M

Pcts

A

DN

R

0d

icrobiological Research 166 (2011) 323—335

Available online at www.sciencedirect.com

www.elsevier.de/micres

henazine-1-carboxylic acid is a more importantontributor to biocontrol Fusarium oxysporumhan pyrrolnitrin in Pseudomonas fluorescenstrain Psd

shutosh Upadhyay, Sheela Srivastava ∗

epartment of Genetics, University of Delhi South Campus, Benito Juarez Road,ew Delhi 110021, India

eceived 31 December 2009 ; received in revised form 22 June 2010; accepted 28 June 2010

KEYWORDSAntibiotics;Gene knockout;Homologousrecombination;HPLC/APCI-MS;Targetron

SummaryPhenazines and pyrrolnitrin (Prn) are broad spectrum antibiotics, produced by bac-teria, more so by the biocontrol strains to kill the phytopathogens in soil. We havestudied a rhizospheric soil isolate of Pseudomonas fluorescens strain Psd producingboth phenazine-1-carboxylic acid (PCA) and Prn. In order to study the contributionof these antibiotics, the phzD and prnC genes involved in PCA and Prn biosynthe-sis, were disrupted in a site-specific manner using a group II intron-based Targetrongene-knockout system, and gene disruption followed by allelic exchange throughhomologous recombination, respectively. The resulting knockout strains Psdphz122s-34 and PsdprnC::gen did not produce PCA and Prn, respectively. In fact, by combiningthese two strategies, a Psdphz122s-34prnC::gen double mutant could also be gener-ated. Identification and lack of PCA production was corroborated by HPLC/APCI-MSanalysis, and TLC detection for both the antibiotics in these mutants. Loss of antifun-gal activity against the phytopathogenic fungus Fusarium oxysporum was observedusing in vitro growth assays on plates or growth chamber experiments with tomato

seedling on an artificial substrate. Based on the characterization of these geneknockout mutants, we propose that PCA and Prn have a major role in antifungalactivity of strain Psd.© 2010 Elsevier GmbH. All rights reserved.

∗Corresponding author. Tel.: +91 11 24110690; fax: +91 11 2411276E-mail address: srivastava sheela@yahoo.com (S. Srivastava).

944-5013/$ – see front matter © 2010 Elsevier GmbH. All rights reseoi:10.1016/j.micres.2010.06.001

1.

rved.

ane2wdaoebeb1acWptfihr1

Paowdb

bsa2icasmasaoto

tcrwmt

324

Introduction

Many fluorescent and non-fluorescent strains of thegenus Pseudomonas are known to produce severalsecondary metabolites, such as phenazines, pyrrol-nitrin, pyoleuteorin, and 2,4-diacetyl phloroglu-cinol. These secondary metabolites or antibioticshave been implicated in plant disease controlenabling the producing strain to serve as a bio-control agent (Mavrodi et al. 1998; Maddula etal. 2008; Costa et al. 2009; Selin et al. 2010).Phenazines are heterocyclic nitrogen-containingsecondary metabolites synthesized by Pseudomonasfluorescens and a few other bacterial genera (Price-Whelan et al. 2006). Biocontrol by phenazinesis connected with their ability to undergooxidation—reduction transformations thus causingthe accumulation of toxic superoxide radicals inthe target cells (Kerr 2000; Laursen and Nielsen2004; Price-Whelan et al. 2006) and pyrrolnitrin hasbeen described as an inhibitor of fungal respiratorychains (Tripathi and Gottlieb 1969). The response ofthe biocontrol strains against phytopathogens mayvary depending upon the types and combination ofthese secondary metabolites produced.

A number of naturally occurring, broad spec-trum, colored phenazines have been reported indifferent studies. P. fluorescens 2-79 is among thefirst few strains from which purified phenazinecompounds were shown to have antifungalactivity (Gurusiddaiah et al. 1986). Differentphenazine derivatives originate either fromphenazine-1-carboxylic acid (PCA) or phenazine-1,6-dicarboxylic acid (PDC) (Leisinger and Margraff1979; Kerr 2000). Genes encoding the phenazinebiosynthetic enzymes are arranged in one coreoperon, phzABCDEFG, in most producer Pseu-domonads, including P. chlororaphis (aureofaciens)30-84 (Pierson and Thomashow 1992; Delaney etal. 2001). The only exception is P. aeruginosa thatcontains two copies of the phenazine operon, withboth the copies showing homology to similar locifrom P. fluorescens, and P. chlororaphis (aureofa-ciens) (Mavrodi et al. 1998, 2006; Delaney et al.2001). Besides these, the genes, phzI and phzR,encode a quorum-sensing circuit that regulatesphenazine production in P. chlororaphis (aureo-faciens) 30-84 (Pierson and Thomashow 1992).In many strains, additional genes such as phzM,phzS, phzO and phzH are involved in phenazinedecoration (Chin-A-Woeng et al. 2001; Mavrodi etal. 2001). The phzD gene is known to code for an

essential enzyme in the phenazine biosyntheticpathway in different bacteria.

Pyrrolnitrin (Prn) is a tryptophan-derived,organohalogenic, antifungal secondary metabolite,

otwt

A. Upadhyay, S. Srivastava

nd is classically known to be produced by aarrow range of Gram-negative bacteria (Hammert al. 1999; Haas and Defago 2005; Costa et al.009). Production of Prn has been correlatedith the ability of some bacteria to control plantiseases caused by fungal pathogens and bears

potential importance in the suppressivenessf the pathogen (Haas and Defago 2005; Costat al. 2009). The pyrrolnitrin (prn) operon haseen completely sequenced; prnABCD genes whichncode different enzymes of the pathway haveeen identified (Hill et al. 1994; Hammer et al.999). Though, the nucleotide and predictedmino acid sequences of the prnABCD genes areonserved among the limited strains available.hen each gene was assessed separately, prnA and

rnC were found to be more strongly conservedhan prnB and prnD (Costa et al. 2009). While therst two genes code for enzymes involved in thealogenation reactions, the latter are required foring rearrangement and oxidation (Kirner et al.998).

We have earlier described the phenazine andrn producing P. fluorescens strain Psd (Upadhyaynd Srivastava 2008). In order to delineate the rolef these antibiotics in overall biocontrol function,e attempted to identify a key pathway gene andisrupted the same to find out its important contri-ution.

Mobile group II introns (“Targetrons”) haveeen used for targeted gene disruption and site-pecific DNA insertion in diverse Gram-negativend Gram-positive bacteria (Yao and Lambowitz007; Malhotra and Srivastava 2008). Group IIntrons are useful for gene targeting because theyan be programmed for insertion into virtuallyny desired DNA target with high frequency andpecificity through retrohoming with efficienciesuch higher than ectopic integration (Karberg et

l. 2001). Such insertions can be selected sub-equently by splice-activated markers (Zhong etl. 2003). In this paper, we report the effectf targeted insertion of a group II intron in tohe phzD gene of the phenazine biosyntheticperon.

For disruption of the Prn biosynthetic operon,he strategy of insertion of an antibiotic-resistanceassette (Gmr) in prnC, combined with the alleliceplacement through homologous recombinationas followed. Subsequently, a double knockoututant of both the phenazine and Prn biosyn-

hetic pathway was generated. Antifungal activity

f the wild-type and mutant strains towards a phy-opathogen, Fusarium oxysporum was correlatedith the production of PCA and Prn. The effect of

hese strains on biocontrol of this phytopathogen

P rum

wp

M

C

Pb2t(sw

oTasctpRP

D

T

S

E

P

CA contributes for biocontrolling Fusarium oxyspo

as also tested using tomato seedlings as the modellant system.

aterials and methods

ulture conditions

. fluorescens strain Psd used in this study haseen described earlier (Upadhyay and Srivastava008). The strain was maintained on the nutri-

ionally defined Standard Succinate Medium, SSMMeyer and Abdallah 1978). For DNA extraction,train Psd was grown in SSM, and Escherichia colias grown in LB medium with appropriate antibi-

p

Tb

able 1. Bacterial strains and plasmids used in this study.

trains/plasmids Relevant characteristics

scherichia coliE. coli XL1Blue supE44 hsdR17 (rk

− mk+) re

gyrA96 relA1 lac− F’[proATn10 (Tcr)

pME6000 ∼7.2 kb cloning vector, pBTcr PLac, maintained in E. c

pACD4K-C 7.7 kb intron expression veexpression from T7 promotkanr as the marker activatretrotransposition (RAM) in

pACDphz122s ∼8 kb pACD4K-C derivative∼350 bp mutated intron spposition 122s of phzD, T7 pconstitutive Pspc, Cmr, in

pMEphz122s ∼7.2 kb pME6000 derivativClaI—NcoI fragment from pinto the ClaI—NcoI sites ofall components necessarydriven by Pspc, Tcr, in E. c

pBKS+ Cloning vector; ColE1 replipBKS-kan Cloning vector; ColE1 repli

to carry ∼900 bp Kmr genepOT2 5.3 kb plasmid containing 5

seudomonas fluorescens2-79 Standard phenazine (PCA)Pf-5 Standard strain for differePsd Wild-type strain, isolated f

roots, Apr

Psdphz122s-34 P. fluorescens Psd derivativmutated intron inserted atphzD gene, Kmr

PsdprnC::gen P. fluorescens strain Psd demutated prnC gene with agene

Psdphz122s-34prnC::gen P. fluorescens strain Psd doderivative carrying the muat position 122s in phzD ancassette inserted in prnC,

325

tics (Ap, Km each at 50 �g/ml and Gm 10 �g/ml).he pigment production medium (PPM, Rosales etl. 1995) was used to recover antibiotics from thepent culture filtrate of strain Psd. All solid mediaontained 15 g l−1 agar. Different strains used inhis study are listed in Table 1. Fungus, F. oxys-orum was obtained from the Indian Agriculturalesearch Institute, New Delhi, and maintained onDA medium at 28 ◦C.

NA extraction and PCR amplification of

hzCD and prnC genes

otal genomic DNA from strain Psd was extracted,y the cetyltrimethylammonium bromide (CTAB)

Reference

cA1 endA1 thi-1B+] lacIq lac Z DM15

Sambrook and Russell(2001)

BR1MCS derivative;oli DH5�

Prof. D. Haas, Universitede Lausanne, Switzerland

ctor drivinger, Cmr, carrying

ed byE. coli XL1Blue

Sigma—Aldrich

, carrying theecific to insertion atromoter replaced by

E. coli XL1Blue

This study

e carrying ∼5.5 kbACDphz122s insertedpME6000, containing

for intron expressionoli DH5�

This study

con; Apr Stratagenecon; Apr, modified Malhotra (2007), Ph.D.

Thesis38 bp Gmr gene Lab Stock

producer strain USDA-ARSnt PGPR traits USDA-ARSrom Vigna mungo Upadhyay and Srivastava

2008e carrying theposition 122s of the

This study

rivative carrying then insertion of Gmr

This study

uble mutanttated intron insertedd gentamicin geneKmr, Gmr

This study

326 A. Upadhyay, S. Srivastava

Table 2. The primers used for targeted disruption of P. fluorescens strain Psd phzD gene.

Primers Sequence (5′—3′) Restriction site

IBSphz AAAAAAGCTTATAATTATCCTTAGATCACCACCCCGTGCGCCCAGATAGGGTG

HindIII

EBS-1dphz CAGATTGTACAAATGTGGTGATAACAGATAAGTCCACCCCGATAACTTACCTTTCTTTGT

BsrGI

EBSphz TGAACGCAAGTTTCTAATTTCGATTTGATCTCGATAGAGGAAAGTGTCT

—

ACCAAA

iwgcttafie

Bd

PirmCp

psRStptPfiPdaPPsaias

Pspc AAAAAAGCTTCCGTTTATTTTTTCTTGTTATAATGCCGCGCCCTCGATAA

protocol (Ausubel et al. 1992). PCR amplificationwas carried out with the phzCD (Raaijmakers et al.1997) and prnC primers, designed from sequencedata of Pf-5 genome (Paulsen et al. 2005; Upadhyayand Srivastava 2010).

Gene targeting in strain Psd

(i) Gene targeting using TargetronThe Targetron gene knockout system

(Sigma—Aldrich) was used for knocking outthe phzD gene in the phenazine biosyntheticoperon. The group II intron’s potential inser-tion sites in the phzD gene were first identifiedby the Sigma—Aldrich computer algorithm.Three primers were then designed to re-target(mutate) the intron to the selected insertionsite (Table 2). The primers, IBSphz, EBS1dphzand EBS2phz were employed to mutate the RNAportion of the intron by a primary PCR. Theamplicon so obtained, was used as a templatein a secondary PCR with the spc promoter,Pspc (constitutive ribosomal protein operonpromoter synthesized on the basis of the E. colisequence information) and EBS1dphz primersso as to replace the T7 promoter in the vectorpACD4K-C. The HindIII—BsrGI digested PCRproduct was ligated to linearized pACD4K-Cyielding pACDphz122s.

(ii) Antibiotic cassette mediated gene disruption byhomologous recombination

Antibiotic cassette mediated site-specificdisruption of the prnC gene was performedand knockouts were screened for homologousrecombination events. Herein, a series of con-structs were created to incorporate the Gmr

(from pOT2) into prnC carried on a vector pBKS+that is not maintained stably in strain Psd.

Construction of pBKS-kan-prnC::gen

The prnC gene first cloned in pBKS was relo-cated in the vector pBKS-kan constructed earlier

sfiic

CATATCCTTGAAGCGGGAGCTTATAATTATCCTTA

HindIII

n the lab. PCR amplified Gmr amplicon (∼500 bp)as ligated in to the unique site of ZraI in prnCene, for which appropriate enzyme site wasreated on the primers as well. The ligated mix-ure was transformed into E. coli XL1Blue andhe transformants were selected on kanamycinnd gentamicin. Transformants were further con-rmed by plasmid isolation and restriction with ZraInzyme.

acterial strain transformation andetection of targeted integration

lasmid containing the phzD specific mutatedntron (pACDphz122s) was put on a broad hostange vector, pME6000. A ∼5.5 kb ClaI—NcoI frag-ent from pACDphz122s was ligated with ∼6.7 kblaI—NcoI digested pME6000, resulting in a ∼12.2 kblasmid, pMEphz122s.

The recombinant plasmids, pMEphz122s andBKS-kan-prnC::gen were electroporated in totrain Psd with the Gene Pulser XcellTM (Bio-ad, USA) as described earlier (Malhotra andrivastava 2006) and screening was done forhe retrotransposition-activated marker, Kmr forMEphz122s and Gmr for pBKS-kan-prnC::genransformants, respectively. The Kmr variant ofsd, strain Psdphz122s-34 so obtained was con-rmed for the presence of the intron byCR with IBSphz/EBS2phz primers and phzDisruption by phzCDF/phzCDR primers (intronnd internal phzCD specific primers). Further,CR was performed from the genomic DNA ofsdphz122s-34 and PsdprnC::gen strains usingpecific primers for phzCD and prnC (forwardnd reverse). With both the strategies work-ng successfully, we also went on to generate

double mutant by transforming the mutanttrain Psdphz122s-34 with pBKS-kan-prnC::gen and

electing for Kmr, Gmr transformants. Once con-rmed, the strains were grown in PPM and

ndividual colonies were screened to visualize anyhange in pigment production associated with PCA

P rum

aa

Eo

Ead

Ho

MaCsi3bmrwbs1ism40aT0(7wrrarcXri(p(1

Bk

I3

watda2wccwiFs3fct

P

IifmtSwstwst(wplP8ed

R

S

PatSa

CA contributes for biocontrolling Fusarium oxyspo

nd subsequent confirmation by TLC and HPLC/MSnalysis.

xtraction and chromatographic detectionf antibiotics

xtraction and detection of phenazine and Prnntibiotics on TLC was performed by the methodsescribed earlier (Upadhyay and Srivastava 2008).

PLC-coupled mass spectrometry analysisf phenazine

ass spectrometric reactions were analyzed withn LTQ mass spectrometer (Thermo Electronorporation) coupled to an Agilent 1100 HPLCystem (Reverse phase). Reaction mixtures contain-ng different concentrations of trans-2,3-dihydro--hydroxyanthranilic acid, DHHA (0.1—5 mmol)uffered in 50 mmol Tris—HCl pH 7.5 and supple-ented with 10 �mol caffeine as internal inert

eference were used. The extracted phenazineas applied to an HPLC system equipped with ainary pump, a diode array detector and a thermo-tatic autosampler. The autosampler was kept at0 ◦C and the reaction mixture was thermostatedn the autosampler before the experiment wastarted. One microlitre sample was injected at 11-in interval and resolved on a dC18 column (3 �m,

.6 mm × 100 mm, Waters Atlantis). Solvent A was

.1% formic acid in water and solvent B, 0.1% formiccid in acetonitrile at a flow rate of 1 ml min−1.he column was developed with the gradient:—2.2 min, 100%A (0%B); 2.2—2.25 min, 100—70%A0—30%B); 2.25—7.9 min, 70—30% A (30—70%B);.9—10 min, 100%A (0%B). Elution was followedith the PDA detector monitoring the wavelength

ange of 200—600 nm. Mass spectroscopy was car-ied out in the positive ion detection mode usingtmospheric pressure chemical ionization (APCI),ecording the mass range from 50—600 m/z in theentroid mode. Mass spectra were analyzed withcalibur software (Thermo Electron Corporation) byeferencing the peak area to that of caffeine. Thedentity of phenazine and phenazine carboxylic acidPCA) was confirmed by spiking the column withhenazine from Alfa Aesar and PCA from InFarmatikGermany) by an established procedure (Flood et al.972).

iocontrol activities of strain Psd and its

nockout strains

n vitro antifungal properties of strains Psdphz122s-4, PsdprnC::gen and Phzphz122s-34prnC::gen

p2s(

327

ere evaluated quantitatively against F. oxysporumnd compared with that of wild-type strain Psd. Forhis, the fungal culture was raised in PD mediumiluted (to 50%) with bacterial culture filtrate,s standardized earlier (Upadhyay and Srivastava010). Inhibition was observed in terms of dryeight of the biomass after 48 h of growth andompared with the untreated control. Untreatedontrol was also raised in 50% diluted PD mediumith the bacterial medium, SSM. Antifungal activ-

ty was also checked by a dual-culture test.or this, the wild-type strain Psd and knockouttrains, Psdphz122s-34, PsdprnC::gen, and Psd122s-4prnC::gen were patched on PDA, with the targetungus, F. oxysporum mycelial plug placed in theenter. Growth was monitored after 48 h of incuba-ion at 28 ◦C.

lant test to monitor biocontrol activity

n order to assess the plant bioprotection activ-ty of the wild-type and the generated strains, aunctional test was carried out using tomato asodel host system and F. oxysporum as the phy-

opathogen, as described earlier (Upadhyay andrivastava 2010). Tomato seedlings (4 leaf stage)ere planted in the culture tubes containing 15 g

oil mixed with three parts of maize flour. The twoest sets, each in triplicate contained plants thatere exposed to ∼108 spores ml−1 of F. oxysporum

uspended either in culture extract of the wild-ype Psd and knockout strains or in normal salinefor control set). Another set contained plantsithout any fungal spores (uninfected control). Alllants were exposed to fungal spores by inocu-ating the suspension around the roots directly.lants were maintained at 28 ◦C with 16 h light andh dark period, watered regularly with a nutri-nt solution and growth was monitored for 10—15ays.

esults

train characteristics

. fluorescens strain Psd produces two majorntibiotics, phenazine and Prn which were iden-ified by their resolution on TLC (Upadhyay andrivastava 2008). Genetic evidence for thesentibiotics was obtained by PCR amplication of

hzCD and prnC genes (Upadhyay and Srivastava010). Partial sequence data of phz ampliconuggested that it contained 423 bp from phzCGenBank EF636691) and 458 bp from phzD (Gen-

faffrpotP(i

poBs

Hb

HfcroP

328

Bank EF636690). The other amplicon carriedthe full sequence of 1482 bp of prnC (GenBankEU491519).

Disruption of the resident phzD gene usingTargetron

The Sigma—Aldrich algorithm identified position122 on the sense strand of phzD gene as an intron-insertion target. Primary PCR using primers, IBSphz,EBS1dphz and EBS2phz mutated the RNA portionof the intron and gave an amplicon of 350 bp. TheT7 promoter in pACD4K-C vector was replaced bythe spc promoter as described in Experimental pro-cedures. The HindIII—BsrGI digested PCR productwhen ligated to linearized vector yielded pACD-phz122s.

The two plasmids, pACDphz122s and the broadhost range vector, pME6000 on double digestionwith ClaI—NcoI yielded ∼5.5 kb and ∼6.7 kb bands,respectively, which were then ligated together. Theresulting ligated product of 12.2 kb carried thephzD specific mutated intron and was labeled aspMEphz122s. The phzD gene was ultimately dis-rupted in strain Psd by electroporation of therecombinant plasmid, pMEphz122s and screening

r

for Km transformants generated by retrotransposi-tion. The Kmr variant of strain Psd, Psdphz122s-34was confirmed for the presence of the intronby PCR with IBSphz/EBS2phz primers. As is clear

(s3T

Figure 1. (a) PCR amplification of the ∼350 bp fragment frostrain Psdphz122s-34 (Lane 1) and the control plasmid, pACDpnot observed from wild-type P. fluorescens strain Psd (Lane 2)that yielded a ∼1.4 kb fragment from P. fluorescens strain Psof the intron and the phzCD coding region, confirming the insDNA size marker in each gel).

A. Upadhyay, S. Srivastava

rom Fig. 1a, this primer set yielded a 350 bpmplicon (same as that obtained for pACDphz122s)rom the variant strain Psdphz122s-34 but notrom the wild-type strain Psd. Similarly, phzD dis-uption was confirmed by primers phzCDF andhzCDR which yielded a ∼1.8 kb fragment (1.4 kbf phzCD + 450 bp from the inserted intron con-aining Pspc). In comparison, the wild-type strainsd showed a band of ∼ 1.4 kb size, as expectedFig. 1b). This indicated that the intron was indeednserted into the phzD gene.

A characteristic feature associated withhenazine production in strain Psd is the releasef a dark red-color pigment, when grown in PPM.ased on this phenotype, the knockout straincreened showed no color.

PLC—mass analyses of phenazine producedy the wild-type and knockout strain

PLC—APCI/MS analysis of extracted phenazinerom strain Psd clearly identified it as phenazine-1-arboxylic acid (PCA), which eluted at two differentetention times i.e. 8.2 and 8.5 min (Fig. 2a). Onef these peaks corresponded to fully aromatizedCA (RT 8.2) whereas the other was reduced PCA

RT 8.5), based on the standards. These two peakshowed typical absorption maxima of phenazine at72 nm. Mass spectra showed a mass of 225 Da.he phenazine knockout strain on the other hand

m the group II intron with IBSphz/EBS2phz primers, fromhz122s (carrying the intron, Lane 3) while the same was. (b) PCR amplification with phzCDF and phzCDR primersd (Lane 1). Lane 2 showed ∼1.8 kb band carrying a partertion of the intron within phzD gene (Lane L represents

P rum

scMops

Dh

Iics

ttwnrfaalti

FiIscm

CA contributes for biocontrolling Fusarium oxyspo

howed a peak at 8.4 min (Fig. 2b), that did notorrespond to the phenazine absorption spectrum.ass spectra showed two atypical peaks with a massf 206 and 246 Da. The chemical nature of theseeaks is currently not known and could only be con-idered as an artifact.

isruption of the resident prnC gene byomologous recombination

n order to disrupt prnC gene, the prnC clonedn pBKS-kan was modified by the insertion of Gmr

assette. This plasmid was electrotransformed intotrain Psd and tranformants were selected on gen-

bwtt

igure 2. HPLC/APCI-MS analysis of extracted phenazine fnjected on to HPLC—MS system and separated with the gradn addition to the chromatogram (i), the MS (ii) and UV/vis sptrain Psd and (b) for Psdphz122s-34. The identity of phenazolumn with PCA standards. Major peak at 8.4 min in strainaxima (372 nm).

329

amicin supplemented medium. For checking forhe stable insertion/presence of Gmr, few coloniesere transferred from the selection medium toon-selective conditions (without gentamicin) andaised for a number of generations. When trans-erred back on to gentamicin, they still retained thebility to resist gentamicin. As pBKS-kan-prnC::gen,pBKS+ derivative, has a narrow host range it is not

ikely to be maintained in Psd thereby suggestinghat the Gmr-cassette carrying prnC gene may havenserted into the Psd genome by homologous recom-

ination. The mutant PsdprnC::gen, so obtainedas confirmed for kanamycin sensitive status as

he transformed plasmid having kanamycin resis-ance gene, after homologous recombination must

rom strain Psd and Psdphz122s-34. One �l sample wasient described in the Experimental procedures section.ectra (iii) are also shown. Panel (a) depicts analysis forine carboxylic acid (PCA) was confirmed by spiking thePsdphz122s-34 did not match with the PCA absorption

330 A. Upadhyay, S. Srivastava

(Con

C

TgocEsrbcPf

Figure 2.

have been eliminated (segregated out). To con-firm this further, PCR was performed from thegenomic DNA of PsdprnC::gen strain using primersprnC (F) and prnC (R). This yielded one ∼2 kbamplicon from PsdprnC::gen, whereas wild typeshowed ∼1.5 kb product for prnC gene as expected(Fig. 3).

Generation of double mutant strainPsdphz122s-34prnC::gen

The recombinant plasmid pBKS-kan-prnC::gen waselectrotransformed in to phenazine knockout

mutant strain, Psdphz122s-34 and Kmr, Gmr

transformants were characterized for the pro-duction of these two antibiotics as describedbelow.

3eFt

tinued ).

haracterization of knockout strains

he antifungal activity of the mutant strains soenerated was probed by growing the fungus, F.xysporum in PD medium mixed with bacterialulture filtrate to the level of 50% as described inxperimental procedures. The spent culture filtratehowed growth inhibitory effects as reflected byeduced biomass generation and dry weight of theiomass after 48 h of growth, when compared toontrol. By this comparison, the strains Psd122s-34,sdprnC::gen, and Psdphz122s-34prnC::gen wereound to show a loss in antifungal activity by 52,

2 and 85%, respectively, as compared to the par-nt strain Psd (Fig. 4a). A dual-culture assay with. oxysporum and all the mutants also supportedhe fact that the Psdphz122s-34 strain is less effec-

PCA contributes for biocontrolling Fusarium oxysporum

Figure 3. PCR confirmation for the gentamicin resis-tance gene insertion in prnC gene, Lane 1 depicts PCRamplification (∼1.5 kb) from wild-type strain Psd. Lane2 shows ∼1.9 kb band with ∼500 bp of gentamicin resis-t1

ttwsfsTp

swtd

P

TFebcPiskpposkt

D

Prv2010). Different strains of P. fluorescens have

FoPdPdi

ance gene inserted in PsdprnC::gen. Lane L representskb DNA size markers.

ive in inhibiting the hyphal growth when comparedo the Prn knockout strain PsdprnC::gen and theild-type strain. In comparison, the double mutant

train was almost ineffective in controlling theungal growth (Fig. 4b). The antibiotic production

tatus of knockout mutants was also confirmed byLC. Extracted PCA and Prn developed on the TLClate showed the characteristic yellow fluorescent

bcb

igure 4. (a) Dry weight of the biomass of Fusarium raised if Psdphz122s-34, PsdprnC::gen, and Psdphz122s-34::gen. Tresd was taken as 100% inhibition. Control represents no inhibitiluted with 50% SSM. Dry weight of biomass was used to com. fluorescens strain Psd (i), Prn knockout mutant PsdprnC::gouble mutant for both the antibiotics Psd122s-34prnC::gen (in the center and test strains were placed around it at equal

331

pot for phenazine and purple spot for Prn in theild-type strain. The two knockout strains showed

he absence of the respective antibiotics and theouble mutant lacked both of them (Fig. 5).

lant bioprotection test

he response of tomato seedlings to phytopathogen. oxysporum in the presence of bacterial culturextracts suggested the role of these antibiotics iniocontrol activities. The results presented in Fig. 6learly showed that the culture extract of strainsd could protect the plants against F. oxysporumnfection, but the same was lacking with the mutanttrains more so in PCA knockout than pyrrolnitrinnockout strain. Set containing fungal spores sus-ended in normal saline showed typical damping offhenotype in the plants. These results confirmedur in vitro tests against F. oxysporum and demon-trated that the loss of antibiotic production by thenockout strains led to their inability in suppressinghe pathogen.

iscussion

. fluorescens strain Psd is a plant growth promotinghizobacterium possessing several activities rele-ant to biological control (Upadhyay and Srivastava

een reported to produce different types andombinations of the antibiotics implicated iniocontrol (Haas and Defago 2005; Maddula et

n PD medium diluted to 50% with the culture supernatantatment with the culture supernatant of wild-type strainion in terms of growth of the fungus raised in PD mediumpare the antifungal activity. (b) Dual-culture assay withen (ii), phenazine knockout mutant Psdphz122s-34 (iii),v), and the fungus F. oxysporum. Mycelial plug was placeddistances. Plate was incubated at 28 ◦C for 48 h.

332 A. Upadhyay, S. Srivastava

Figure 5. TLC profile for antibiotics production (phenazine and Prn) from knockout mutant strains Psdphz122s-34(Lane 2), PsdprnC::gen (Lane 3), and Psdphz122s-34prnC::gen (Lane 4) along with wild-type strain Psd (Lane 5).

-5 (Ly.

ee1k

Extracts from standard P. fluorescens strains 2-79 and Pffor phenazine and Prn antibiotics production, respectivel

al. 2008; Selin et al. 2010). The potential ofstrain Psd is based on the unique combinationof phenazine and Prn, it produces and releasesin the culture medium (Upadhyay and Srivastava2008). We derived the contribution of these antibi-

otics towards biocontrol function by attempting toknock out one of the important genes each fromthe respective biosynthetic operons both singly ortogether.

stPc

Figure 6. Effect of culture extract of Psd and its mutant strrum, (a) control plant grown in the absence of fungal strain,knockout strains, (c) infected control plant showing dampingstrain Psd; set 2, strain PsdprnC::gen, and set 3 represents th

ane 1 in both the cases) were taken as positive control

The technique of targeted insertional mutagen-sis is a well established and useful tool for geneticngineering of different bacterial species (Alexeyer999; Perutka et al. 2004). The Targetron genenockout system (Sigma—Aldrich) that allows the

ite-specific disruption of bacterial genes by inser-ion of a group II intron, proved handy in strainsd. We have earlier used a similar strategy forreation of an ipdC gene knockout in Azospirillum

ains in plant protection against the pathogen, F. oxyspo-(b) fungal spores mixed with culture extract of Psd andoff. Note: spore count in set b and c was 108/ml, set 1,e response to strain Psdphz122s-34.

P rum

bStbhgacbacmerudIdstei

ksrPtpptwbg1

awitsawPuoagtdcFoaPi

dtssTtoaIswdebmprvadpacsPA

eigofvptp

A

AWlisfiT

R

CA contributes for biocontrolling Fusarium oxyspo

rasilense SM, a PGP rhizobacterium (Malhotra andrivastava 2008). The present study demonstratedhat this method could be employed as a broad-ased strategy in different types of bacteria. Itas been described in earlier studies that the Tar-etron expressed via an inducible promoter frombroad host range vectors can be used for effi-

ient gene targeting in a variety of Gram-negativeacteria (Yao and Lambowitz 2007). We have usedbroad host range vector with promoter modifi-

ation for efficient expression in strain Psd. Thisodification has been necessitated based on our

arlier observation that the plasmid pACD4K-C car-ying the components of the group II intron placednder the T7 promoter could not be used directly inifferent bacteria (Malhotra and Srivastava 2008).n the present study, P. fluorescens strain Psd alsoid not seem to express T7 polymerase (data nothown). The constitutive E. coli spc ribosomal pro-ein operon promoter Pspc, used here was foundffective for efficient intergration of group II intronn the desired gene.

That the strategy of generating a phenazinenockout mutation was successful, was demon-trated by the dual-culture assay and lack ofelease of pigmented compound by mutant strain,sdphz122s-34. The analysis of the PCA produc-ion by HPLC coupled with the online atmosphericressure chemical ionization-mass spectrometryrofile suggested the lack of phenazine produc-ion by the mutant strain in comparison to theild-type strain. The HPLC/APCI-MS method haseen used routinely in chemical analysis and offersood linearity and reproducibility (Combs et al.999).

To gain insight into the contribution of Prnntibiotic, disruption of its biosynthetic genesas carried out in strain Psd. This comprised the

nsertion of an antibiotic-resistance cassette inhe desired gene, followed by allelic replacementtrategy through homologous recombination. Suchstrategy has been employed in many bacteria andas also successful in getting a prnC null mutant,sdprnC::gen of strain Psd. Such a mutant enableds to evaluate the share of this antibiotic in theverall biocontrol function against Fusarium. Alsonull mutant of both the antibiotic biosynthetic

enes could also be generated by combininghe two strategies. With these mutants, weemonstrated the contribution of PCA and Prn inonferring the biocontrol of damping off pathogen. oxysporum through an in vitro assessment. Based

n our results, it can be suggested that both thentibiotics are required for antifungal action, butCA has a major contribution in comparison to Prnn P. fluorescens strain Psd.

A

333

Despite the inhibitory activity of antibiotics,efinite evidence for their importance in biocon-rol is lacking mainly because the production initu could not be directly correlated with diseaseuppression under biologically relevant conditions.he most common diseases associated with phy-opathogenic fungi are take all disease, dampingff and wilts (Maddula et al. 2008; Couillerot etl. 2009; Mazurier et al. 2009; Selin et al. 2010).n order to fulfil this requirement, we have demon-trated the biocontrol activity against F. oxysporumith tomato as the host system. That the antibiotic-eficient mutants are compromised in this functionstablished that these antibiotics are involved iniocontrol. Studies involving antibiotic-deficientutants have clearly demonstrated that antibioticroduction in natural habitats play an importantole in ecological competence and long term sur-ival of strains in the environment (Mazzola etl. 1992; Selin et al. 2010). It has been amplyocumented that the different biocontrol strainsroduce different types and combinations of thesentibiotics. Their relative contribution towards bio-ontrol and their effectivity range also vary fromtrain to strain (Thomashow and Weller 1988;ierson and Thomashow 1992; Hill et al. 1994; Chin--Woeng et al. 2001; Selin et al. 2010).

We have studied P. fluorescens strain Psd andstablished the genetic basis of biocontrol activ-ty by successfully knocking out the phzD and prnCenes of the phenazine and Prn biosynthetic oper-ns, respectively. The altered phenotype and theunctional differences of the knockout strains initro as well as with a plant system in soil, against ahytopathogen F. oxysporum suggested the impor-ant contribution these antibiotics may have inotential biocontrol activity.

cknowledgements

uthors gratefully acknowledge the help of Dr.ulf Blankenfeldt, Max-Planck Institute of Molecu-

ar Physiology, Dortmund, Germany, in the chemicaldentification of phenazines, and USDA-ARS fortrains Pf-5 and 2-79. Authors also acknowledge thenancial support from Department of Science andechnology, UGC-SAP and DST-FIST, Govt. of India.

eferences

lexeyer MF. The pKNOCK series broad-range mobilizablesuicide vectors for gene knockout and targeted DNAinsertion into the chromosome of Gram-negative bac-teria. BioTechniques 1999;26:824—8.

L

M

M

M

M

M

M

M

M

M

M

P

P

P

334

Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG,Smith JA, et al. Short protocols in molecular biology.3rd ed. New York: John Wiley and Sons; 1992.

Chin-A-Woeng TFC, Thomas-Oates JE, Lugtenberg BJJ,Bloemberg GV. Introduction of the phzH geneof Pseudomonas chlororaphis PCL1391 extends therange of biocontrol ability of phenazine-1-carboxylicacid-producing Pseudomonas spp. strains. Mol PlantMicrobe Interact 2001;14:1006—15.

Combs MT, Ashraf-Khorassani M, Taylor LT. HPLC:atmospheric pressure chemical ionization-mass spec-troscopy of eight regulated sulfonamides. J PharmBiomed Anal 1999;25:301—8.

Costa R, van Aarle IM, Mendes R, Van Elsas JD.Genomics of pyrrolnitrin biosynthetic loci: evi-dence for conservation and whole-operon mobilitywithin Gram-negative bacteria. Environ Microbiol2009;11:159—75.

Couillerot O, Prigent-Combaret C, Caballero-MelladoJ, Moenne-Loccoz Y. Pseudomonas fluorescens andclosely-related fluorescent pseudomonads as biocon-trol agents of soil-borne phytopathogens. Lett ApplMicrobiol 2009;48:505—12.

Delaney SM, Mavrodi DV, Bonsall RF, Thomashow LS. phzO,a gene for biosynthesis of 2-hydroxylated phenazinecompounds in Pseudomonas aureofaciens 30-84. J Bac-teriol 2001;183:318—27.

Flood ME, Herbert RB, Holliman FG. Pigments of Pseu-domonas species. V. Biosynthesis of pyocyanin and thepigments of Pseudomonas aureofaciens. J Chem Soc[Perkin 1] 1972;4:622—6.

Gurusiddaiah S, Weller DM, Sarkar A, Cook RJ. Char-acterization of an antibiotic produced by a strainof Pseudomonas fluorescens inhibitory to Gaeuman-nomyces graminis var. tritici and Pythium spp.Antimicrobial Agents Chemother 1986;29:488—95.

Haas D, Defago G. Biological control of soil-bornepathogens by fluorescent pseudomonads. Nat RevMicrobiol 2005;3:307—19.

Hammer PE, Burd W, Hill DS, Ligon JM, van Pee K. Conser-vation of the pyrrolnitrin biosynthetic gene clusteramong six pyrrolnitrin-producing strains. FEMS Micro-biol Lett 1999;180:39—44.

Hill DS, Stein JI, Torkewitz NR, Morse AM, Howell CR,Pachlatko JP. Cloning of genes involved in the synthe-sis of pyrrolnitrin from Pseudomonas fluorescens androle of pyrrolnitrin synthesis in biological control ofplant disease. Appl Environ Microbiol 1994;60:78—85.

Karberg M, Guo H, Zhong J, Coon R, Perutka J, LambowitzA. Group II introns as controllable gene targetingvectors for genetic manipulation of bacteria. NatBiotechnol 2001;19:1162—7.

Kerr JR. Phenazine pigments: antibiotics and virulencefactors. Rev Infect Dis 2000;2:184—94.

Kirner S, Hammer PE, Hill DS, Altmann A, Fischer I, WeisloLJ. Functions encoded by pyrrolnitrin biosyntheticgenes from Pseudomonas fluorescens. J Bacteriol

1998;180:1939—43.

Laursen JB, Nielsen J. Phenazine natural products:biosynthesis, synthetic analogues, and biologicalactivity. Chem Rev 2004;104:1663—85.

P

A. Upadhyay, S. Srivastava

eisinger T, Margraff R. Secondary metabolites ofthe fluorescent Pseudomonads. Microbiol Rev1979;43:422—42.

addula VSRK, Pierson EA, Pierson III LS. Altering the ratioof phenazines in Pseudomonas chlororaphis (aureofa-ciens) strain 30-84: effects on biofilm formation andpathogen inhibition. J Bacteriol 2008;190:2759—66.

alhotra M, Srivastava S. Targeted engineering ofAzospirillum brasilense strain SM with IndoleAcetamide pathway for IAA over-expression. Can JMicrobiol 2006;52:1078—84.

alhotra M, Srivastava S. An ipdC gene knock-outof Azospirillum brasilense strain SM and its impli-cations on indole-3-acetic acid biosynthesis andplant growth promotion. Antonie van Leeuwenhoek2008;93:425—33.

alhotra M. Genetic analysis and manipulation of IAAbiosynthesis in Rhizospheric isolates: Azospirillumbrasilense and Pseudomonas fluorescens Psd. PhD the-sis 2007.

avrodi DV, Blankenfeldt W, Thomashow LS. Phenazinecompounds in fluorescent Pseudomonas spp.biosynthesis and regulation. Annu Rev Phytopathol2006;44:417—45.

avrodi DV, Bonsall RF, Delaney SM, Soule MJ, PhillipsG, Thomashow LS. Functional analysis of genesfor biosynthesis of pyocyanin and phenazine-1-carboxamide from Pseudomonas aeruginosa PAO1. JBacteriol 2001;183:6454—65.

avrodi DV, Ksenzenko VN, Bonsall RF, Cook RJ, BoroninAM, Thomashow LS. A seven-gene locus for synthesisof phenazine-1-carboxylic acid by Pseudomonas fluo-rescens. J Bacteriol 1998;180:2541—8.

azurier S, Corberand T, Lemanceau P, Raaijmakers JM.Phenazine antibiotics produced by fluorescent pseu-domonads contribute to natural soil suppressivenessto Fusarium wilt. ISME 2009;3:977—91.

azzola M, Cook RJ, Thomashow LS, Weller DM, PiersonIII LS. Contribution of phenazine antibiotic biosyn-thesis to the ecological competence of fluorescentpseudomonads in soil habitats. Appl Environ Microbiol1992;58:2616—24.

eyer JM, Abdallah MA. The fluorescent pigment ofPseudomonas fluorescens: biosynthesis, purificationand physicochemical properties. J Gen Microbiol1978;107:319—28.

aulsen IT, Press CM, Ravel J, Kobayashi DY, Myers GS,Mavrodi DV, et al. Complete genome sequence of theplant commensal Pseudomonas fluorescens Pf-5. NatBiotechnol 2005;23:873—8.

erutka J, Wang W, Goerlitz D, Lambowitz AM. Useof computer-designed group II introns to disruptEscherichia coli DExH/D-box protein and DNA helicasegenes. J Mol Biol 2004;336:421—39.

ierson LS, Thomashow LS. Cloning and heterologousexpression of the phenazine biosynthetic locus fromPseudomonas aureofaciens 30-84. Mol Plant: Microbe

Interact 1992;5:330—9.

rice-Whelan A, Dietrich LEP, Newman DK. Rethink-ing ‘secondary’ metabolism: physiological roles forphenazine antibiotics. Nat Chem Biol 2006;2:71—8.

P rum

R

R

S

S

T

T

U

U

Y

Z

CA contributes for biocontrolling Fusarium oxyspo

aaijmakers JM, Weller DM, Thomashow LS. Frequencyof antibiotics producing spp. in natural environments.Appl Environ Microbiol 1997;63:881—7.

osales AM, Thomashow L, Cook RJ, Mew TW. Isolationand identification of antifungal metabolites producedby rice-associated antagonistic Pseudomonas sp. Phy-topathology 1995;85:1028—32.

ambrook J, Russell DW. Molecular cloning. In: A labora-tory manual. Cold Spring Harbor: New York Cold SpringHarbor Laboratory Press; 2001.

elin C, Habibian R, Poritsanos N, Athukorala SNP,Fernando D, Teresa R, et al. Phenazines are notessential for Pseudomonas chlororaphis PA23 biocon-trol of Sclerotinia sclerotiorum but do play a rolein biofilm formation. FEMS Microbiol Ecol 2010;71:73—83.

homashow LS, Weller DM. Role of a phenazine antibioticfrom Pseudomonas fluorescens in biological controlof Gaeumannomyces graminis var. tritici. J Bacteriol1988;170:3499—508.

335

ripathi RK, Gottlieb D. Mechanism of action ofthe antifungal antibiotic pyrrolnitrin. J Bacteriol1969;100:310—8.

padhyay A, Srivastava S. Characterization of a newisolate of Pseudomonas fluorescens strain Psd asa potential biocontrol agent. Lett Appl Microbiol2008;47:98—105.

padhyay A, Srivastava S. Evaluation of multiple plantgrowth promoting traits of an isolate of Pseu-domonas fluorescens strain Psd. Indian J Exp Biol2010;48:601—9.

ao J, Lambowitz AM. Gene targeting in Gram-Negativebacteria by use of a mobile group II intron (“tar-getron”) expressed from a broad-host-range vector.Appl Environ Microbiol 2007;73:2735—43.

hong J, Karberg M, Lambowitz A. Targeted and ran-

dom bacterial gene disruption using a group IIintron (targetron) vector containing a retrotranspo-sition activated selectable marker. Nucleic Acids Res2003;31:1656—64.