Ga

MAa

b

ARRAA

KAGGPVV

1

akvsdsicgebtanmp

PB

0d

Journal of Virological Methods 155 (2009) 167–174

Contents lists available at ScienceDirect

Journal of Virological Methods

journa l homepage: www.e lsev ier .com/ locate / jv i romet

rapevine virus A-mediated gene silencing in Nicotiana benthamianand Vitis vinifera

ookkan Murugananthama, Yoni Moskovitza, Sabrina Haviva, Tamar Horesha,nnie Fenigsteina, Jacques du Preezb, Dirk Stephanb, Johan T. Burgerb, Munir Mawassia,∗

The S. Tolkowsky Laboratory, Department of Plant Pathology-The Virology Unit, Agricultural Research Organization, Volcani Center, Bet Dagan 50250, IsraelDepartment of Genetics, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa

rticle history:eceived 12 June 2008eceived in revised form 4 October 2008ccepted 9 October 2008vailable online 2 December 2008

a b s t r a c t

Virus-induced gene silencing (VIGS) is an attractive approach for studying gene function. Although thenumber of virus vectors available for use in VIGS experiments has increased in recent years, most of thesevectors are applied in annual or herbaceous plants. The aim of this work was to develop a VIGS vector basedon the Grapevine virus A (GVA), which is a member of the genus Vitivirus, family Flexiviridae. The GVA vectorwas used to silence the endogenous phytoene desaturase (PDS) gene in Nicotiana benthamiana plants. In

eywords:groinoculationrapevinerapevine virus Ahytoene desaturase

addition, an Agrobacterium-mediated method for inoculating micropropagated Vitis vinifera cv. Primeplantlets via their roots was developed. Using this method, it was possible to silence the endogenous PDSgene in V. vinifera plantlets. The GVA-derived VIGS vector may constitute an important tool for improvingfunctional genomics in V. vinifera.

© 2008 Elsevier B.V. All rights reserved.

Htov

rsTmfavfGg

irus-induced gene silencingirus vector

. Introduction

Infection of a plant by a virus often triggers a defense responsend activates post-transcriptional gene silencing in the plant, aey mechanism for protecting plants against viral invasion. In thisirus-induced gene silencing (VIGS) response, the plant defenseystem specifically targets RNAs derived from the viral genome foregradation (Bernstein et al., 2001; Nykanen et al., 2001). As a con-equence, if the virus is harboring a fragment of a plant gene ints genome, the defense system will trigger the destruction of theorresponding plant mRNA, as well. On this basis, in recent years,enomes of plant viruses have been widely utilized to knock downxpression of either transgenes or endogenous genes, and haveeen proven to be important tools for the analysis of gene func-ion in plants (Atkinson et al., 1998; Baulcombe, 1999; Burton et

l., 2000; Lu et al., 2003; Burch-Smith et al., 2004). Although theumber of developed VIGS vectors has increased in recent years,ost of these vectors are meant to be used in annual or herbaceous

lants (Kumagai et al., 1995; Ruiz et al., 1998; Ratcliff et al., 2001;

∗ Corresponding author at: The S. Tolkowsky Laboratory, Department of Plantathology-The Virology Unit, Agricultural Research Organization, the Volcani Center,et Dagan 50250, Israel. Tel.: +972 3 9683844; fax: +972 3 9604180.

E-mail address: mawassi@volcani.agri.gov.il (M. Mawassi).

owgpps

ubp

166-0934/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.jviromet.2008.10.010

olzberg et al., 2002; Brigneti et al., 2004). However, new virus vec-ors are required to expand the application of VIGS to a wider rangef plants. The aim of this study was to test whether the Grapevineirus A (GVA) could be used as a VIGS vector.

GVA is closely associated with the economically importantugose wood (RW) disease of grapevine, specifically with Kobertem grooving (Garau et al., 1994; Chevalier et al., 1995, 1997).his virus, which is spread through infected propagation plantaterials and by mealybugs, is a member of the genus Vitivirus,

amily Flexiviridae (Martelli et al., 2007). It is a filamentous particlebout 800 nm long, and is considered to be a phloem-associatedirus. The GVA genome (∼7.4 kb) consists of five open readingrames (ORFs; Galiakparov et al., 1999, Saldarelli et al., 2000;aliakparov et al., 2003c). ORF1, located at the 5′-terminus of theenome, encodes a 194-kDa polypeptide with conserved motifsf replication-related proteins. ORF2 encodes a ∼20-kDa proteinhose function is unknown. ORF3 is the movement protein (MP)

ene, ORF4 encodes the coat protein (CP) and ORF5 encodes a smallrotein that exhibits sequence similarities to small RNA bindingroteins of various plant viruses (Galiakparov et al., 2003b) and

uppresses RNA silencing (Chiba et al., 2006; Zhou et al., 2006).

Recently a GVA-derived expression vector was developed andsed to express the beta-glucuronidase (GUS) gene in Nicotianaenthamiana plants (Haviv et al., 2006). Infection of Vitis viniferalants with the GVA vector has not been developed so far.

1 f Virol

Idode(iGde

2

2

Ndpoaa

2

uepc1fGtP

NDp -3GseNBt

(dftvV3GauIvpt

wt

awC

2

Act1tpJ

atdttaWlatap1ww5

2b

Rid

u(dtg

eNApVsgt(w

aprobes (Roche Molecular Biochemicals). The probes, specific to the

68 M. Muruganantham et al. / Journal o

nfection of this host with a cloned infectious cDNA of GVA isifficult to accomplish using the simple agro-infiltration meth-ds that are used for most plant viruses. The present studyescribes: (i) the use of the GVA-derived vector for silencing thendogenous phytoene desaturase (PDS) gene in N. benthamiana,ii) an Agrobacterium-mediated method developed for inoculat-ng in vitro-propagated V. vinifera plantlets with cloned infectiousVA cDNAs, through their roots, and (iii) the use of the methodeveloped for applying the GVA-derived vector for silencing thendogenous PDS gene in V. vinifera plantlets.

. Materials and methods

.1. Plant material

N. benthamiana and V. vinifera cv. Prime were used in this study.. benthamiana plants were grown in pots under greenhouse con-itions: 25 ◦C; a 16-h light cycle and 60% humidity. V. viniferalantlets, which were obtained from plant material treated by cry-preservation to ensure virus elimination, were micropropagatednd maintained at a temperature of 26 ◦C under a 16-h photoperiod,s described by Wang et al. (2003).

.2. GVA constructs

Fig. 1 presents schematic maps of the GVA-derived constructssed in this study. The vector pGVA-118, which is described by Havivt al. (2006), contains the GVA cDNA under the promoter of T7 RNAolymerase. The entire cDNA of GVA-118 was PCR-amplified andloned into the Stu I cleavage site of the pCass2 vector (Shi et al.,997) between the 35S promoter and terminator sequences derivedrom Cauliflower mosaic virus (CaMV). The cassette, consisting of theVA cDNA and the CaMV 35S promoter and terminator, was next

ransferred into the pCAMBIA2301 binary vector with the help ofvu II digestion, to generate pGVA-378.

The cDNA fragment corresponding to the nts 1–500 of the. benthamiana PDS gene (NbPDS) (GenBank accession no.Q469932) was PCR-amplified using Taq DNA polymerase and therimers NbPDSf1 (5′-AATCATGCGGCCGCATGCCCCAAATCGGACTTGT′) (Italics Not I cleavage site) and NbPDSr1 (5′-CTCTTA-GGCCCAATATGTGCAACCCAGTCTC-3′) (Italics Apa I cleavageite). The PCR product was digested with Not I and Apa I restrictionnzymes and cloned into a similarly digested pGVA-118 vector.ext, the resulting GVA-nbPDS cDNA was inserted into the pCAM-IA2301 vector, under the control of the CaMV 35S promoter anderminator, to generate pGVA-nbPDS-349.

To obtain sequence information about the V. vinifera PDSVvPDS) gene, sequence alignments of at least 16 PDS cDNAserived from various plants and available in GenBank were per-ormed. Sequences of conserved regions were selected and usedo design specific primers. A 304-nts cDNA fragment of the V.inifera cv. Prime PDS gene was PCR-amplified with the primersvPDSf1 (5′-AATCATGCGGCCGCGGCCTTCTTAGATGGTAATCCT-′) (Italics Not I cleavage site) and VvPDSr1 (5′-CTCTTA-GGCCCCTCAAACCATATATGAACATTGA-3′) (Italics Apa I cleav-ge site). The PCR product (submitted for publication in GenBanknder the accession no. EU816356) was digested with Not I and Aparestriction enzymes and cloned into a similarly cleaved pGVA-118ector. The resulting GVA-nbPDS cDNA was next inserted into the

CAMBIA2301, under the control of the CaMV 35S promoter anderminator, to generate pGVA-VvPDS-377.

The green fluorescent protein (GFP) gene was PCR-amplifiedith primers specific to the 18 nts of the 5′ and the 3′ termini of

he gene. The product was cloned into pGVA-118, between the Not I

N7tfi

ogical Methods 155 (2009) 167–174

nd Apa I cleavage sites. The resulting recombinant GVA-GFP cDNAas then inserted into the pCAMBIA2301, under the control of theaMV 35S promoter and terminator, to generate pGFP-GVA-160.

.3. Agro-inoculation of N. benthamiana and V. vinifera plantlets

Virus infections of N. benthamiana were achieved throughgrobacterium-mediated transient expression of infectious GVADNAs cloned into the binary vector pCAMBIA2301. Agrobacteriumumefaciens strain EHA 105, re-suspended to an OD600 of 1.0 in0 mM MgCL2, 10 mM MES and 100 �M acetosyringone, was infil-rated to the underside of leaves of 3-week-old N. benthamianalants using a 2 ml syringe without a needle, as described by

ohansen and Carrington (2001).Virus infections of micropropagated V. vinifera plantlets were

chieved by inoculating the roots of the plantlets with Agrobac-erium, using a modified version of the agrodrenching technologyescribed by Ryu et al. (2004). Micropropagated plantlets, withwo to four leaves each, were detached from the agar mediumogether with their roots. The roots were then gently injured with

sterile needle. The plantlets were next transferred onto sterilehatman No. 1 filter paper immersed in 10 ml half-strength MS

iquid medium (Murashige and Skoog, 1962) with BA at 1 mg/lnd NAA at 0.5 mg/l. For inoculation, a 1 ml suspension of Agrobac-erium inoculum, re-suspended in MS medium containing 100 �Mcetosyringone to an OD600 of 1.0, was added to the media. Thelantlets were kept with the Agrobacterium-containing media for0 days and then transferred to fresh MS liquid media, where theyere kept for another 14 days. At the end of this period, the plantletsere transferred to fresh media supplemented with cefotaxime at00 mg/l and maintained for subsequent analyses.

.4. RNA extraction, RT-PCR, semi-quantitative PCR and Northernlot analyses

Total RNA from N. benthamiana was extracted using the Tri-eagent kit (Sigma–Aldrich, USA) according to the manufacturer’s

nstructions. Total RNA of V. vinifera plants was extracted asescribed by Chang et al. (1993).

For RT-PCR, total RNA extracts were treated with DNAase andsed for cDNA synthesis with the primer 18dT composed of polyT). PCR amplifications were performed using primers that wereesigned based on the sequences of the N. benthamiana PDS gene,he V. vinifera PDS gene (obtained as described above) and the GVAenome (GenBank accession no. AY244516).

Semi-quantitative PCR was performed as described by Liut al. (2002), using the N. benthamiana PDS-specific primersbPDSf2 (5′-GGTTGCAGTGGAAGGAACAT-3′) and NbPDSr2 (5′-GCGTACACACTGAGCAACG-3′), or the V. vinifera PDS-specificrimers VvPDSf2 (5′-CTTACCCAAATGTGCAGAACCTGTT-3′) andvPDSr2 (5′-CCTGGGTTCAAAGCAATCAATATACA-3′). As a control,emi-quantitative PCR was conducted to amplify the host elon-ation factor (EF)-1 � mRNA from the same cDNA product withhe primers rEF1 (5′-GATTGGTGGTATTGGAACTGTC 3′) and fEF-15′-AGCTTCGTGGTGCATCTC-3′). The intensities of the PCR productsere analyzed using Gel Doc 2000 (Bio-Rad, USA).

Northern blot and hybridization analyses were performedccording to Galiakparov et al. (2003a), with DIG-labeled ribo-

. benthamiana PDS and the V. vinifera PDS, were made of the same60- and 420-nts fragments. These fragments were obtained fromhe semi-quantitative PCR assays described above. The probe usedor GVA detection was designed to hybridize with the 5′-1.0 kb ofts genome.

M. Muruganantham et al. / Journal of Virological Methods 155 (2009) 167–174 169

Fig. 1. Schematic diagrams of the GVA-derived clones used in this study. (a) GR-5, containing the full-length cDNA of GVA genome cloned into the pUC57 vector under thepromoter T7 RNA polymerase. Boxes represent open reading frames (ORFs); CP, coat protein; MP, movement protein. (b) GVA-118; a GVA-based expression vector containingtwo heterologous MP-sgRNA promoters derived from two strains of GVA and a sequence of multiple cleavage sites (MCSs) of restriction enzymes (Haviv et al., 2006). TheGVA cDNA was cloned into the pUC57 vector under the control of the T7 RNA polymerase promoter. (c) GVA-378; a GVA-based expression vector as in (b), but with the cDNAc a termt ing thc

2

gci

3

3

wp(Iw3bn

iopwPt7vtHl(ivn

loned into the pCAMBIA2301 binary vector between two CaMV 35S promoters andhe N. benthamiana PDS gene. (e) GVA-VvPDS-377; a GVA-378-based vector containontaining the GFP gene.

.5. Electron microscopy

Leaves and stems of micropropagated V. vinifera plantlets wereround with 0.1 M phosphate buffer (pH 7.0). The sap was thenentrifuged at 3000 × g for 3 min and the virus particles presentn the sap were observed by means of electron microscopy.

. Results and discussion

.1. VIGS in N. benthamiana

The vector GVA-118, described previously by Haviv et al. (2006),as designed to express the foreign gene through the MP-sgRNAromoter. A 500 nts product of the PDS gene of N. benthamianaNbPDS) was PCR-amplified and cloned between the Not I and Apa

cleavage sites of the GVA-118. The resulting GVA-NbPDS cDNAas then inserted into the pCAMBIA2301 to generate pGVA-nbPDS-49, as described in Materials and Methods (Fig. 1). Plants of N.enthamiana were inoculated with Agrobacterium carrying pGVA-bPDS-349 through infiltration of leaves. As a control, plants were

ltvmu

inator. (d) GVA-nbPDS-349; a GVA-378-based vector containing the 5′-500 nts ofe 306 nts of the Vitis vinifera PDS gene. (f) GFP-GVA-160; a GVA-378-based vector

nfiltrated with Agrobacterium carrying the pCAMBIA2301, in whichnly a similar NbPDS fragment was cloned between the CaMV 35Sromoter and terminator. For simplicity, the Agrobacterium cultureill be referred to here by the name of the assigned construct.

lants inoculated with GVA-nbPDS-349 started to display pheno-ypes that indicated the silencing of the endogenous PDS gene at–14 days post-infection (dpi). The leaves exhibited photobleachedeins (Fig. 2), as a result of a lack of carotenoids and the destruc-ion of chlorophyll by photo-oxidation (Kumagai et al., 1995).owever, neither the infiltrated leaves nor the newly developed

eaves of the control plants exhibited photobleached phenotypesresults not shown). These results indicate that the photobleach-ng symptoms observed in GVA-nbPDS-349-inoculated plants wereirus-vector-mediated. The photobleaching symptoms in GVA-bPDS-349-inoculated plants were prominent in the midrib and

ateral veins relatively soon after inoculation (7–14 dpi). At 20 dpi,he photobleaching symptoms were observed in other branched-offeins and veinlets; whereas the surrounding leaf tissue remainedostly green. At later stages (>30 dpi), the upper leaves of the inoc-

lated plants displayed only a few photobleached-veins, though

170 M. Muruganantham et al. / Journal of Virological Methods 155 (2009) 167–174

Fig. 2. Virus-induced silencing of the PDS gene in N. benthamiana. (A) Phenotypes of leaves of N. benthamiana inoculated with GVA-378 (a) or with GVA-nbPDS-349, 14 dpi (band c); (c shows phenotypes at 40 dpi.) (B) Northern blot analysis of GVA RNA extracted from N. benthamiana plants inoculated with GVA-378 or GVA-nbPDS-349. RNA wasextracted from green veins (lane 1), photobleached veins (lane 2) and green foliar tissue from the areas between the veins (lane 3). M; RNA extracted from mock-inoculatedplants. (C) Northern blot analysis of the PDS mRNA extracted from N. benthamiana plants inoculated with GVA-378 at 21 dpi, or with GVA-nbPDS-349 at 21 and 30 dpi. (D)S mRNi an olig( amplr

toama

emi-quantitative RT-PCR analysis was conducted to examine the silencing of PDSsolated from silenced and non-silenced plants and used for cDNA synthesis, withlower panel) plant genes. Lanes 1–6 show the products at 15, 18, 21, 24, 27 and 30eaction. Lane M, marker sizes. Arrows indicate the positions of the PCR products.

hese leaves exhibited symptoms of virus infection (Fig. 2Ac). Thisbservation suggests that the inserted PDS sequence can be splicednd removed from the GVA vector or, alternatively, that GVA-ediated suppression of the silenced PDS mRNA is more effective

t a later time point post-infection.

opvs

A in N. benthamiana inoculated with GVA-378 or GVA-nbPDS-349. Total RNA waso-(dT) primer, followed by PCR amplification of the PDS (upper panel) and EF-1�

ification cycles. Lane 7 shows the amplified sequences obtained in a template-less

The expression of bleaching phenotypes mainly in the leaf veinsf plants inoculated with GVA-nbPDS-349 resembles the findingsreviously reported by Haviv et al. (2006). In that study, the GVAector was applied to express the GUS gene through the MP-gRNA promoter and subsequent GUS accumulation was detectable

f Virol

mbaNgtrGstlwSeteitGopGontmptRv5gplddss

aGaopb5ptl

mvwipwNfiPahoTlv

t8cftlb

3G

rvoitvcpaottb

tecpepltp

tfiadtabwtprpksTmiaapwprG

M. Muruganantham et al. / Journal o

ainly within and adjacent to leaf veins. These observations maye attributed to several possible factors. First, it may be that GVAccumulation is mostly limited to leaf veins. To test this possibility,orthern blot analyses were performed on total RNA obtained fromreen leaf veins of GVA-378-infiltrated plants and from green foliarissue of GVA-nbPDS-349-infiltrated plants. Hybridization with aiboprobe specific to the 5′-GVA terminus revealed the presence ofVA RNAs corresponding to the genomic RNA and to the 5′-terminalub-genomic RNAs (Galiakparov et al., 2003a), in both veins and theissue between the veins (Fig. 2B, lanes 1 and 3). Nevertheless, theevels of the GVA RNAs differed in both tissues; levels were some-

hat higher within the veins (lane 1) than between them (lane 3).econd, the sequence carried by the GVA vector could be mostlyxpressed in the phloem, perhaps as a result of its being controlledhrough the MP-sgRNA promoter, which is known to control thexpression of the product dedicated to virus movement. To exam-ne this possibility, a GVA vector that expresses the foreign genehrough a CP-sgRNA promoter derived from the related vitivirusrapevine virus B (GVB) was designed. The 5′ 500-nts fragmentf NbPDS was then inserted into the designed vector to produceGVA-nbPDS-351 (not described in Fig. 1). Plants infiltrated withVA-nbPDS-351 exhibited a photobleaching phenotype in the veinsf their leaves, similar to that induced by GVA-nbPDS-349 (resultsot shown). This result indicates that restriction of photobleachingo leaf veins is apparently not linked to the type of the viral pro-

oter. Third, it is possible that VIGS is activated differently in thehotobleached leaf veins than in the green leaf tissue. To examinehis possibility, Northern blot analyses were performed using totalNA obtained from green foliar tissue and from photobleached leafeins of GVA-nbPDS-349-infiltrated plants. Hybridization with a′ GVA-specific riboprobe demonstrated that smaller amounts ofenomic and 5′-terminal sub-genomic RNAs accumulated in thehotobleached leaf veins than in the green foliar tissue (Fig. 2B,

anes 2 and 3). This demonstrates that the viral RNA accumulatedifferently in the two tissues, suggesting that VIGS expression isifferent as well. This would be possible if the suppression of RNAilencing by the virus is activated differently in these different tis-ues.

Because photobleaching symptoms are believed to be associ-ted with silencing and lower concentrations of the PDS mRNA,VA-nbPDS-349-infiltrated plants were subjected to Northern blotnalyses. Total RNA was extracted from photobleached leaf veinsf GVA-nbPDS-349-infiltrated plants and from green leaf veins ofGVA-378-infiltrated plants. The RNA was subjected to Northernlot and hybridization analyses with a riboprobe specific to the nts58–1318 of the NbPDS mRNA (a portion of the gene which is notresent in GVA-nbPDS-349). The results presented in Fig. 2C showhat smaller amounts of PDS mRNA accumulated in photobleachedeaf veins at 21 and 30 dpi, as compared to green leaf veins.

In addition to Northern blot analyses, reduction of the PDSRNA in photobleached leaf veins, as compared to green leaf

eins, was examined using a semi-quantitative PCR assay. Total RNAas extracted from photobleached leaf veins of GVA-nbPDS-349-

nfiltrated plants and from green leaf veins of GVA-378-infiltratedlants and used for cDNA synthesis with primer 18dT. The cDNAas subjected to a semi-quantitative PCR assay using the primersbPDSf2 and NbPDSr2. The use of these primers ensures the ampli-cation of a 760 nts fragment (nts 558–1318) of the N. benthamianaDS gene, which is not present in pGVA-nbPDS-349. As a control,semi-quantitative PCR procedure was performed to amplify the

ost EF-1 � mRNA from the same cDNA samples. The intensitiesf the PCR products were analyzed using Gel Doc 2000 (Bio-Rad).he results, presented in Fig. 2D, show that there was remarkablyess amplification of the plant PDS mRNA in photobleached leafeins than in the green leaf veins. Quantitation analysis revealed

ti

ea

ogical Methods 155 (2009) 167–174 171

hat the mean level of PDS mRNA in the photobleached veins was5% lower than that in the green leaf veins. In contrast, the PCRontrols exhibited similar levels of amplification of the EF-1� RNAor both the photobleached and green leaf veins. Taken together,hese findings suggest that the photobleaching symptoms in theeaf veins of GVA-nbPDS-349-infiltrated N. benthamiana plants cane attributed to the silencing of the plant PDS mRNA.

.2. Attempts to infect V. vinifera plants with a cloned cDNA ofVA

Since infiltration of N. benthamiana plants with Agrobacteriumesulted in efficient virus inoculation, the first attempts to infect V.inifera cv. Prime plants with an infectious GVA clone were basedn the infiltration of leaves of mature plants. Agrobacterium carry-ng the GFP-tagged GVA (named GFP-GVA-160, Fig. 1) was usedo directly infiltrate expanded leaves. Yet, no GFP expression orirus accumulation was detected. Next, additional inoculation pro-edures were tested, including stem-slashing, which has been usedreviously for inoculating citrus plants with CTV (Satyanarayana etl., 2001), and rubbing or biolistic bombardment of cambial facesf bark patches with the Agrobacterium-carrying GFP-GVA-160 andhen grafting them in place on V. vinifera plants. However, none ofhe inoculated plants (total of ∼50) became infected, as examinedy RT-PCR and by confocal microscopy assays.

It proved difficult to inoculate mature grapevine plants usinghe Agrobacterium-mediated procedures. Therefore, the next strat-gy was to examine agroinfection of micro-propagated V. viniferav. Prime plantlets, an approach which would provide a year-round,lentiful supply of plants. To develop a protocol for the deliv-ry of Agrobacterium-carrying GFP-GVA-160 to micropropagatedlantlets, vacuum infiltration, combined with piercing the plantlet

eaves with a needle was examined initially. However, the applica-ion of this approach resulted in infection of only 5% of the testedlantlets (∼150), according to confocal microscopy assays (Fig. 3A).

Because agroinfection of the grapevine foliage did not appearo be sufficiently effective, we then attempted to use a modi-ed version of the agrodrenching method described by Ryu etl. (2004). The original agrodrenching method was used to infectiverse Solanoceae species through soil drenched with an Agrobac-erium carrying a Tobacco rattle virus (TRV)-derived vector (Ryu etl., 2004). Two-week-old V. vinifera cv. Prime plantlets, which hadeen cultured on a solid BM rooting medium (Wang et al., 2003),ere detached and their roots were wounded and inserted into

ubes with filter papers immersed in an MS liquid medium. Thelantlets were inoculated with a suspension of Agrobacterium car-ying the pGVA-378, as described in Section 2. Ten days later, thelantlets were transferred to fresh liquid media, where they wereept for 14 days. After this period, the plantlets were transferred toimilar media supplemented with cefotaxime, as described above.he agrodrenched plantlets were examined for virus infectivity byeans of RT-PCR 28 days after the agrodrenching (dpd). Fig. 3B

llustrates the RT-PCR results obtained with representative mock-nd GVA-378-inoculated plantlets, and shows the amplification ofGVA fragment (nts 6047–6251) from the GVA-378-inoculated

lantlet. However, no amplification was obtained when the PCRas conducted directly on the total RNA of the GVA-378-inoculatedlant, without cDNA synthesis (lane 3). Taken together, theseesults provide evidence for infection of V. vinifera plantlet withVA-378. When this inoculation protocol was applied it was found

hat approximately 50% of the plantlets tested (>200) becamenfected by 28 dpd, as indicated by RT-PCR.

To further confirm the infection of V. vinifera plantlets, sap wasxtracted from leaves of plantlets that were agrodrenched with GVAnd this sap was examined under an electron microscope (Fig. 3C).

172 M. Muruganantham et al. / Journal of Virological Methods 155 (2009) 167–174

F confoG e arroG d direM ract f

Fbtdv

3

mvat3sta

mGt3i

te

iipapdobtr

onaoownmtn

ig. 3. Infection of V. vinifera plantlets with GVA-derived clones. (A) RepresentativeFP-GVA-160. (B) RT-PCR amplification of a 250-nts GVA fragment (indicated by thVA-378-inoculated plantlets (lane 2). Lane 3 contains the product of PCR conducte; marker sizes. (C) Electron micrograph showing a vitivirus-like particle in sap ext

lexuous and filamentous particles, ca. 800 nm in length, resem-ling vitivirus virions were visualized in some of tested plantlets,hough at an extremely low titer. However, no such particles wereetected in the mock-inoculated plantlets. These observations pro-ide additional evidence of infection of V. vinifera with GVA-378.

.3. VIGS in V. vinifera

The feasibility of the agrodrenching technology for infectingicropropagated plantlets led to the examination of VIGS in V.

inifera, as well. The 304 nts RT-PCR product of the PDS gene wasmplified from V. vinifera cv. Prime and inserted into GVA-118, andhen inserted into the pCAMBIA2301 vector between the CaMV5S promoter and terminator to generate pGVA-VvPDS-377. Auspension of Agrobacterium-carrying pGVA-VvPDS-377 was usedo agrodrench micropropagated V. vinifera plantlets, as describedbove.

The inoculated plantlets were examined for virus infectivity byeans of RT-PCR 28 dpd. The results of analyses of representativeVA-378- and GVA-VvPDS-377-inoculated plantlets (Fig. 4B) show

he amplification of a GVA::PDS fused fragment from GVA-VvPDS-

77-inoculated plantlets (Fig. 4B, lane 2), but not from GVA-378-noculated plantlets (Fig. 4B, lane 1).

Plantlets inoculated with GVA-VvPDS-377 began to display pho-obleaching symptoms that were attributed to silencing of thendogenous gene PDS at 14–20 dpd. These symptoms were exhib-

t

mwd

cal images showing green fluorescence in leaf tissue from plantlets inoculated withw) conducted on RNA extracted from non-inoculated plantlets (lane 1) and fromctly on the total RNA of a GVA-378-inoculated plant without cDNA synthesis. Lane

rom V. vinifera plantlets inoculated with GVA-378. Bar, 200 nm.

ted by 40–60% of the inoculated plantlets, and appeared mainlyn leaf veins and at leaf margins (Fig. 4A). At later dates (>30 dpd),hotobleaching was observed on the whole surface of some leavesnd, in some cases, in the leaf petioles. Thus the photobleachedhenotype observed on the plantlets of the V. vinifera leaves wasifferent from that observed on N. benthamiana, probably becausef the nature of the soft tissue of plantlets. In addition, the photo-leaching symptoms on V. vinifera were observed on only two orhree leaves of GVA-VvPDS-377-inoculated plantlets, probably as aesult of the subsequent suppression of VIGS by GVA (Fig. 4A).

To ascertain that the appearance of photobleaching symptomsn GVA-VvPDS-377-inoculated plantlets was GVA-mediated andot a result of RNA silencing induced by the transient expression ofVvPDS-containing sequence followed by long distance traffickingf RNA gene silencing, an experiment demonstrating the expressionf VvPDS in the absence of the GVA vector was conducted. Plantletsere agrodrenched with pCAMBIA2301 vector containing the 304ts VvPDS fragment, which was cloned between the CaMV 35S pro-oter and terminator sequences. The newly developed leaves of

hese plantlets did not exhibit photobleached phenotypes (resultsot shown), providing evidence that the photobleached pheno-

ypes are linked to virus vector infection.

In order to confirm VIGS and the decrease in the level of PDSRNA in photobleached leaves of V. vinifera plantlets inoculatedith GVA-VvPDS-377, a semi-quantitative RT-PCR assay was con-ucted. Total RNA was extracted from photobleached leaves and

M. Muruganantham et al. / Journal of Virological Methods 155 (2009) 167–174 173

Fig. 4. Virus-induced silencing of the PDS gene in V. vinifera plantlets. (A) Phenotypes appearing on leaves of silenced (s) plantlets inoculated with GVA-VvPDS-377 or onleaves of non-silenced (ns) plantlets inoculated with GVA-378. A whole plantlet with a photobleached phenotype, maintained in a culture tube, is shown on the right. (B)RT-PCR analyses performed to amplify the fused sequences of GVA ORF2 and V. vinifera PDS from RNA extracted from V. vinifera plantlets inoculated with GVA-378 (lane 1) orwith GVA-VvPDS-377 (lane 2). Lane 3 shows the amplification product obtained from a template-less reaction. (C) Semi-quantitative RT-PCR analysis conducted to examinethe silencing of PDS mRNA in V. vinifera inoculated with GVA-VvPDS-377 or GVA-378. Total RNA was isolated from leaves of silenced and non-silenced plantlets and used forc PDS (1 tainedo

fwaVfinqtllPe

toltlP

DNA synthesis, with an oligo-(dT) primer, followed by PCR amplification of partial8, 21, 24, 27 and 30 amplification cycles. Lane 6 shows the amplification product obf the PCR products.

rom green leaves of plantlets inoculated with GVA-378. The RNAas used for cDNA synthesis with the primer 18dT, followed bysemi-quantitative PCR assay using the primers VvPDSf2 and

vPDSr2. The use of these primers was meant to ensure the ampli-cation of a 420 nts fragment of the V. vinifera PDS gene, which isot present in pGVA-VvPDS-377. As a control, we performed a semi-uantitative PCR procedure to amplify the host EF-1 � mRNA from

he same cDNA. The results presented in Fig. 4C show remarkablyess amplification of the plant PDS mRNA from the photobleachedeaves than of the PDS mRNA from the green leaves. Analysis of theCR products by Gel Doc 2000 revealed that the PDS mRNA lev-ls in the photobleached leaves were, on average, 70% lower than

4

to

upper panel) and EF-1� (lower panel) plant genes. Lanes 1–5 show the products atfrom a template-less reaction. Lane M; marker sizes. Arrows indicate the positions

hose of the green leaves. The PCR controls exhibited similar levelsf amplification of the EF-1� RNA in both photobleached and greeneaves. These findings suggest that the photobleached phenotypeshat appeared on the micropropagated V. vinifera plantlets inocu-ated with GVA-VvPDS-377 are linked to the silencing of the plantDS mRNA.

. Conclusions

VIGS is considered to be a powerful tool for studying the func-ions of host genes (Burch-Smith et al., 2004). Until recently, thenly VIGS vectors available were suitable mainly for use in annual

1 f Virol

pibkvvcpo

wiitAiautiptittto

A

0Rtl5

R

A

B

B

B

B

B

C

C

C

C

G

G

G

G

G

H

H

J

K

L

L

M

M

N

R

R

R

S

S

S

74 M. Muruganantham et al. / Journal o

lants. To the best of our knowledge, the GVA-derived VIGS vectors the first to be used for silencing the endogenous PDS genes inoth N. benthamiana and V. vinifera plants. Unlike in most of thenown VIGS vectors, silencing of the PDS gene in N. benthamianaia the GVA-derived VIGS vector is apparently restricted to the leafeins, which could limit the use of this tool for genes essentiallyonfined to the veins. The phenotype of PDS silencing in micro-ropagated V. vinifera plantlets, however, was different, as it wasbserved throughout the veins and also in the inter-vein areas.

Application of silencing vectors to the woody grapevine plantould enable high-throughput analysis of gene function in this

mportant crop. The ability to induce VIGS depends on the abil-ty of the virus vector to infect the plant tissue, but it is difficulto infect grapevine with cloned GVA by means of the standardgrobacterium-based delivery strategies. The results of this study

ndicate that we have overcome these inoculation difficulties bydapting the agrodrench method of Ryu et al. (2004) for the inoc-lation of micropropagated V. vinifera plantlets and for silencinghe endogenous PDS gene through the GVA vector, thereby provid-ng a method and a tool for reverse genetic studies of a variety ofrocesses occurring in micropropagated V. vinifera plants. In addi-ion, the present work resulted in the generation of grapevine plantsnfected with a pure population of GVA derived from a cloned infec-ious cDNA sequence. This development is of major importance forhe investigation of molecular events associated with viral infec-ions responsible for disease establishment and the developmentf disease symptoms.

cknowledgements

This research was supported by grants IS-3784-05 and CB-9020-5 from BARD, the United States–Israel Binational Agriculturalesearch and Development Fund, and by grant number 565/05 fromhe Israeli Science Foundation. This report was approved for pub-ication as Agricultural Research Organization, Journal Series No.08/08.

eferences

tkinson, R.G., Bieleski, L.R.F., Gleave, A.P., Janssen, B.J., Morris, B.A.M., 1998. Post-transcriptional silencing of chalcone synthase in petunia using a geminivirus-based episomal vector. Plant J. 15, 593–604.

aulcombe, D.C., 1999. Fast forward genetics based on virus-induced gene silencing.Curr. Opin. Plant Biol. 2, 109–113.

ernstein, E., Caudy, A.A., Hammond, S.M., Hannon, G.J., 2001. Role for a bidentateribonuclease in the initiation step of RNA interference. Nature 409, 363–366.

rigneti, G., Martin-Hernandez, A.M., Jin, H., Chen, J., Baulcombe, D.C., 2004. Virus-induced gene silencing in Solanum species. Plant J. 39, 264–272.

urch-Smith, T.M., Anderson, J.C., Martin, G.B., Dinesh-Kumar, S.P., 2004. Applica-tions and advantages of virus-induced gene silencing for gene function studies

in plants. Plant J. 39, 734–746.

urton, R.A., Gibeaut, D.M., Bacic, A., Findlay, K., Roberts, K., Hamilton, A., Baulcombe,D.C., Fincher, G.B., 2000. Virus-induced gene silencing of a plant cellulose syn-thase gene. Plant Cell 12, 691–706.

hang, S., Puryear, J., Cairney, J., 1993. A simple and efficient method for isolatingRNA from pine trees. Plant Mol. Biol. Rep. 11, 113–116.

W

Z

ogical Methods 155 (2009) 167–174

hevalier, S., Greif, C., Clauzel, J.M., Walter, B., Fritsch, C., 1995. Use of animmunocapture-polymerase chain reaction procedure for the detection ofgrapevine virus A in Kober stem grooving infected grapevines. J. Phytopathol.143, 369–373.

hevalier, S., Digiaro, M., Savino, V., 1997. Further evidence that grapevine virus Ais the agent of Kober stem grooving. Extended Abstracts, 12th Meeting of ICVG,Lisbon, pp. 39–40.

hiba, M., Reed, J.C., Prokhnevsky, A.I., Chapman, E.J., Mawassi, M., Koonin, E.V., Car-rington, J.C., Dolja, V.V., 2006. Diverse suppressors of RNA silencing enhanceagroinfection by a viral replicon. Virology 346, 7–14.

aliakparov, N., Tanne, E., Sela, I., Gafny, R., 1999. Infectious RNA transcripts fromGrapevine virus A cDNA clone. Virus Genes 19, 235–242.

aliakparov, N., Goszczynski, D.E., Che, X., Batuman, O., Bar-Joseph, M., Mawassi,M., 2003a. Two classes of subgenomic RNA of Grapevine Virus A produced byinternal controller elements. Virology 312, 434–448.

aliakparov, N., Tanne, E., Mawassi, M., Gafny, R., Sela, I., 2003b. ORF 5 of GrapevineVirus A encodes a nucleic acid-binding protein and affects pathogenesis. VirusGenes 27, 257–262.

aliakparov, N., Tanne, E., Sela, I., Gafny, R., 2003c. Functional analysis of theGrapevine Virus A genome. Virology 306, 42–50.

arau, R., Prota, V.A., Piredda, R., Boscia, D., Prota, U., 1994. On the possible rela-tionship between Kober stem grooving and Grapevine Virus A. Vitis 33, 161–163.

aviv, S., Galiakparov, N., Goszczynski, E.D., Batuman, O., Czosnek, H., Mawassi, M.,2006. Engineering the genome of Grapevine Virus A into a vector for expressionof proteins in herbaceous plants. J. Virol. Meth. 132, 227–231.

olzberg, S., Brosio, P., Gross, C., Pogue, G.P., 2002. Barley stripe mosaic virus-inducedgene silencing in a monocot plant. Plant J. 30, 193–200.

ohansen, L.K., Carrington, J.C., 2001. Silencing on the spot. Induction and suppressionof RNA silencing in the Agrobacterium-mediated transient expression system.Plant Physiol. 126, 930–938.

umagai, M.H., Donson, J., Della-Cioppa, G., Harvey, D., Hanley, K., Grill, L.K., 1995.Cytoplasmic inhibition of carotenoid biosynthesis with virus-derived RNA. Proc.Natl. Acad. Sci. U.S.A. 92, 1679–1683.

iu, Y., Schiff, M., Marathe, R., Dinesh-Kumar, S.P., 2002. Tobacco Rar1, EDS1 andNPR1/NIM1 like genes are required for N-mediated resistance to tobacco mosaicvirus. Plant J. 30, 415–429.

u, R., Martin-Hernandez, A.M., Peart, J.R., Malcuit, I., Baulcombe, D.C., 2003. Virus-induced gene silencing in plants. Methods 30, 296–303.

artelli, G.P., Adams, M.J., Kreuze, J.F., Dolja, V.V., 2007. Family Flexiviridae:a case study in virion and genome plasticity. Annu. Rev. Phytopathol. 45,73–100.

urashige, T., Skoog, F., 1962. A revised medium for rapid growth and bioassays withtobacco tissue cultures. Plant Physiol. 15, 473–497.

ykanen, A., Haley, B., Zamore, P.D., 2001. ATP requirements and small interferingRNA structure in the RNA interference pathway. Cell 107, 309–321.

atcliff, F., Martin-Hernandez, A.M., Baulcombe, D.C., 2001. Tobacco rattle virus as avector for analysis of gene function by silencing. Plant J. 25, 237–245.

uiz, M.T., Vionnet, O., Baulcombe, D.C., 1998. Initiation and maintenance of virus-induced gene silencing. Plant Cell 10, 937–946.

yu, C.M., Anand, A., Kang, L., Mysore, K.S., 2004. Agrodrench: A novel and effectiveagroinoculation method for virus-induced gene silencing in roots and diverseSolanaceous species. Plant J. 40, 322–331.

aldarelli, P., Dell’Orco, M., Minafra, A., 2000. Infectious cDNA clones of twograpevine viruses. Arch. Virol. 145, 397–405.

atyanarayana, T., Bar-Joseph, M., Mawassi, M., Albiach-Marti, M.R., Ayllon, M.A.,Gowda, S., Hilf, M.E., Moreno, P., Garnsey, S.M., Dawson, W.O., 2001. Amplificationof citrus tristeza virus from a cDNA clone and infection of citrus trees. Virology280, 87–96.

hi, B.J., Ding, S.W., Symons, R.H., 1997. Plasmid vector for cloning infectious cDNAsfrom plant RNA viruses: high infectivity of cDNA clones of tomato aspermycucumovirus. J. Gen. Virol. 78, 1181–1185.

ang, Q., Mawassi, M., Li, P., Gafny, R., Sela, I., Tanne, E., 2003. Elimination ofGrapevine Virus A (GVA) by cryopreservation of in vitro-grown shoot tips ofVitis vinifera L. Plant Sci. 165, 321–327.

hou, Z.Sh., Dell’Orco, M., Saldarelli, P., Turturo, C., Minafra, A., Martelli, G.P., 2006.Identification of an RNA-silencing suppressor in the genome of Grapevine VirusA. J. Gen. Virol. 87, 2387–2395.