R E S E A R C H A R T I C L E

Transcriptional activityofantifungalmetabolite-encoding genesphlDand hcnBC inPseudomonas spp.using qRT-PCRMelanie M. Paulin1, Amy Novinscak1, Marc St-Arnaud2, Claudia Goyer3, Nadine J. DeCoste1,Jean-Pierre Prive4, Josee Owen4 & Martin Filion1

1Department of Biology, Universite de Moncton, Moncton, NB, Canada; 2Institut de recherche en biologie vegetale, Jardin botanique de Montreal,

Montreal, QC, Canada; 3Agriculture and Agri-Food Canada, Fredericton, NB, Canada; and 4Agriculture and Agri-Food Canada, Bouctouche,

NB, Canada

Correspondence: Martin Filion, Department

of Biology, Universite de Moncton, Moncton,

NB, Canada E1A 3E9. Tel.: 11 506 858 4329;

fax: 11 506 858 4541; e-mail:

martin.filion@umoncton.ca

Received 21 October 2007; revised 17

December 2008; accepted 14 February 2009.

First published online 18 March 2009.

DOI:10.1111/j.1574-6941.2009.00669.x

Editor: Kornelia Smalla

Keywords

Pseudomonas; qRT-PCR; 2,4-

diacetylphloroglucinol (DAPG); hydrogen

cyanide (HCN); transcriptional activity.

Abstract

Production of 2,4-diacetylphloroglucinol (2,4-DAPG) and hydrogen cyanide

(HCN) by Pseudomonas spp. shows great potential for controlling soilborne plant

pathogens. However, little is known about the transcriptional activity of phl and

hcn genes encoding 2,4-DAPG and HCN, respectively. To progress toward a better

understanding of what triggers phl and hcn expression under rhizosphere condi-

tions, novel PCR primers and TaqMan probes were designed to monitor relative

phlD and hcnBC expression in quantitative real time-PCR assays. Transcriptional

activity of phlD and hcnBC was studied in time-course confrontational assays using

combinations of Pseudomonas spp. isolated in this study: LBUM300 (producing

2,4-DAPG and HCN) and LBUM647 (producing HCN only); pathogens Phy-

tophthora cactorum and Verticillium dahliae; and solid growth media King’s B

medium and potato dextrose agar. In correlation with the antagonistic activity

observed, expression of phlD and hcnBC and production of 2,4-DAPG was

detected throughout the 14-day course of the experiment in LBUM300 on both

media, while hcnBC expression diminished to undetectable levels in LBUM647. In

LBUM300 expression of phlD and hcnBC significantly changed over time and

was also influenced by the presence of pathogen and growth media following

time-dependent responses.

Introduction

Pseudomonas spp. producing the antifungal metabolites 2,4-

diacetylphloroglucinol (2,4-DAPG) and hydrogen cyanide

(HCN) can play a key role in the suppression of diseases

caused by various soilborne fungal pathogens. 2,4-DAPG

producers are best know for their implication in the

suppression of important root diseases such as take-all of

wheat caused by Gaeumannomyces graminis var. tritici

(Raaijmakers & Weller, 1998; de Souza et al., 2003b), black

root rot of tobacco caused by Thielaviopsis basicola (Keel

et al., 1992), and Pythium damping-off of cucumber (Gir-

landa et al., 2001; Notz et al., 2001) and sugarbeet (Shana-

han et al., 1992; Bergsma-Vlami et al., 2005). With the

exception of one Pseudomonas strain (Rezzonico et al.,

2007), all strains discovered to date that are capable of

producing 2,4-DAPG also produce HCN, a volatile antifun-

gal compound involved in suppression of diseases such as

black root rot of tobacco (Voisard et al., 1989). Pseudomonas

strains producing both 2,4-DAPG and HCN are generally

more effective at protecting plants against diseases than their

HCN-deficient counterparts (Sharifi-Tehrani et al., 1998).

Production of 2,4-DAPG and HCN is regulated by com-

plex operons and genetic mechanisms (Wei & Zhang, 2006).

Multiple biotic and abiotic factors such as the presence of

nutrients may also significantly modulate antifungal meta-

bolite production (Duffy & Defago, 1999; Notz et al., 2001;

Duffy et al., 2004). Although the influence of nutritional

sources on antifungal metabolite production has been

studied, results have almost exclusively been obtained under

liquid culture conditions. For most Pseudomonas strains, it

has been determined that the production of 2,4-DAPG is

stimulated by glucose, Zn21 and NH4Mo21, while inorganic

phosphate has a general inhibitory effect (Duffy & Defago,

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

1999). Addition of glycine, a precursor for HCN produc-

tion, and iron, are known to favor cyanogenesis (Voisard

et al., 1989; Laville et al., 1998; Blumer & Haas, 2000).

However, little is known about the impact of nutritional

factors under nonliquid culture conditions, as well as the

impact of biotic factors, such as the presence of pathogen

that may onset the expression of key antifungal metabolite-

encoding genes.

The use of gene reporter systems such as lacZ/b-galacto-

sidase (Notz et al., 2001; Siddiqui & Shaukat, 2003), luxAB

(Seveno et al., 2001), and more recently green fluorescent

protein-labelling (Baehler et al., 2005) have so far been the

methods of choice to study antifungal metabolite gene

expression in response to different treatments. However,

results obtained with such systems should be used with

caution. For example, in lacZ reporter systems, b-galactosi-

dase may remain in the medium as a stable product for a

certain period of time, leading to an inaccurate detection of

downshifts in gene expression (Schnider-Keel et al., 2000).

Furthermore, because the possibility that genetic transfor-

mation may interfere with gene expression of interest

cannot be completely ruled out, the development of novel

and more powerful methodological approaches should be

favored. Such tools should, however, first be validated under

controlled conditions to clearly define their limits before

being applied under complex soil rhizosphere conditions.

This step-by-step approach may help solve technical pro-

blems inherently associated with gene expression detection

under soil conditions. In this study, novel PCR primers and

TaqMan probes were designed to monitor relative phlD and

hcnBC gene expression in quantitative real time-PCR (qRT-

PCR) assays. Validation of the technique was performed

under solid growth media conditions. Effect of pathogen

presence and media on phlD and hcnBC gene expression was

then studied in time-course confrontational assays using

combinations of Pseudomonas spp. isolated in this study

from the rhizosphere of strawberry: LBUM300 (producing

2,4-DAPG and HCN; classified as a D genotype) and

LBUM647 (producing HCN only); pathogens Phytophthora

cactorum and Verticillium dahliae; and growth media King’s

B medium (KB) and potato dextrose agar (PDA).

Materials and methods

Bacterial strains and growth conditions

Two Pseudomonas spp. isolates designated as strains

LBUM647 and LBUM300, were isolated on KB (King et al.,

1954) from strawberry (Fragaria� ananassa Duch. cv.

Veestar) rhizosphere soil samples. The soil samples were

collected in 2004 from experimental plots located at the

Agriculture and Agri-Food Canada S. H. J. Michaud Re-

search Farm (Bouctouche, NB, Canada). Field soil was

characterized as a gleyed podzolic gray luvisol, a subgroup

of the Canadian System of Soil Classification, with a pH of

5.2, 62% sand, 25% silt, 13% clay, and 2.6% organic matter.

Strains were routinely grown at 25 1C for 48 h with shaking

at 250 r.p.m. in 9 mL of tryptic soy broth (TSB) (Difco). A

preliminary study (unpublished data) has determined that

Pseudomonas spp. LBUM300 carries the operons for the

production of HCN and 2,4-DAPG, while Pseudomonas

spp. LBUM647 carries only the operon for the production

of HCN.

Time-course in vitro fungal inhibition assays

Pathogenic isolates of V. dahliae Kleb and P. cactorum

(Lebert & Cohn) J. Schrot. were obtained from the Labor-

atoire de Diagnostic en Phytoprotection (QC, Canada) and

routinely grown at 25 1C on PDA (Difco). Inhibition assays

were performed on two different solid growth media, PDA

and KB, using the following combinations: (1) V. dahliae

1LBUM647, (2) V. dahliae1LBUM300, (3) V. dahliae alone

(V. dahliae control), (4) P. cactorum1LBUM647, (5)

P. cactorum1LBUM300, (6) P. cactorum alone (P. cactorum

control), (7) LBUM647 alone (LBUM647 control), and (8)

LBUM300 alone (LBUM300 control), for a total of 16

different treatments. Actively growing mycelial plugs

(0.5 cm diameter) were placed in the center of PDA or KB

agar plates, and for each treatment (except V. dahliae and

P. cactorum controls), plates were inoculated at two opposite

ends with 20 mL spots of a log culture of the same bacterial

isolate. LBUM647 and LBUM300 control plates were only

inoculated with bacteria as described above. In order to

study the transcriptional activity of hcnBC (for LBUM647

and LBUM300) and phlD (for LBUM300 only) over time,

for each treatment combination, a group of 13 identical

plates was prepared per block, corresponding to predeter-

mined sampling times for RNA extraction: 0, 6, 12, 24, 36,

48, 72, 96, 144, 192, 240, 288, and 336 h. After inoculation,

the plates were incubated at 25 1C. The experiment consisted

of three blocks, each containing 16 treatment combinations

and 13 sampling times, for a total of 624 plates. When

excluding V. dahliae and P. cactorum controls (not contain-

ing bacteria), 468 plates were used for bacterial RNA

isolation.

Bacterial RNA isolation

Total bacterial RNA was extracted directly from bacterial

colonies on each plate using the Mobio UltraClean Micro-

bial RNA Isolation Kit (Mobio, Carlsbad, CA). RNA extrac-

tions were performed according to the manufacturer’s

protocol with the following modification: bacterial colonies

were scraped from the plates with an inoculating loop

and placed in 1.8 mL of TSB. Extractions then immediately

followed as indicated. To eliminate genomic DNA

FEMS Microbiol Ecol 68 (2009) 212–222 c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

213Transcriptional activity of phlD and hcnBC using qRT-PCR

contaminations, DNA digestion reactions were carried out

using 50mL of extracted RNA diluted 1 : 2 with diethyl

pyrocarbonate (DEPC)-treated water. DNAse treatments

were carried out according to the manufacturer’s protocol,

first using 6 U of TURBO DNA-free enzyme (Ambion,

Austin, TX) followed by two treatments using 6 U and 4 U

of DNA-free enzyme (Ambion), respectively. Total RNA

quantity and quality were then evaluated by spectrophoto-

metry using a NanoDrops ND-1000 spectrophotometer

(NanoDrop Technologies, Wilmington, DE).

Primer design

PCR primers and TaqMan probes for qRT-PCR and qPCR

reactions were designed using the software BIOEDIT v. 7.0.4.1

(Hall, 1999) based on sequences from partial hcnBC, phlD,

and 16S rRNA genes of Pseudomonas spp. LBUM300 and

LBUM647 (DQ788986, DQ788990, DQ788992, DQ788999,

DQ789001) and homologous sequences retrieved from the

GenBank database (Table 1). Specificity was confirmed

using BLASTN (Altschul et al., 1997) and PCR amplifications

on DNA extracted from a collection of Pseudomonas sp.

known for harboring or not HCN and/or 2,4-DAPG coding

genes (data not shown). Each TaqMan probe was labelled

with an FAM reporter dye at the 50-end and a BHQ-1

quencher dye at the 30-end (Alpha DNA, Montreal, Cana-

da). To confirm that a single amplicon of interest was

obtained from each primer-probe combination, qRT-PCR

amplification products were separated by agarose gel elec-

trophoresis, amplicon size was determined by comparison

with a 1-kb Plus DNA ladder (Invitrogen, Burlington,

Canada), and fragments were purified and then sequenced.

qRT-PCR and qPCR

All qRT-PCR reactions were performed using the Super-

script III Platinum One-Step Quantitative RT-PCR System

(Invitrogen) in an MJ-Research DNA Engine Opticon 2

system (Bio-Rad, Hercules, CA). Each qRT-PCR run in-

cluded cDNA synthesis and amplification of the target phlD

and/or hcnBC, as well as the 16S rRNA reference gene

transcripts. Reactions were prepared containing 12.5 mL of

2�Reaction Mix (Invitrogen), 10 mM of each correspond-

ing primer (Alpha DNA), 10mM of the TaqMan fluorogenic

probe (Invitrogen), 0.5mL of Superscript III RT/Platinum

Taq Mix (Invitrogen), 2.5 mL of DNAse-treated RNA diluted

1 : 2, and sterile DEPC-treated water for a final volume of

25 mL. The qRT-PCR thermal cycling program consisted of

an initial 15 min incubation at 50 1C for cDNA synthesis

followed by denaturation at 95 1C for 2 min and then 40

cycles at 95 1C for 15 s, 60 1C for 30 s, and optical plate reads

following each cycle amplification. All qRT-PCR reactions

were replicated twice.

To ensure complete elimination of DNA contaminations

following DNAse digestions, RNA samples were also sub-

jected to qPCR amplifications with appropriate primer sets

and probes for phlD, hcnBC, and 16S rRNA genes (Table 1).

Reactions were prepared using the Platinum Quantitative

PCR SuperMix-UDG kit (Invitrogen) containing 12.5mL

of Platinum Quantitative PCR SuperMix-UDG, 10mM of

each corresponding primer (Alpha DNA), 10 mM of the

TaqMan fluorogenic probe (Invitrogen), 2.5 mL of DNAse-

treated RNA diluted 1 : 2, and sterile DEPC-treated water for

a final volume of 25 mL. The thermal cycling program was

identical to the qRT-PCR assays with the exception of the

initial incubation at 50 1C. qPCR amplifications were car-

ried out in an MJ-Research DNA Engine Opticon 2 system

(Bio-Rad).

Relative quantification

The relative quantification ratios of phlD and hcnBC gene

transcripts standardized to 16S rRNA reference gene tran-

scripts were calculated according to a mathematical model

in which no calibration curve is required (Pfaffl, 2001). Fold

changes were determined by the cycle value at which

fluorescence of the genes of interest cross threshold levels

(Ct) while taking into account the correction for reaction

efficiencies of each primer pair, as follows:

Fold change¼ðEtargetÞDCttarget

ðcontrol�sampleÞ=ðEref ÞDCt

ref

ðcontrol�sampleÞ

Therefore, before qRT-PCR measurements, dilution

curves were generated from extracted RNA samples serially

Table 1. Primers and probes used in this study

Primers and probes Sequence (50 ! 30) Target gene Product length (bp) References

Phl2a GAGGACGTCGAAGACCACCA phlD 127 Raaijmakers et al. (1997)

Phl2q CGGCGGACGGAAAATTCTTGA This study

phlD probe CGCGACCCGACCGGGTTCCA This study

HCNn ATGTCGGCCAACCGCAAG hcnBC 233 This study

Acb ACGATGTGCTCGGCGTAC Ramette et al. (2003)

hcn probe CCGCAGTTGCACCGCGAGCT This study

F311Ps CTGGTCTGAGAGGATGATCAGT 16S rRNA 202 Milling et al. (2005)

Pseudo2n TCGGTAACGTCAAAACACTAACGT Purohit et al. (2003) modified

16S probe CGAAAGCCTGATCCAGCCATGCCGC This study

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

214 M.M. Paulin et al.

diluted 10-fold in order to calculate individual reaction

efficiency (E = 10[�1/slope]). Reactions for qRT-PCR were set

up for phlD, hcnBC, and 16S rRNA genes as described

previously and the slope of each dilution curve was used to

calculate primer efficiency. In order to follow fold changes in

gene expression over time, Ct values of target genes from

each sample treatment were determined relative to the Ct

values of corresponding qRT-PCR amplifications when first

detected in all treatments (6 h), and then normalized to

levels of 16S rRNA gene transcripts amplification.

2,4-DAPG isolation and detection

A log-phase culture of LBUM300 was used to apply 20 mL

aliquots at two opposite ends on the surface of PDA and KB

agar plates and incubated at 25 1C as described above. To

extract 2,4-DAPG, two agar disks of 2 cm in diameter (total

weight c. 2 g) that included the bacterial colonies were cut

with a cork borer from each growth media plate at time 6,

12, 24, 48, 96, 144, 240, and 336 h in triplicate. Extraction of

2,4-DAPG was carried out according to the protocol of

Bonsall et al. (1997) that was modified as follows: the two

disks were combined and homogenized using a Polytron

homogenizer (Brinkmann Instruments, Westbury, NY) for

1 min in 10 mL of 80% acetone. The extracts were then

centrifuged in a clinical centrifuge at maximum speed for

10 min and the acetone was removed from the supernatant

at 35 1C using a rotary evaporator (Buchi, Switzerland).

The remaining water (c. 2 mL) was acidified to pH 2.0 with

88 mL of 10% trifluoroacetic acid (TFA), and extracted

twice with 10 mL of ethyl acetate. The ethyl acetate extracts

were combined, evaporated to dryness using a rotary

evaporator at 35 1C, and then stored at � 20 1C until

processed. Dried extracts were redissolved in 5 mL of

35% acetonitrile, diluted as appropriate, filtered through

0.45 m syringe filters, and injected into a LiChrosorb RP-18

column (4.6� 200 mm) using a Rheodyne Model 7125

injector with a 20mL sample loop. The Agilent 1100 Series

HPLC system (Agilent Technologies, Wilmington, DE) con-

sisted of a quaternary pump, photodiode array detector

and CHEMSTATION chromatographic software (Agilent Tech-

nologies). Solvent conditions included a flow rate of

1.0 mL min�1 with a 2-min initialization at 10% acetonitrile

(ACN)-0.1% TFA followed by a 20-min linear gradient to

100% ACN–0.1% TFA. HPLC gradient profiles were mon-

itored at 270 nm. 2,4-DAPG eluted at a retention time of

15.9 min in c. 73% ACN. Six-point standard curves yielded a

correlation coefficient of 0.9999. An efficiency of 70% for

2,4-DAPG extraction was obtained for both PDA and KB

agar plates by extracting a known quantity of pure 2,4-

DAPG (Toronto Research Chemicals, North York, ON,

Canada) that had been incorporated into each medium

before extraction.

Statistical analysis

Statistical analyses were performed using the Correlation

and General Linear Model procedure of the SAS statistical

software (SAS Institute Inc., 1992). For LBUM300 the effects

of time, pathogens, and media on gene expression fold

changes, as well as interactions between treatments, were

analyzed by repeated measures ANOVA. Gene expression fold

changes were submitted to power transformations when

needed to meet the requirements of the tests. For LBUM647,

the effect of pathogens and media on gene expression fold

changes were analyzed per sampling time using Friedman

nonparametric tests. The effect of media and time on 2,4-

DAPG quantification, as well as interactions between treat-

ments were analyzed by ANOVA. 2,4 DAPG quantification

data were submitted to log transformation to normalize

skewed distributions before performing the ANOVA. For

ANOVA and repeated measures ANOVA, a posteriori compar-

isons of the means between treatment levels were carried out

using Tukey’s studentized range tests at a 5% level of

significance. For Friedman analyses, a posteriori compari-

sons of the means between treatment levels were carried out

using Tukey–Kramer honestly significant differences tests at

a 5% level of significance.

Results

Time-course in vitro fungal inhibition assays

For each treatment combination, fungal and bacterial

growth was visually assessed and progression was documen-

ted by photography at each sampling time for RNA extrac-

tion over the entire 2 weeks of the experiment. When

compared with fungal growth of V. dahliae alone on KB

plates, coinoculations of V. dahliae with LBUM300 or

LBUM647 yielded a fungal inhibition (moderate inhibition)

that persisted throughout 336 h (Fig. 1). Moderate levels of

inhibition were similarly observed when P. cactorum was

coinoculated with either LBUM300 or LBUM647. On PDA,

coinoculation of V. dahliae with LBUM300 yielded a mod-

erate level of fungal inhibition, while coinoculation with

LBUM647 only yielded a slight growth inhibition (Fig. 2).

Phytophthora cactorum coinoculated with LBUM300 yielded

a high level of inhibition, while coinoculation with

LBUM647 was only slightly inhibitory to mycelium growth.

qRT-PCR primer efficiency and linearity

Before assessing expression levels of target genes by qRT-

PCR, the efficiency of each primer set was determined by

establishing 10-fold dilution curves of RNA samples ex-

tracted from strains LBUM300 and LBUM647. For each

gene, qRT-PCR amplification of RNA was always detected

for at least the first four dilutions. Linear regression analysis

FEMS Microbiol Ecol 68 (2009) 212–222 c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

215Transcriptional activity of phlD and hcnBC using qRT-PCR

Fig. 1. Time course in vitro fungal inhibition assays on KB growth medium. Growth progression of (a) Verticillium dahliae control, (b) V. dahliae

1LBUM300, (c) V. dahliae1LBUM647, (d) Phytophthora cactorum control, (e) P. cactorum1LBUM300, (f) P. cactorum LBUM647, (g) LBUM300 control,

and (h) LBUM647 control, from 6 to 336 h of incubation at 25 1C.

Fig. 2. Time course in vitro fungal inhibition assays on PDA growth medium. Growth progression of (a) Verticillium dahliae control, (b) V. dahliae

1LBUM300, (c) V. dahliae1LBUM647, (d) Phytophthora cactorum control, (e) P. cactorum1LBUM300, (f) P. cactorum LBUM647, (g) LBUM300 control,

and (h) LBUM647 control, from 6 to 336 h of incubation at 25 1C.

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

216 M.M. Paulin et al.

revealed that all dilution curves were highly linear with a

correlation coefficient (r2) of Z0.938. The qRT-PCR effi-

ciency for each primer set was calculated with slopes ranging

from � 3.45 to � 6.81, indicating a PCR efficiency of

1.84, 1.40, and 1.95 for phlD, hcnBC, and 16S rRNA genes,

respectively.

Primer specificity and DNAse treatments

qRT-PCR amplification products generated single bands of

appropriate size when run on gel electrophoresis and of

appropriate composition when sequenced, therefore con-

firming specificity of primers and probes for phlD, hcnBC,

and 16S RNA gene transcripts detection (data not shown).

qPCR amplification of DNAse-treated RNA samples did not

cross threshold levels before reaching at least 40 cycles of

amplification, clearly indicating sufficient elimination of

contaminating DNA for further RNA expression analysis by

qRT-PCR.

Time-course qRT-PCR relative gene expression

The relative expression of phlD and hcnBC genes was

analyzed over 13 time points ranging from 0 to 336 h of

incubation. The Ct of target gene transcripts were normal-

ized to those of endogenous 16S rRNA reference gene

transcripts while taking into account primer efficiencies.

This allowed to determine total fold change of 2,4-DAPG

and HCN coding genes over the entire course of the

experiment. RNA extracted at 0 h of incubation did not

yield consistent gene amplification among treatments. Con-

sequently, Ct values obtained after 6 h of incubation were

chosen as the control values in order to calculate fold change

of gene expression over time for each treatment. For each

combination of treatment and incubation time, mean values

of three experimental runs (i.e. three separate replications of

the experiment) that each underwent two different qRT-

PCR assays are presented in Figs 3, 4, and 5.

Overall, the transcriptional activity of phlD from

LBUM300 significantly changed over time (Po 0.0001).

There was a significant interaction between combinations

of time�media (Po 0.0001) and time� pathogen

(Po 0.05). On KB medium, LBUM300 phlD transcriptional

activity was consistently lower over time than initial levels

observed at 6 h of incubation, whether in the presence of

either pathogen or when inoculated alone, although bacteria

control plates underwent a smaller expression decrease

than when coinoculated with pathogens (Fig. 3a). While

LBUM300 phlD expression on PDA medium followed

similar patterns when coinoculated with P. cactorum or on

bacterial control plates, the presence of V. dahliae generally

increased phlD expression, which remained upregulated at

336 h (Fig. 3b).

The overall transcriptional activity of hcnBC from

LBUM300 followed similar trends and was also significantly

altered over time (Po 0.0001). There was a significant

interaction between combinations of time�media

(Po 0.0001) and time� pathogen (Po 0.0001). On KB

medium, the expression of hcnBC from LBUM300 was

mostly upregulated for all treatment combinations, espe-

cially when inoculated alone or when coinoculated with

P. cactorum (Fig. 4a). In these two cases, expression of

hcnBC gradually rose over time reaching a peak around

72 h of incubation; thereafter, the expression gradually

diminished but could still be detected after 336 h. On PDA

medium, the expression of hcnBC from LBUM300 was

mostly upregulated when confronted with V. dahliae, where

hcnBC underwent important fold changes reaching a peak

later into the experiment around 192 h of incubation

Fig. 3. Relative expression levels of phlD in Pseudomonas sp. LBUM300

normalized to 16S rRNA gene expression when inoculated on KB (a), and

PDA (b) plates. On each media, fold change of the bacterial target gene

was determined in the presence of Verticillium dahliae, Phytophthora

cactorum, and on control plates without pathogen. Bars are SE of the

mean. For each time, values followed by a different letter are significantly

different using Tukey’s studentized range test (Po 0.05).

FEMS Microbiol Ecol 68 (2009) 212–222 c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

217Transcriptional activity of phlD and hcnBC using qRT-PCR

(Fig. 4b). Unlike hcnBC expression patterns on KB in the

presence of P. cactorum, LBUM300 hcnBC transcriptional

activity on PDA medium was generally downregulated in

comparison with initial expression at 6 h, but could still be

detected after 336 h. While expression of hcnBC on

LBUM300 control plate fluctuated between down- and

upregulation with no clear trend, expression could still be

detected after 336 h.

In Pseudomonas sp. LBUM647, expression patterns of

hcnBC were completely different than the ones found for

Pseudomonas sp. LBUM300. No repeated measures ANOVA

analyses could be performed on these data because require-

ments of the tests were not fulfilled, due in part to a high

number of samples in which no transcriptional activity was

detected, and the inherent skewed distributions of residuals.

Even when only analyzing the first 72 h time period of the

experiment, where expression data were obtained for each

treatment, it was impossible to meet the requirements of the

tests. Instead, nonparametric Friedman tests were per-

formed on data obtained up to 72 h. Friedman tests were

first performed to compare hcnBC expression for each

pathogen treatment per time and per media (Fig. 5). For

each sampling time, no effect of pathogens was found on

either KB or PDA media. The general effect of media on

hcnBC expression, regardless of pathogen treatments, was

then tested per time using Friedman tests. Except for 12 and

24 h, no media effect was found. Although it was impossible

to statistically analyze the general time effect on hcnBC

expression, clear trends could be observed. Coinoculation

of LBUM647 with either P. cactorum or V. dahliae yielded a

general downregulation of hcnBC gene expression, leading

Fig. 4. Relative expression levels of hcnBC in Pseudomonas sp.

LBUM300 normalized to 16S rRNA gene expression when inoculated on

KB (a), and PDA (b) plates. On each media, fold change of the bacterial

target gene was determined in the presence of Verticillium dahliae,

Phytophthora cactorum, and on control plates without pathogen. Bars

are SE of the mean. For each time, values followed by a different letter

are significantly different using Tukey’s studentized range test (Po 0.05).

Fig. 5. Relative expression levels of hcnBC in Pseudomonas sp.

LBUM647 normalized to 16S rRNA gene expression when inoculated on

KB (a), and PDA (b) plates. On each media, fold change of the bacterial

target gene was determined in the presence of Verticillium dahliae,

Phytophthora cactorum, and on control plates without pathogen. Bars

are SE of the mean. For each time up to 72 h, values followed by the

same letter are not significantly different as determined using a

Tukey–Kramer honestly significant differences test (Po 0.05).

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

218 M.M. Paulin et al.

to no detectable expression after 72 h, with the exception

of a slight 0.1-fold change detected in the presence of

P. cactorum at 144 h (Fig. 5a). On LBUM647 control plate,

hcnBC gene expression followed the same general pattern

gradually decreasing with no expression detected after 72 h.

On PDA, hcnBC expression of LBUM647 was downregu-

lated in all treatments with no expression detectable after

96 h of incubation (with three exceptions at 144, 192, and

288 h showing very low levels of expression) (Fig. 5b).

Overall, expression of hcnBC genes was notably stronger

and longer lasting in LBUM300 capable of producing both

2,4-DAPG and HCN than in strain LBUM647 producing

only HCN.

2,4-DAPG isolation and detection

To confirm that phlD gene expression leads to 2,4-DAPG

production, 2,4-DAPG was extracted and quantified from

PDA and KB plates containing 20 mL aliquots of a log-phase

culture of LBUM300 at two opposite ends of each plate. 2,4-

DAPG isolation and detection performed 6, 12, 24, 48, 96,

144, 240, and 336 h following bacteria inoculation revealed

the presence of the antifungal metabolite at each time point

on both media. For each combination of treatment and

incubation time, mean values of three experimental runs

(i.e. three separate replications of the experiment) are

presented in Fig. 6. Overall, the amount of 2,4-DAPG

detected was significantly altered by time (Po 0.001) and

growth media (Po 0.001). There was also a significant

interaction between media� time (Po 0.001). On PDA

medium, amounts of 2,4-DAPG detected was consistently

lower over time, except at 336 h, than amounts detected on

KB medium. Increased growth of LBUM300 was, however,

observed on KB as compared with PDA media. As plugs of

defined diameters were extracted from both media for 2,4-

DAPG quantification, increased cell density on KB could

have accounted for the increase in 2,4-DAPG detection

observed.

Discussion

In this study, novel PCR primers and TaqMan probes were

designed to monitor relative phlD and hcnBC gene expres-

sion in qRT-PCR assays. Validation of the specificity,

linearity, and efficiency of the assays was first performed,

and qRT-PCR was then used to study the temporal effect

pathogens and solid growth media have on phlD and hcnBC

gene expression in two different Pseudomonas sp. strains,

LBUM300 producing 2,4-DAPG and HCN, and LBUM647

producing only HCN. LBUM300 displayed the strongest

antagonism toward the growth of V. dahliae and P. cactorum

on both growth media throughout the 336 h of the experi-

ment. Keel et al. (1996) also reported that Pseudomonas

strains Q69c-80 and mutant Q2-87<Tn5-1, both nonpro-

ducers of 2,4-DAPG but still capable of producing HCN,

inhibited the growth of the pathogen G. graminis var. tritici

on KB and MA agar, but to a lesser extent than most other

strains capable of producing both metabolites. In correla-

tion with the antagonistic activity observed, the transcrip-

tional activity of phlD in LBUM300 was detected during the

entire course of the experiment on both media for all

pathogen inoculation treatments. While expression patterns

of hcnBC differed greatly between both Pseudomonas strains,

strain LBUM300 also carrying genes for the production of

2,4-DAPG was the only strain able to maintain hcnBC

expression throughout the 336 h of the experiment. These

results suggest a correlation between the capacity to main-

tain antifungal metabolite gene expression over time and the

respective antagonistic activity observed.

The expression of phlD in Pseudomonas sp. LBUM300

was upregulated when confronted to V. dahliae on PDA,

while it was downregulated in the same conditions on KB.

The effect of the medium itself is not surprising because

PDA, contains dextrose (D-glucose) and that glucose has

previously been found to stimulate production of 2,4-DAPG

in strains of Pseudomonas (Duffy & Defago, 1999). As for

phlD, the expression of hcnBC in LBUM300 was more

upregulated when confronted to V. dahliae on PDA than on

KB medium. However, these trends did not lead to a

significant higher level of antagonism on PDA than on KB.

The studies that have so far described gene expression

associated with 2,4-DAPG production have been mainly

carried out in liquid culture using reporter systems, with

expression usually peaking between 20 and 25 h of incuba-

tion, corresponding to the late exponential phase of growth,

Time (h)6 h 12 h 24 h 48 h 96 h 144 h 240 h 336 h

µg 2

,4-D

AP

G p

er s

ampl

e

0

50

100

150

200

250

300KBPDA

∗

∗

∗ ∗∗

∗

∗

Fig. 6. Time course 2,4-DAPG production assays by Pseudomonas sp.

LBUM300 on KB and PDA plates. On each medium, 2,4-DAPG quantifi-

cation was achieved from 6 to 336 h following inoculation with

LBUM300. Bars are SE of the mean. For each time, an asterisk indicates

a significant difference determined using Tukey’s studentized range test

(Po 0.05).

FEMS Microbiol Ecol 68 (2009) 212–222 c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

219Transcriptional activity of phlD and hcnBC using qRT-PCR

and gradually declining thereafter, with no expression

detectable after a few hours (Schnider-Keel et al., 2000; Notz

et al., 2001). It has, however, been noted before that the

production of 2,4-DAPG could be more important on solid

than liquid media, but the transcriptional activity of the

genes associated with such production on solid media has to

our knowledge never been described before. 2,4-DAPG

production in Pseudomonas fluorescens Q2-87 was 47.7

times greater on solid YM medium than in its liquid

counterpart, with a production of 32mg mL�1 of culture

medium (Bonsall et al., 1997). Similar production rates were

described for the same strain on solid MA and KB media,

producing 35.6 and 37.4 mg mL�1, respectively (Keel et al.,

1996). In this study, strain LBUM300 achieved production

up to 120.1 and 34.0 mg mL�1 on KB and PDA, respectively.

This could further explain why we have been able to detect

the expression of phlD on solid KB and PDA media over a

much longer period of time than under liquid culture

conditions, even after 2 weeks of incubation.

Expression patterns of hcnBC genes in LBUM647 were

strikingly different than in LBUM300. When confronted

with either pathogen, hcnBC expression in LBUM647 was

mostly downregulated and became undetectable after

72–96 h of incubation. However, pathogen inhibitions were

maintained on both media, especially on KB medium, even

though expression could not be detected following 72–96 h

of incubation. This suggest that other growth-inhibiting

compounds might have been produced and released by

Pseudomonas sp. LBUM647 such as iron-sequestering side-

rophores, lytic enzymes, or other unidentified metabolites

(Keel et al., 1996; Duffy & Defago, 1999; Haas & Defago,

2005). Furthermore, volatile HCN could have remained in

the sealed Petri dishes, especially if large initial amounts

were released. Varying levels of pathogen inhibition might

also have been due to variations in pathogen sensitivity to

antifungal metabolites production, as it has previously been

reported (Mazzola et al., 1995; de Souza et al., 2003a; Duffy

et al., 2004).

These results clearly show that factors such as pathogen

presence and media composition can alter patterns of phlD

and hcnBC expression in Pseudomonas strains over time

when grown on solid media. But does gene expression

accurately reflect production of antifungal metabolites in

the medium? In a previous study (Notz et al., 2001),

expression of phlA in P. fluorescens CHA0, was compared

with 2,4-DAPG production in vitro, using a phlA–lacZ

reporter system. Results indicated that production of 2,4-

DAPG in KB closely followed the paralleled phlA expression

over 160 h of incubation. At 160 h, neither expression of

phlA nor 2,4-DAPG could be detected in the medium.

Because 2,4-DAPG is known for being specifically degraded

by the producing bacterium via the action of a hydrolase

termed PhlG that converts 2,4-DAPG to MAPG (Bottiglieri

& Keel, 2006), it is usually assumed that 2,4-DAPG does not

accumulate as a stable product in the medium in most, if not

all 2,4-DAPG-producing pseudomonads. We have shown in

this study that phlD expression and 2,4-DAPG production

could still be detected after 336 h of incubation on solid

media, either KB or PDA. This indicates that the antifungal

metabolite was still released by the bacterium, even after

336 h of incubation, and could have contributed, in parallel

with HCN production, to the inhibition of growth observed

for both fungal pathogens. Even if gene expression and 2,4-

DAPG production fluctuated according to the different

treatments, production of 2,4-DAPG has been maintained

at concentrations that appear sufficient to cause mycelium

growth inhibition. It has previously been suggested that a

stronger production of 2,4-DAPG does not necessarily

translate into enhanced biocontrol activity in vitro or in the

rhizosphere, as pathogen sensitivity may vary, and concen-

trations necessary for inhibition may attain thresholds (Keel

et al., 1992). Other factors such as root colonization ability,

population densities sufficient for biocontrol, and environ-

mental factors triggering gene expression may also come

into play. Thus, it might be favorable for pseudomonad

strains to maintain basal levels of 2,4-DAPG, and also

possibly HCN production, over time instead of overprodu-

cing these potentially phytotoxic compound.

Monitoring antifungal gene expression over time under

in situ rhizosphere soil conditions will most likely represent

the next important challenge that will provide valuable

information about the ecology of antifungal metabolite-

producing strains. Sensitive techniques such as the real-time

detection of mRNA by qRT-PCR described in this study

will undoubtedly become valuable tools to monitor gene

expression of key antifungal metabolites under complex

environmental conditions.

Acknowledgements

This work was supported by National Sciences and Engi-

neering Research Council of Canada and New Brunswick

Innovation Foundation grants to Martin Filion. We thank

Lucie Bijeau, Melanie Demers, Mame Daro Faye, France

LeBlanc, Sophie LeBlanc, Francoise Morin, and Chadi

Nassar for their technical assistance.

References

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W

& Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new

generation of protein database search programs. Nucleic Acids

Res 25: 3389–3402.

Baehler E, Bottiglieri M, Pechy-Tarr M, Maurhofer M & Keel C

(2005) Use of green fluorescent protein-based reporters to

monitor balanced production of antifungal compounds in the

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

220 M.M. Paulin et al.

biocontrol agent Pseudomonas fluorescens CHA0. J Appl

Microbiol 99: 24–38.

Bergsma-Vlami M, Prins ME & Raaijmakers JM (2005)

Influence of plant species on population dynamics, genotypic

diversity and antibiotic production in the rhizosphere by

indigenous Pseudomonas spp. FEMS Microbiol Ecol 52:

59–69.

Blumer C & Haas D (2000) Mechanism, regulation, and

ecological role of bacterial cyanide biosynthesis. Arch Microbiol

173: 170–177.

Bonsall RF, Weller DM & Thomashow LS (1997) Quantification

of 2,4-diacetylphloroglucinol produced by fluorescent

Pseudomonas spp. in vitro and in the rhizosphere of wheat.

Appl Environ Microbiol 63: 951–955.

Bottiglieri M & Keel C (2006) Characterization of PhlG, a

hydrolase that specifically degrades the antifungal compound

2,4-diacetylphloroglucinol in the biocontrol agent

Pseudomonas fluorescens CHA0. Appl Environ Microbiol 72:

418–427.

de Souza JT, Arnould C, Deulvot C, Lemanceau P, Gianinazzi-

Pearson V & Raaijmakers JM (2003a) Effect of 2,4-

diacetylphloroglucinol on Pythium: cellular responses and

variation in sensitivity among propagules and species.

Phytopathology 93: 966–975.

de Souza JT, Weller DM & Raaijmakers JM (2003b) Frequency,

diversity, and activity of 2,4-diacetylphloroglucinol-producing

fluorescent Pseudomonas spp. in Dutch take-all decline soils.

Phytopathology 93: 54–63.

Duffy B & Defago G (1999) Environmental factors modulating

antibiotic and siderophore biosynthesis by Pseudomonas

fluorescens biocontrol strains. Appl Environ Microbiol 65:

2429–2438.

Duffy B, Keel C & Defago G (2004) Potential role of pathogen

signaling in multitrophic plant–microbe interactions involved

in disease protection. Appl Environ Microbiol 70: 1836–1842.

Girlanda M, Perotto S, Moenne-Loccoz Y, Bergero R, Lazzari A,

Defago G, Bonfante P & Luppi AM (2001) Impact of

biocontrol Pseudomonas fluorescens CHA0 and a genetically

modified derivative on the diversity of culturable fungi in the

cucumber rhizosphere. Appl Environ Microbiol 67: 1851–1864.

Haas D & Defago G (2005) Biological control of soil-borne

pathogens by fluorescent pseudomonads. Nat Rev Microbiol 3:

307–319.

Hall TA (1999) BioEdit: a user-friendly biological sequence

alignment editor and analysis program for Windows 95/98/

NT. Nucleic Acids Symp Ser 41: 95–98.

Keel C, Schnider U, Maurhofer M, Voisard C, Laville J, Burger U,

Wirthner P, Haas D & Defago G (1992) Suppression of root

diseases by Pseudomonas fluorescens CHA0: importance of the

bacterial secondary metabolite 2,4-diacetylphloroglucinol.

Mol Plant Microbe In 5: 4–13.

Keel C, Weller DM, Natsch A, Defago G, Cook RJ & Thomashow

LS (1996) Conservation of the 2,4-diacetylphloroglucinol

biosynthesis locus among fluorescent Pseudomonas strains

from diverse geographic locations. Appl Environ Microbiol 62:

552–563.

King E, Ward M & Raney D (1954) Two simple media for the

demonstration of pyocyanin and fluorescin. J Lab Clin Med 44:

301–307.

Laville J, Blumer C, Von Schroetter C, Gaia V, Defago G, Keel C &

Haas D (1998) Characterization of the hcnABC gene cluster

encoding hydrogen cyanide synthase and anaerobic regulation

by ARN in the strictly aerobic biocontrol agent Pseudomonas

fluorescens CHA0. J Bacteriol 180: 3187–3196.

Mazzola M, Fujimoto DK, Thomashow LS & Cook RJ (1995)

Variation in sensitivity of Gaeumannomyces graminis to

antibiotics produced by fluorescent Pseudomonas spp. and

effect on biological control of take-all of wheat. Appl Environ

Microbiol 61: 2554–2559.

Milling A, Smalla K, Maidl F, Schloter M & Munch J (2005)

Effects of transgenic potatoes with an altered starch

composition on the diversity of soil and rhizosphere bacteria

and fungi. Plant Soil 266: 23–39.

Notz R, Maurhofer M, Schnider-Keel U, Duffy B, Haas D &

Defago G (2001) Biotic factors affecting expression of the 2,4-

diacetylphloroglucinol biosynthesis gene phlA in Pseudomonas

fluorescens biocontrol strain CHA0 in the rhizosphere.

Phytopathology 91: 873–881.

Pfaffl MW (2001) A new mathematical model for relative

quantification in real-time RT-PCR. Nucleic Acids Res 29:

2002–2007.

Purohit HJ, Raje DV & Kapley A (2003) Identification of

signature and primers specific to genus Pseudomonas using

mismatched patterns of 16S rDNA sequences. BMC

Bioinformatics 4: 19.

Raaijmakers JM & Weller DM (1998) Natural plant protection by

2,4-diacetylphloroglucinol-producing Pseudomonas spp. in

take-all decline soils. Mol Plant Microbe In 11: 1–2.

Raaijmakers JM, Weller DM & Thomashow LS (1997)

Frequency of antibiotic producing Pseudomonas spp.

in natural environments. Appl Environ Microbiol 63:

881–887.

Ramette A, Frapolli M, Defago G & Moenne-Loccoz Y (2003)

Phylogeny of HCN synthase-encoding hcnBC genes in

biocontrol fluorescent Pseudomonas and its relationship with

host plant species and HCN synthesis ability. Mol Plant

Microbe In 16: 525–535.

Rezzonico F, Zala M, Keel C, Duffy B, Moenne-Loccoz Y &

Defago G (2007) Is the ability of biocontrol fluorescent

pseudomonads to produce the antifungal metabolite 2,4-

diacetylphloroglucinol really synonymous with higher plant

protection? New Phytol 173: 861–872.

Schnider-Keel U, Seematter A, Maurhofer M et al. (2000)

Autoinduction of 2,4-diacetylphloroglucinol biosynthesis in

the biocontrol agent Pseudomonas fluorescens CHA0 and

repression by the bacterial metabolites salicylate and

pyoluteorin. J Bacteriol 182: 1215–1225.

Seveno NA, Morgan JA & Wellington EM (2001) Growth of

Pseudomonas aureofaciens PGS12 and the dynamics of HHL

FEMS Microbiol Ecol 68 (2009) 212–222 c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

221Transcriptional activity of phlD and hcnBC using qRT-PCR

and phenazine production in liquid culture, on nutrient agar,

and on plant roots. Microbial Ecol 41: 314–324.

Shanahan P, O’Sullivan DJ, Simpson P, Glennon JD & O’Gara F

(1992) Isolation of 2,4-diacetylphloroglucinol from a

fluorescent pseudomonad and investigation of physiological

parameters influencing its production. Appl Environ Microbiol

58: 353–358.

Sharifi-Tehrani A, Zala M, Moenne-Loccoz Y, Natsch A & Defago

G (1998) Biocontrol of soil-borne fungal plant diseases by 2,4-

diacetylphloroglucinol-producing fluorescent pseudomonads

with different restriction profiles of amplified 16S rDNA. Eur J

Plant Pathol 104: 631–643.

Siddiqui M & Shaukat S (2003) Suppression of root-knot disease

by Pseudomonas fluorescens CHA0 in tomato: importance of

bacterial secondary metabolite, 2,4-diacetylpholoroglucinol.

Soil Biol Biochem 35: 1615–1623.

Voisard C, Keel C, Haas D & Defago G (1989) Cyanide

production by Pseudomonas fluorescens helps suppress black

root rot of tobacco under gnotobiotic conditions. EMBO J 8:

351–358.

Wei H-L & Zhang L-Q (2006) Quorum-sensing system influences

root colonization and biological control ability in

Pseudomonas fluorescens 2P24. Antonie van Leeuwenhoek 89:

267–280.

FEMS Microbiol Ecol 68 (2009) 212–222c� 2009 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

222 M.M. Paulin et al.