuberosum (17, 26). Consequently,it causes important losses within cool temperate regions, wherepotatoes have traditionally been grown. The optimal temperaturefor pathogenicity, estimated to be around 20°C (19, 21),is a good compromise allowing both a fast multiplication (optimalat 24°C) and an efficient production of lytic enzymes, whichis optimal at temperatures ranging between 12 and 24°C (26).As the thermoregulation of bacterial multiplication differsfrom that of exoenzyme production, we investigated the thermoregulationof QS that triggers exoenzyme synthesis. Therefore, HSL productionand accumulation were measured during the three growth phasesof a typical P. atrosepticum strain (Table 1) cultured at sixtemperatures. The role of the expI gene encoding the HSL synthasewas determined by cloning this gene in Escherichia coli andby engineering a mutated strain of P. atrosepticum. Finally,the transcription of expI was estimated by relative reversetranscription-PCR (RT-PCR) under the same range of temperatures.

View this table:
[in this window]
[in a new window]

TABLE 1. Bacterial strains and plasmids^a

Effect of temperature on N-acyl HSL diversity and levels.

Top

Effect of temperature on...

Bacterial cultures were grown at 8, 12, 15, 20, 24, and 28°Cin minimal medium with polygalacturonic acid (PGA) as the solesource of carbon. This vegetable compound induces synthesisof virulence factors involved in plant disease or resistancethat have been demonstrated to be under QS control (25). Characterizationof HSLs was performed by high-performance liquid chromatographycoupled with mass spectrometry (16). In these conditions, P.atrosepticum 6276 produces four HSLs: mainly N-3-oxo-octanoyl-L-HSL(3-oxo-C8-HSL) and minor quantities of N-octanoyl-L-HSL (C8-HSL),N-3-oxo-hexanoyl-L-HSL (3-oxo-C6-HSL), and N-3-oxo-decanoyl-L-HSL(3-oxo-C10-HSL). The HSL levels increase from the middle tothe end of the logarithmic phase and then decline during thestationary phase (Fig. 1). Growth temperature significantlyinfluences HSL quantities produced, and optimal HSL productionoccurs at 24°C. In this psychrotolerant species, significantHSL levels were measured at 12°C and HSL production wasstill observed at 8°C. In contrast, a negative effect appearedat a warmer temperature (28°C). Lastly, the four HSLs wereproduced and accumulated in the same way at each temperature(Fig. 1). These data do not allow an association of the productionof a particular HSL with a typical growth phase or temperature.

View larger version (20K):
[in this window]
[in a new window]

FIG. 1. Effect of temperature on HSL production and accumulation by P. atrosepticum 6276. The HSLs were extracted from supernatant at exponential phase (white columns), late exponential phase (gray columns), and stationary growth phase (black columns). For each temperature point, at least three independent cultures were made. Each bar represents the standard deviation. OD₅₈₀U, unit of optical density at 580 nm.

Cloning, sequencing, and role of expI.

Top

Cloning, sequencing, and role...

The primers ExpI1C (5'TGAATTGGGCGGTAAAAATGT3') and ExpI2C (5'AATTCACCGTTGCCAAGAAG3'),deduced from the E. carotovora subsp. carotovora SCC1 genome(22), were used for amplification of expI by PCR. The singlePCR product (651 bp) was cloned into the pMOS Blue vector asdescribed by the manufacturer (pMOS Blue blunt-ended cloningkit; Amersham Pharmacia Biotech, Little Chalfont, United Kingdom)and introduced into E. coli XL1-Blue cells by electroporationaccording to standard procedures (13). Positive clones werechecked by PCR using the two above-described oligonucleotidesas primers, and the sequence of the cloned fragment was determined(Genome Express, Meylan, France). In silico sequence analyseswere performed with programs available at the Infobiogen website (Bioinformatic Resources center; http://www.infobiogen.fr/).A unique gene, named expI for exoenzyme production inducer (20),was identified in P. atrosepticum 6276 (GenBank accession no.AJ580599). Comparison of sequences of the 216-amino-acid-residuepeptide deduced from in silico translation showed, respectively,98% and 97% identities with E. carotovora subsp. carotovoraEC153 and E. carotovora SCC3193 HSL synthase proteins (4, 20).

An HSL-deficient P. atrosepticum 6276 mutant was obtained byrandom transposon mutagenesis. A bank of random insertion mutantswas made by using the mini-Tn5 transposon delivery system carriedby the pAG408 suicide vector (27). The Escherichia coli S17-1 ${lambda}$ pirstrain was used as the donor in a mating experiment to transferthe suicide plasmid into P. atrosepticum 6276. Transconjugantswere selected on PGA agar (25) plates supplemented with kanamycin(50 µg·liter^–1) and gentamicin (30 µg·liter^–1)and then tested on indicative LB plates containing an HSL monitorbacterium (Chromobacterium violaceum CV026) (14). Among transconjugantsanalyzed, one did not display purple color in the surroundingagar. For this mutant strain, named P. atrosepticum 6276-EI,PCR experiments show that the transposable element is integratedapproximately 500 bp downstream from the expI translation startcodon.

At the optimal temperature of production (24°C), HSL synthesiswas totally altered in P. atrosepticum 6276-EI, since no HSLwas detected. In contrast, tenuous but quantifiable amountsof 3-oxo-C8-HSL and C8-HSL were detected in the E. coli XL1-Bluestrain harboring the complete expI gene when grown at 24°Cin LB medium, but no HSLs were detected in the wild strain (datanot shown). Soft rot in tuber was assessed by testing at theoptimal temperature of HSL production as described previously(25). The test confirmed, in planta, the importance of the expIgene in the QS and in the consequent tissue maceration. No maceratedtissue was found on tubers 7 days after inoculation with P.atrosepticum 6276-EI (Fig. 2). This lack of maceration is attributedto the absence of HSL synthesis in the potato tuber.

View larger version (54K):
[in this window]
[in a new window]

FIG. 2. Role of expI gene on tuber disease. Potato tubers were infected by P. atrosepticum 6276 and incubated at 24°C. The development of symptoms was evaluated after 7 days.

Effect of temperature on expI expression.

Top

Effect of temperature on...

The kinetics of the expI transcription were determined by semiquantitativeRT-PCR experiments with P. atrosepticum 6276 cultured in PGAmedium. Total RNAs were extracted as described by the manufacturer(RNA isolation kit; Roche Diagnostics, Mannheim, Germany). Forcalibration, an RT-PCR was also performed using R16S1C (5'-GGGGGTAGAATTCCAGGTGT-3')and R16S2C (5'-CGGCAGTCTCCTTTGAGTTC-3') primers, amplifyinga unique 495-bp fragment of the P. atrosepticum E1 16S ribosomalgene (6). At each studied temperature, expI transcript amountswere most important at the end of the exponential growth phase.From 8°C to 24°C, amounts of transcripts increased inthe same manner as the growth temperature, and a negative effectappeared at 28°C (Fig. 3). The thermoregulation of expItranscription and that of HSL levels are identical throughoutthe bacterial culture (Fig. 1 and 3).

View larger version (27K):
[in this window]
[in a new window]

FIG. 3. Effect of temperature on expI gene expression in P. atrosepticum 6276. Transcription of expI was determined by RT-PCR experiments with cells harvested from exponential phase, late exponential phase, and stationary growth phase. Results obtained on 1% agarose gels were expressed as the ratio of expI transcripts to 16S transcripts.

Pectobacterium atrosepticum 6276 harbors the QS signaling systemof class I recently defined by Chatterjee et al. (4). This traitis linked to the structural characteristics of its HSL synthase(2, 28). We suggest that only 3-oxo-C8-HSL corresponds to aQS signal and that the others are less specific products ofExpI synthase or catabolites that appear during HSL turnover.This hypothesis is supported by (i) HSL concentrations usuallymeasured in other Pectobacterium strains (1, 3, 30), (ii) theabsence of HSL synthesis in an expI-deficient mutant, (iii)the major production of this HSL by another bacterial species(E. coli) complemented with expI, and (iv) the identical thermoregulationof each HSL.

The effect of temperature on extracellular HSL concentrationis the concomitant result of HSL production and degradation.Previous studies showed that HSL degradation could be influencedby pH or temperature: nonenzymatic lysis is increasingly favoredin LB medium with increasing pH (from neutral to mildly alkaline)and temperature over a range from 22 to 37°C (3, 31). However,in PGA medium, during the bacterial growth, a decrease of pHfrom 8 to 7.2 is observed (data not shown). In these conditions,the HSL autolysis seems to be minor, since we observed at eachtemperature tested an increase of HSL production from the logexponential to the early stationary phase over a decrease inthe pH of the medium (Fig. 1). Here we show that the thermoregulationof QS molecules reflects the thermoregulation of HSL synthesis.

Another report in this work is that the QS thermoregulationis identical to that of the bacterial growth that we have previouslydescribed for strain 6276 under the same culture conditions(26). However, by engineering P. atrosepticum 6276 which expressesa lactonase and so is unable to produce high concentrationsof HSLs, we have shown that QS does not control bacterial growthin vitro (PGA medium) and in planta (25). This suggests thatthe HSL synthesis pathway is not thermoregulated differentlyfrom the main pathways involved in the basal metabolism. Besides,our results show that one of the mechanisms involved in theQS thermoregulation consists of controlling HSL synthase productionby modulating the amounts of expI transcripts. This temperature-dependentregulation could occur at the level of transcription initiationinvolving temperature-sensitive sigma factors or by productionof small RNAs which would act as thermosensors (9, 10). In contrast,the QS thermoregulation differs from that one of the main virulencefactors studied in the soft rot bacterium Pectobacterium. Indeed,these bacteria secrete a large variety of extracellular lyticenzymes. Among them, the pectate lyase has been described asthe most important cause of maceration and cell killing (11).In P. atrosepticum 6276, the optimal temperature for pectatelyase production (between 12 and 15°C) is lower than theoptimal temperature for HSL synthesis (24°C), while thisexoenzyme production is under QS control (25, 26). These resultsstrongly suggest that the thermoregulation of pectate lyaseproduction occurs downstream of quorum signaling.

ACKNOWLEDGMENTS

We thank A. Dufour for critical review of the manuscript.

This study was supported by a grant from Conseil Régionalde Haute-Normandie and Comité Nord.

FOOTNOTES

* Corresponding author. Mailing address: Laboratoire de Microbiologie du Froid - UPRES 2123, Université de Rouen, 55 rue Saint-Germain, F-27000 Evreux, France. Phone: 33 2 32291549. Fax: 33 2 32291550. E-mail: xavier.latour{at}univ-rouen.fr

Published ahead of print on 27 April 2007.

REFERENCES

Top