Viral infection enables phloem loading of GFP and long-distance trafficking of the protein: Viral...

transcript

Viral infection enables phloem loading of GFP andlong-distance trafficking of the protein

Gadi Peleg, Dikla Malter and Shmuel Wolf*

Institute of Plant Sciences and Genetics in Agriculture and Otto Warburg Minerva Center for Agricultural Biotechnology, The

Hebrew University of Jerusalem, Faculty of Agricultural, Food and Environmental Quality Sciences, PO Box 12, Rehovot 76100,

Israel

Received 8 February 2007; revised 7 March 2007; accepted 12 March 2007*For correspondence (fax +972 8 9462385; e-mail swolf@agri.huji.ac.il).

Summary

It is generally accepted that viral systemic infection follows the source-to-sink symplastic pathway of sugar

translocation. In plants that are classified as apoplastic loaders, the boundary between the companion cell–

sieve element (CC–SE) complex and neighboring cells is symplastically restricted, and the potential passage of

macromolecules between the two domains has yet to be explored. Transgenic tobacco plants expressing

green fluorescence protein (GFP) and cucumber mosaic virus (CMV)-encoded proteins fused to GFP under the

control of the fructose-1,6-bisphosphatase (FBPase) promoter were produced in order to localize the encoded

proteins in mesophyll and bundle sheath cells and to explore the influence of viral infection on the functioning

of plasmodesmata interconnecting the two domains. GFP produced outside the vascular tissue could

overcome the symplastic barrier between the CC–SE complex and the surrounding cells to enter the

vasculature in CMV-infected plants. Grafting of control (non-transgenic) tobacco scions to CMV-infected

FBPase-GFP-expressing root stocks confirmed that GFP could move long distances in the phloem. No

movement of the gfp mRNA was noticeable in this set of experiments. The ability of GFP to enter the

vasculature and move long distances was also evident upon infection of the grafting plants with other viruses.

These results provide experimental evidence for alteration of the functioning of plasmodesmata intercon-

necting the CC–SE complex and neighboring cells by viral infection to enable non-selective trafficking of

macromolecules from the mesophyll into the sieve tube.

Keywords: coat protein, cucumber mosaic virus, movement protein, plasmodesmata, tobacco.

Introduction

Following viral replication in infected plant cells, systemic

infection requires the short-distance cell-to-cell movement

of viral genomes and their long-distance transport through

the vascular system. This systemic transport is governed by

a series of mechanisms involving virus and plant factors

(Lazarowitz and Beachy, 1999; Lucas and Gilbertson, 1994;

Scholthof, 2005). From its initial entry into plant cells, the

virus’s infectious material is transported to neighboring cells

via plasmodesmata. The plasmodesmata, together with

phloem conductivity, create a symplastic continuity that

enables the traffic of macromolecules and viral genomes

between cells and organs. However, in order to move long

distances, viruses have to overcome symplastic restric-

tion(s) imposed by plasmodesma-mediated selective traf-

ficking that forms symplastic domains. Accumulation of

cowpea chlorotic mottle virus in virus-resistant soybean

plants (Goodrick et al., 1991) or tomato aspermy virus in

tomato plants (Thompson and Garcıa-Arenal, 1998) is evi-

dent in bundle sheath cells but not in phloem parenchyma

cells, indicating the presence of a symplastic barrier to viral

movement at this boundary. A more distinct domain is the

sieve element–companion cell (SE–CC) complex in the

phloem tissue. The extremely low abundance, and sugges-

ted altered function, of plasmodesmata interconnecting the

SE–CC complex and neighboring cells (Ding et al., 1996,

1998) raise questions regarding the mechanism by which the

viral cargo overcomes this barrier and enters into the sieve

tube system.

Cucumber mosaic virus (CMV) is a single-stranded

positive-sense RNA virus divided into three segments

ª 2007 The Authors 165Journal compilation ª 2007 Blackwell Publishing Ltd

The Plant Journal (2007) 51, 165–172 doi: 10.1111/j.1365-313X.2007.03128.x

(Palukaitis et al., 1992). RNA-1 and 2 encode the 1A and 2A

proteins involved in RNA replication. RNA-2b encodes a

suppressor of post-transcriptional gene silencing, and RNA-

3 encodes the 3a gene as well as the coat protein (CP) gene,

which is also translated from sub-genomic RNA. Each of the

five encoded proteins participates in viral movement, but

both 3a protein and CP have been shown to be indispens-

able for cell-to-cell and long-distance movement of CMV

(Canto et al., 1997). The CMV 3a protein has the capacity to

bind single-stranded RNA (Li and Palukaitis, 1996), can

interact with plasmodesmata to alter their size exclusion

limits, and facilitates the traffic of viral RNA from cell to cell

(Canto et al., 1997; Ding et al., 1995; Kaplan et al., 1995). It

was therefore defined as a movement protein (MP). Immu-

nolocalization studies of the CMV MP in infected Nicotiana

clevelandii plants revealed that, even in minor veins, the

protein localizes mainly to plasmodesmata interconnecting

the CCs and SEs (Blackman et al., 1998). Expression of CMV

MP fused to green fluorescent protein (GFP) under a CC-

specific promoter (from commelina yellow mottle virus)

established that the fused protein moves through the

plasmodesmata interconnecting the CCs and SEs (Itaya

et al., 2002). Interestingly, the ability of the fused protein to

move outside the vascular tissue was organ- or develop-

mental stage-specific, indicating that CMV MP can mediate

its cell-to-cell movement by interaction with plasmodesmal

factors in a specific manner.

Earlier studies indicated that functional CP is also essen-

tial for cell-to-cell (Kaplan et al., 1998; Nagano et al., 2001)

and long-distance (Taliansky and Garcia-Arenal, 1995)

movement, but its mode of operation has yet to be explored.

The requisite co-existence of CP and MP for the short- and

long-distance movement of CMV led to speculation that

their interaction is required for viral infection (Blackman

et al., 1998; Canto et al., 1997). As CMV virions were not

detected in plasmodesmata interconnecting CCs and SEs of

infected plants, it was suggested that the virus moves into

the SE as a ribonucleoprotein complex that contains the viral

RNA, CP and MP (Blackman et al., 1998).

Despite ample information on the influence of viral

movement on plasmodesmata interconnecting mesophyll

cells, the mode by which viral components interact with

plasmodesmata interconnecting the SE–CC complex with

neighboring cells is still unknown. One might expect that

loading of the virus into the phloem would be associated

with alteration of the functioning of these plasmodesmata.

However, it is still not clear which viral protein(s) are involved

in the modification of these specific plasmodesmata, and

whether such a modification influences the traffic of endog-

enous macromolecules between the two distinct domains.

In this study, we produced transgenic plants expressing

GFP and CMV-encoded proteins fused to GFP under the

control of the fructose-1,6-bisphosphatase (FBPase) promo-

ter, which is specific to mesophyll cells (Lloyd et al., 1991).

We found that, upon viral infection, GFP that is produced

outside the vascular tissue can overcome the symplastic

barrier between the SE–CC complex and the surrounding

cells to enter the vasculature and move long distances in the

phloem. Our findings suggest that viral infection disrupts the

selectivity of macromolecular trafficking into the phloem.

Results

Tissue-specific expression of free GFP, CMV-MP:GFP, and

CMV-CP:GFP in transgenic tobacco plants

The FBPase promoter was employed to express free GFP,

CMV-MP:GFP or CMV-CP:GFP in transgenic tobacco plants

outside the boundary of the plant vasculature (Figure 1a).

These transgenic plants provide an experimental system to

study the role of the viral proteins in cell-to-cell and long-

distance transport of macromolecules. Detailed PCR analy-

sis (representative analyses are presented in Figure 1b–d)

indicated that 6, 12 and 11 independent tobacco lines

contained the genes coding for free GFP, CMV-CP:GFP and

CMV-MP:GFP, respectively.

Tissue and cellular localization of the fused proteins

CMV-MP:GFP and CMV-CP:GFP and free GFP in transgenic

tobacco plants

Fluorescence microscopy indicated that, when expressed

under the control of the FBPase promoter, free GFP and both

CMV-MP:GFP and CMV-CP:GFP were localized predomin-

antly to mesophyll cells with no expression in the epidermis

(Figure 2). Free GFP was localized to the cytoplasm and

nuclei (Figure 2a,b), CMV-MP:GFP was specifically localized

to the cell periphery, with bright fluorescent dots indicating

targeting to the plasmodesmata (Figure 2c,d), and CMV-

CP:GFP was localized predominantly to the nuclei. Bright

circles within the nuclei indicated specific localization of

CMV-CP:GFP to the nucleolus (Figure 2e,f).

Interestingly, when expressed under the control of the

FBPase promoter, CMV-MP:GFP was also detected between

vascular parenchyma cells or between vascular parenchyma

and bundle sheath cells (Figure 3a), while no fluorescent

signals from CMV-CP:GFP were detected in any of the

phloem cells (Figure 3b).

Phloem loading of GFP in virally infected tobacco plants

Transgenic plants expressing free GFP, CMV-MP:GFP and

CMV-CP:GFP under the control of the FBPase promoter

were infected with CMV to determine whether systemic

viral movement would enable traffic of the various

proteins into the vascular system. CMV infection caused a

significant reduction (and often elimination) of GFP fluor-

escence in plants containing CMV-MP:GFP or CMV-CP:GFP

166 Gadi Peleg et al.

ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 51, 165–172

(Figure S1c). Bright fluorescence was observed in plants

expressing free GFP (Figure 4b). Perhaps the most signifi-

cant outcome, in this set of experiments, were the bright

fluorescent signals within the vascular system of plants

expressing free GFP under the control of the FBPase pro-

moter as a result of CMV infection (Figure 4c,d). These

results indicate that the viral infection significantly alters

the function of plasmodesmata interconnecting the SE–CC

complex with the neighboring cells, thereby enabling the

traffic of GFP from the mesophyll into the vascular system.

In this respect, it is interesting to note that GFP fluorescence

was also detected in the epidermis of FBPase-GFP-expres-

sing CMV-infected plants.

Grafting experiments were employed to verify that the

GFP present in the vasculature of CMV-infected plants can

move long distances within the sieve tubes. As indicated in

Figure 4(e,g) GFP fluorescence was detected in control

FBPase promoter

GFPCMV-MP

EcoRI/BamHI XbaI XbaIBamHI

GFPCMV-CP

EcoRI/BamHI SmaI SalIBamHI

EcoRI/BamHI XbaI

Figure 1. Tobacco transformation.

(a) Structure of the chimeric binary plant expression cassettes used for plant

transformation.

(b)–(d) PCR analyses of leaves from various transgenic N. tabacum lines. (b)

Transgenic plants containing gfp (lanes 1–5) and non-transgenic control line

(lane 6).

(c) Transgenic plants containing cmv-cp:gfp (lanes 1–3 and 5), a transgenic

line that does not contain the gene (lane 4) and non-transgenic control line

(lane 6).

(d) Transgenic plants containing cmv-mp:gfp (lanes 1–5) and non-transgenic

control line (lane 6). M, molecular marker.

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Figure 2. Cellular localization of free GFP (a, b), CMV-MP:GFP (c, d), CMV-

CP:GFP (e, f), in transgenic plants expressing the various gene constructs

under the control of the FBPase promoter, compared with control non-

transgenic plants (g, h).

GFP expression is demonstrated in spongy mesophyll (a, c, e, g) and palisade

(b, d, f, h) cells of young mature leaves. Fluorescence was detected and

imaged by confocal laser scanning microscopy. Bar = 10 lm.

Viral infection and phloem loading of proteins 167

tobacco scions, which did not contain gfp, when grafted

onto CMV-infected transgenic plants expressing GFP under

the control of the FBPase promoter. No fluorescence signal

was detected in a control tobacco plant grafted onto FBPase-

GFP non-infected transgenic rootstock (Figure 4f,h).

To further explore the general influence of viral infection

on phloem loading and long-distance transport of GFP,

similar grafting plants were inoculated with either tobacco

mosaic virus (TMV) or potato virus Y (PVY). Western blot

analyses confirmed the presence of GFP in the scion (sink

leaves) of control non-transgenic plants grafted onto

FBPase-GFP rootstocks infected with TMV, whereas GFP

was absent from control scions grafted onto healthy FBPase-

GFP-expressing rootstocks (Figure 5). GFP was also detec-

ted in control (non-transgenic) scions following infection of

FBPase-GFP stocks with PVY (data not shown).

The presence of GFP in the scion of control non-transgenic

tobacco plants may result from movement of the protein

translated in the rootstock expressing the gene, or from

traffic of the gfp mRNA. The shoot apex, upper stem

segment and young sink leaf of the control scions were

subjected to RT-PCR analyses to explore whether gfp

transcripts might be moving long distances. As indicated

in Figure 6, gfp transcripts were not detected in any tissue of

the control scions, indicating that the gfp mRNA cannot

move long distances from the transgenic rootstock.

Discussion

To explore the influence of viral proteins on phloem loading

of macromolecules, expression of the target molecule out-

side the vascular tissue must be assured. A tissue-specific

promoter (FBPase) was employed to express GFP and CMV-

encoded MP and CP fused to GFP as markers for phloem

loading and long-distance trafficking of macromolecules.

When expressed in transgenic tobacco plants under the

control of the FBPase promoter, all three proteins were

detected mainly in the spongy mesophyll and palisade cells

of source leaves. This tissue-specific expression is in agree-

ment with an earlier study in which the GUS reporter gene

was expressed under the control of the same promoter in

tobacco plants (Lloyd et al., 1991). In that study, GUS

expression was evident in photosynthetically active and

meristematic cells. As expected, fluorescence decoration of

the cell wall was observed in plants expressing CMV-MP:GFP

(Figure 2c,d), suggesting that the fused protein was targeted

to plasmodesmata. The specific localization (fluorescence) of

the fused protein on cell walls interconnecting neighboring

cells and not those facing air spaces (Figure 2c) provided

further support for interaction of the fused protein with

plasmodesmata. In contrast, CMV CP was localized exclu-

sively to the nuclei. Bright fluorescence signals in the center

of the nuclei suggested that the CP is localized specifically to

the nucleolus (Figure 2e). Targeting of virus-encoded CPs to

the nucleolus has been observed in tobacco plant cells

infected with tobacco etch virus (Langenberg and Zhang,

1997), in N. benthamiana epidermal cells infected with a

TMV derivative expressing the GFP-tagged ORF3 protein

(CP-like protein) of groundnut rosette virus (Kim et al., 2004),

and in cells infected with CMV (Lin et al., 1996). Mutational

analysis identified a particular sequence within the potato

leafroll virus CP that is involved in nucleolar targeting (Haupt

et al., 2005), while micro-injection and transient expression

experiments have demonstrated localization of the tomato

yellow leaf curl virus CP fused to GFP to nucleoli of

N. benthamiana mesophyll cells (Rojas et al., 2001). It has

been proposed that CP is involved in the shuttling of viral

nucleic acids between the nucleoplasm and the cytoplasm;

however, the biological significance of its nucleolar local-

ization is still unknown. Functional analyses of mutants in

the nucleolus localization domain of the groundnut rosette

virus ORF3 protein have indicated a correlation between its

nucleolar localization and its ability to form cytoplasmic

ribonucleoprotein particles and transport viral RNA long

distances via the phloem (Kim et al., 2004). These latter

results suggest involvement of the nucleolus in the control of

long-distance transport of RNA and possibly proteins.

Movement of free GFP between epidermal cells and from

the epidermis to mesophyll cells has been observed follow-

ing biolistic bombardment of plasmid encoding the gene

(a) (b) Figure 3. Cross-sections of third-order veins of

mature transgenic tobacco leaves expressing

CMV-MP:GFP (a) or CMV-CP:GFP (b) under the

control of the FBPase promoter.

In CMV-MP:GFP-expressing plants (a), GFP fluor-

escence is seen in the vascular parenchyma cells

or between phloem parenchyma and bundle

sheath cells (see arrows), compared with ab-

sence of the CMV-CP:GFP in the vascular area (b).

Typical autofluorescence is detected in xylem

elements. Fluorescence was detected and im-

aged by confocal laser scanning microscopy.

Bar = 10 lm. BS, bundle sheath; C, companion

cell; P, phloem parenchyma; S, sieve element; X,

xylem.

onto sink leaves (Itaya et al., 1998; Oparka et al., 1999).

During the sink-to-source transition, the capacity of the GFP

to move from cell to cell decreased substantially, and this was

associated with the developmental shift from simple to

branched forms of plasmodesmata (Itaya et al., 1998; Oparka

et al., 1999). However, when fused to either CMV-MP (Itaya

et al., 1998) or TMV-MP (Crawford and Zambryski, 2001), cell-

to-cell traffic of the 57 kDa fused protein was found in source

leaves as well. In the current study, localization of free GFP

and CMV-CP:GFP was restricted to mesophyll and bundle

sheath cells (Figures 2 and 3), but CMV-MP:GFP also decor-

ated the walls of the vascular parenchyma cells (Figure 3).

The presence of CMV-MP:GFP in vascular parenchyma cells

suggests that the MP can also interact with and move

through plasmodesmata interconnecting the bundle sheath

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Figure 4. Influence of CMV infection on the transport of GFP into the vascular

tissue in transgenic tobacco plants expressing GFP under the control of the

FBPase promoter.

(a) Infected (left) and healthy (right) young source leaves.

(b) GFP fluorescence in spongy mesophyll cells of infected plants.

(c, d) Cross- and longitudinal sections, respectively, of a third-order vein of a

young mature leaf infected with CMV. Arrows indicate GFP fluorescence in

companion cells.

(e) GFP fluorescence in companion cells (see arrow) of a young leaf of control

non-transgenic tobacco scion grafted onto an infected FBPase:GFP-expres-

sing rootstock.

(f) No fluorescence signal, except for the typical autofluorescence of xylem

cells, was detected in a control non-transgenic tobacco scion grafted onto a

non-infected FBPase-GFP transgenic rootstock.

(g) GFP fluorescence in bundle sheath (see arrow) of a young leaf of a control

non-transgenic tobacco scion grafted onto an infected FBPase:GFP-expres-

sing rootstock.

(h) No fluorescence signal was detected in a control non-transgenic tobacco

scion grafted onto a non-infected FBPase-GFP transgenic rootstock.

For (e)–(h), longitudinal sections were performed on the third-order vein of a

young leaf (number 2 from the shoot apex). Typical autofluorescence is

detected in xylem elements in (c), (d) and (f). Fluorescence was detected and

imaged by confocal laser scanning microscopy. C, companion cell; X, xylem.

Figure 5. Presence of GFP as indicated by immunoblots.

GFP is present in the sink leaf of a non-transgenic control scion (1) and the

source leaf of FBPase-GFP transgenic rootstocks (2) infected with TMV, and is

present in the source leaf of non-infected FBPase-GFP-transgenic rootstock (3)

but absent from the sink leaf of non-transgenic control non-infected scion (4).

Figure 6. Presence of the gfp transcript as indicated by RT-PCR analyses.

The gfp transcript is present in healthy (lane 1) and CMV-infected (lane 2)

source leaves of FBPase-GFP rootstock, but absent from the sink leaf, stem

segment and apex of control non-transgenic scions in non-infected plants

(lanes 3, 5 and 7, respectively) and CMV-infected plants (lanes 4, 6 and 8,

respectively). Lane 9 is a positive control comprising plasmid containing the

gfp gene. Lanes 10 and 11 are negative controls comprising reaction mixture

with no cDNA template or in which the reverse transcriptase was absent,

respectively. The presence of the actin gene was seen in all plant tissues

(lower panel). M, molecular marker.

and vascular parenchyma cells of source leaves. In this

respect, it is interesting to note that this boundary provides a

symplastic barrier for the trafficking of some viruses (Good-

rick et al., 1991; Thompson and Garcıa-Arenal, 1998).

CMV infection was induced to explore the influence of all

viral components on phloem loading of macromolecules.

GFP expression was evident in CCs of CMV-infected FBPase-

GFP-expressing plants (Figure 4c,d). The presence of GFP in

the phloem of plants expressing this gene under the control

of the FBPase promoter indicates that CMV infection disen-

gages the symplastic barrier between the SE–CC complex

and neighboring cells. Moreover, grafting experiments

established that, upon CMV infection, GFP could enter the

sieve tube and move long distances to a non-transgenic

control tobacco scion in which gfp was not present (Fig-

ure 4e,g). Phloem loading and long-distance movement of

GFP upon infection with viruses from distant families (TMV

and PVY) suggest that the ability to interact with and alter the

functioning of plasmodesmata interconnecting the SE–CC

complex and neighboring cells is a general characteristic of

plant viruses.

Interestingly, no GFP fluorescence was observed in CMV-

infected plants expressing either CMV-MP:GFP or CMV-

CP:GFP. It is logical to assume that elimination of GFP

fluorescence in these plants was due to silencing of the

virally encoded mRNA via a post-transcriptional mechanism

(Baulcombe, 1999). Inoculation of these plants with either

TMV or PVY did not cause a significant reduction in GFP

fluorescence (Figure S1). These results support the assump-

tion that the reduction in expression of the CMV-encoded

proteins fused to GFP is only due to CMV infection. Never-

theless, plants infected with CMV exhibited typical symp-

toms, indicating that the viral genes were not silenced

effectively. It may well be possible that partial silencing was

sufficient to prevent detection and monitoring of GFP

fluorescence, but was not sufficient to prevent rapid replica-

tion of the viral genome. On the other hand, we cannot rule

out the possibility that the disappearance of GFP fluores-

cence was due to another, as yet unknown, mechanism.

No movement of CMV-CP:GFP or CMV-MP:GFP outside

the domain in which they were expressed was observed in

PVY- or TMV-infected plants. One possible explanation for

the non-movement of the virally encoded proteins fused to

GFP may relate to their cellular localization. Targeting of

CMV-encoded proteins to the nucleolus or plasmodesmata

(in the case of CP or MP, respectively) may result in

anchoring of those proteins to the cellular organelles,

thereby preventing (or inhibiting) their cell-to-cell and

long-distance movement. In this case, cell-to-cell and long-

distance trafficking of non-specific proteins occurred for

cytosolic proteins only. Alternatively, we cannot rule out the

possibility that the size exclusion limit of plasmodesmata

interconnecting the phloem with neighboring cells is not

sufficient to allow phloem loading of the fused proteins.

Tobacco plants are classified as apoplastic loaders, which

are characterized by a small number of plasmodesmata

interconnecting the bundle sheath/phloem parenchyma

cells and the SE–CC complexes. Intercellular and inter-organ

communication are required for the plant to function as a

whole organism, and it is therefore logical to assume that the

role of these plasmodesmata is to enable the long-distance

exchange of information molecules (Oparka and Turgeon,

1999; Wolf and Lucas, 1997). On the other hand, plasmodes-

mata interconnecting these two domains must be extremely

selective with regard to the passage of macromolecules in

order to enable distinct cellular processes in specialized

tissues. Unique protein profiles in the phloem sap of

cucurbits (Walz et al., 2004) and Brassica napus (Giavalisco

et al., 2006) reflect the selectivity of protein transport into the

sieve tube. Our data indicate that this restricted communi-

cation pathway is exploited by plant viruses to establish

systemic infection, altering the functioning of plasmodes-

mata interconnecting the two boundaries and allowing non-

selective trafficking of endogenous macromolecules.

Using a 2D gel electrophoresis method, we recently

compared the protein profiles in the phloem sap of healthy

and CMV-infected melon plants (Malter and Wolf, unpub-

lished data). We were able to identify 10–20 additional

proteins in the phloem sap of CMV-infected plants, exhib-

iting an increase of 5–10% in their phloem sap protein

content compared with healthy plants. These results further

support the hypothesis that CMV infection releases the

plasmodesma-imposed restriction of macromolecule traf-

ficking. Nevertheless, one should remember that the altered

protein profile in CMV-infected melon plants may be due to

traffic of endogenous proteins from the CCs and not all the

way from the mesophyll into the sieve tube. The fact that

only a small number of additional proteins were found in the

phloem sap of infected melon plants suggests that the plant

maintains some degree of selectivity, even during viral

infection. A current study is aimed at characterizing the

nature of those proteins that are present in the sap of virally

infected versus healthy plants.

Experimental procedures

Reagents and enzymes

DNA restriction enzymes and all nucleic acid modification enzymeswere obtained from Boehringer Mannheim (http://www.roche.com/home.html), unless otherwise specified. Cefotaxime (claforan) waspurchased from Laboratoires Roussel (http://www.sanofi-aventis.com). All other reagents were obtained from Sigma Chemical Co.(http://www.sigmaaldrich.com/) and Bio-Rad (http://www.bio-rad.com/).

Plasmids and tobacco transformation

The gfp (Chiu et al., 1996) was constructed by PCR using primers5¢-GGATCCGGACCTCCTCCTGGAATGGTGAGCAAGGGCGAG-3¢ and

5¢-GAGCTCTTACTTGTACAGCTCGTCC-3¢, thus adding a spacer(Gly-Pro-Pro-Pro-Gly) to the gfp gene, and BamHI and SacI sites atthe 5¢ and 3¢ ends of the gene, respectively. The cmv-mp and cmv-cpgenes were fused to the gfp gene using PCR as follows. The geneswere obtained by PCR from the RNA-3 clone of CMV (Fny strain).cmv-mp was excised using primers 5¢-CCATGGCTTTCCAAGGTAC-3¢ and 5¢-GGATCCAAGACCGTTAACCACC-3¢, thus creating NcoI andBamHI sites at the 5¢ and 3¢ ends of the gene, respectively. cmv-cpwas excised using primers 5¢-AAAACCATGGACAAATCTGAATCA-3¢and 5¢-AAAAGGATCCGACTGGGAGCACTCCA-3¢, also creating NcoIand BamHI sites at the 5¢ and 3¢ ends of the gene, respectively.

The three constructs were first fused into the bacterial expres-sion vector pET29a (Novagen, http://www.clinafa.com/html/NVG/home.html) at the NcoI, BamHI and SacI sites, and expressed inEscherichia coli BL21 as CMV-MP:GFP or CMV-CP:GFP fusionproteins. The fusion proteins were examined by SDS–PAGE andfluorescence microscopy (data not shown).

The amplified gfp was digested with BamHI and HindIII at the 5¢and 3¢ ends, respectively, and was ligated into pBluescript (pBS) atthe same restriction sites. The amplified cmv-mp:gfp gene wasdigested with XbaI and SalI at the 5¢ and 3¢ ends, respectively, andligated into pBS at the same restriction sites. The amplified cmv-cp:gfp gene was digested with SmaI and XbaI at the 5¢ and 3¢ ends,respectively, and then ligated into pBS at the same restriction sites.The three genes were inserted into the binary vector pBinPlus (pBP;van Engelen et al., 1995) containing the FBPase promoter (Lloydet al., 1991), kindly provided by Professor Sonnewald (Lehrstuhl furBiochemie, Friedrich-Alexander Universitat Erlangen-Nurnberg,91058, Erlangen, Germany). The gfp gene was inserted as an XbaI/SalI fragment, cmv-mp:gfp was inserted as an XbaI/XbaI fragment,and cmv-cp:gfp was inserted as a SmaI/SalI fragment.

The three binary vectors containing the genes coding for GFP,CMV-MP:GFP and CMV-CP:GFP were transformed into Agrobacte-rium strain LBA4404. Transgenic plants were produced by conven-tional methods (Horsch et al., 1985). Briefly, leaf discs of Nicotianatabacum (Xanthi nc) were inoculated with the recombinant Agro-bacterium cells, using kanamycin (300 mg l)1) and claforan(500 mg l)1) as selective agents. GFP-expressing plants were selec-ted using a fluorescent microscope (model BH-2; Olympus, http://www.olympus-global.com/). Rooted transformants were grownunder sterile conditions at 25�C, and then removed to soil mixturein a greenhouse with natural sunlight.

Grafting experiments and virus inoculation

Non-transgenic tobacco plants were grafted onto rootstocks (with5–7 source leaves) of FBPase:GFP transgenic tobacco plants (Fig-ure S2). Seven to 10 days after grafting, the rootstock’s youngestsource leaf was inoculated with CMV (strain Fny) using carborun-dum as an abrasive and 500 ll inoculum containing 40 lg virus.Control grafts, including non-transgenic rootstock, were subjectedto a similar inoculation procedure with H2O. Additional virusinfections included TMV and PVY using similar procedures.

Confocal laser scanning microscopy

For cross-sections, cut leaves were embedded in 7% low-meltingagarose (Conda, http://www.condalab.com), after which they weresectioned (60–80 lm width) using a vibratom (VT1000S, Leica,http://www.leica.com). GFP expression in phloem cells was visu-alized in cross-sections of third-order veins (the first-order vein isthe midrib which branches at frequent intervals to form second-order veins, and these are branched to form the third-order veins).

The pattern of GFP expression was detected by confocal laserscanning microscope (CLSM model LSM510; Zeiss, http://www.zeiss.com/) using blue laser (25 mW argon) excitation light(488 nm).

GFP detection gel blots

An alternative method for the identification of GFP in grafting plantswas based on Western blot analyses. Tissue samples were takenfrom young, healthy and virally infected sink leaves of control (non-transgenic) scions and from source leaves of FBPase:GFP trans-genic tobacco rootstocks.

Plant tissue (1:4 w/w) was homogenized in extraction buffer (9 M

urea, 75 mM Tris buffer, pH 6.8, 4.5% v/v SDS and 7.5% v/v b-mercaptoethanol). After grinding, samples were left at roomtemperature for 30 min, boiled for 5 min, and then centrifuged at10 000 g for 10 min. The supernatant was brought to a finalconcentration of 10% v/v glycerin. Equal amounts of protein wereseparated on 12% SDS–polyacrylamide gels containing 0.3% bis-acrylamide at 200 V and 18 mA for 2 h. Proteins were transferred toa nitrocellulose membrane and GFP was detected using polyclonalanti-GFP antiserum (1:100; Clontech, http://www.clontech.com/).

RNA isolation and RT-PCR analyses

Total RNA was extracted from 100 mg of the shoot apex, stemsegments and a young sink leaf of CMV-infected and healthy control(non-transgenic) tobacco scions of the same age, using Tri-reagent(Sigma) according to the manufacturer’s recommendations. TotalRNA extracted from source leaves of FBPase:GFP transgenic to-bacco rootstocks served as a positive control. RNA was quantifiedusing an RNA/DNA analyzer (GeneQuant II, Pharmacia Biotech,http://www.apbiotech.com). From 2 lg total RNA per sample,cDNA was prepared using a 22-mer-encoded oligo(dT) primerand Superscript II reverse transcriptase (Lifetechnologies, http://www.gentaur.com).

Two specific primers within the conserved region of the gfp, 5¢-AAAATGGTGAGCAAGGGCGAG-3¢ and 5¢-AAAGTACAGCTCGTCCATGCC-3¢, forward and reverse, respectively, were used for theidentification of gfp mRNA in the various plant tissues. For the sameset of cDNA samples, another PCR analysis was performed for thehousekeeping gene actin, serving as an internal probe for the mRNAlevel. Parallel RT-PCR analyses were performed, including controlsin which the reverse transcriptase was omitted to verify thatgenomic template was absent from the reaction mixture.

Acknowledgements

This paper is a contribution from the Uri Kinamon Laboratory.G.P. and D.M. were supported by a scholarship from theKinamon Foundation. Support of this project by the GermanFederal Ministry of Education and Research (BMBF) within theframework of German–Israeli Project Cooperation (DIP) is gratefullyacknowledged.

Supplementary Material

The following supplementary material is available for this articleonline:Figure S1. Cellular localization of CMV-CP:GFP in transgenictobacco plants expressing the gene construct under the control ofthe FBPase promoter following viral infection.

Figure S2. Graft union of a control Xanthi tobacco plant ontransgenic rootstock expressing GFP under the control of theFBPase promoter.This material is available as part of the online article from http://www.blackwell-synergy.com.

References

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

Blackman, L.M., Boevink, P., Santa Cruz, S., Palukaitis, P. and

Oparka, K.J. (1998) The movement protein of cucumber mosaicvirus traffics into sieve elements in minor veins of Nicotianaclevelandii. Plant Cell, 10, 525–537.

Canto, T., Prior, D.A.M., Hellwald, K.H., Oparka, K.J. and Palukaitis,

P. (1997) Characterization of cucumber mosaic virus. IV. Move-ment protein and coat protein are both essential for cell-to-cellmovement of CMV. Virology, 237, 237–248.

Chiu, W.L., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H. and Sheen,

J. (1996) Engineered GFP as a vital reporter in plants. Curr. Biol. 6,325–330.

Crawford, K.M. and Zambryski, P.C. (2001) Non-targeted and tar-geted protein movement through plasmodesmata in leaves indifferent developmental and physiological stages. Plant Physiol.125, 1802–1812.

Ding, B., Li, Q., Nguyen, L., Palukaitis, P. and Lucas, W.J. (1995)Cucumber mosaic virus 3a protein potentiates cell-to-cell traf-ficking of CMV RNA in tobacco plants. Virology, 207, 345–353.

Ding, X.S., Shintaku, M.H., Carter, S.A. and Nelson, R.S. (1996)Invasion of minor veins of tobacco leaves inoculated with tobaccomosaic virus mutants defective in phloem-dependent movement.Proc. Natl Acad. Sci. USA, 93, 11155–11160.

Ding, X.S., Carter, C., Deom, M.C. and Nelson, R.S. (1998) Tob-amovirus and potyvirus accumulation in minor veins of inocula-ted leaves from representatives of the Solanaceae and Fabaceae.Plant Physiol. 116, 125–136.

van Engelen, F.A., Molthoff, J.W., Conner, A.J., Nap, J.P., Pereira, A.

and Stiekema, W.J. (1995) pBinplus: an improved plant trans-formation vector based on pBin19. Trans. Res. 4, 288–290.

Giavalisco, P., Kapitza, K., Kolasa, A., Buhtz, A. and Kehr, J. (2006)Towards the proteome of Brassica napus phloem sap. Proteo-mics, 6, 896–909.

Goodrick, B.J., Kuhn, C.W. and Hussey, R.S. (1991) Restricted sys-temic movement of cowpea chlorotic mottle virus in soybeanwith non necrotic resistance. Phytopathology, 81, 1426–1431.

Haupt, S., Stroganova, T., Ryabov, E., Hyon Kim, S., Fraser, G.,

Duncan, G., Mayo, M.A., Barker, H. and Taliansky, M. (2005) Nu-cleolar localization of potato leafroll virus capsid proteins. J. Gen.Virol. 86, 2891–2896.

Horsch, R.B., Fry, J.E., Hoffman, N.L., Eichholtz, D., Rogers, S.G. and

Fraley, R.T. (1985) A simple and general method for transferringgenes into plants. Science, 227, 1229–1231.

Itaya, A., Woo, Y.-M., Masuta, C., Bao, Y., Nelson, R.S. and Ding, B.

(1998) Developmental regulation of intercellular protein traffick-ing through plasmodesmata in tobacco leaf epidermis. PlantPhysiol. 118, 373–385.

Itaya, A., Ma, F., Qi, Y., Matsuda, Y., Zhu, Y., Liang, G. and Ding, B.

(2002) Plasmodesma-mediated selective protein traffic between‘symplasmically isolated’ cells probed by a viral movement pro-tein. Plant Cell, 14, 2071–2083.

Kaplan, I.B., Shintaku, M.H., Li, Q., Zhang, L., Marsh, L.E. and

Palukaitis, P. (1995) Complementation of virus movement in

transgenic tobacco expressing the cucumber mosaic virus 3agene. Virology, 209, 188–199.

Kaplan, I.B., Zhang, L. and Palukaitis, P. (1998) Characterization ofcucumber mosaic virus. V. Cell-to-cell movement requires capsidprotein but not virions. Virology, 246, 221–231.

Kim, S.H., Ryabov, E.V., Brown, J.W.S. and Taliansky, M. (2004)Involvement of the nucleolus in plant virus systemic infection.Biochem. Soc. Trans. 32, 557–560.

Langenberg, W.G. and Zhang, L. (1997) Immunocytology shows thepresence of tobacco etch virus P3 protein in nuclear inclusions. J.Struct. Biol. 118, 243–247.

Lazarowitz, S.G. and Beachy, R.N. (1999) Viral movement proteinsas probes for intracellular and intercellular trafficking in plants.Plant Cell, 11, 535–548.

Li, Q. and Palukaitis, P. (1996) Comparison of the nucleic acid- andNTP-binding properties of the movement protein of cucumbermosaic cucumovirus and tobacco mosaic virus. Virology, 216, 71–79.

Lin, N.S., Hsieh, C.E. and Hsu, Y.H. (1996) Capsid protein of cucumbermosaic virus accumulates in the nuclei and at the periphery of thenucleoli in infected cells. Arch. Virol. 141, 727–732.

Lloyd, J.C., Raines, C.A., John, U.P. and Dyer, T.A. (1991) Thechloroplast FBPase gene of wheat: structure and expression ofthe promoter in photosynthetic and meristematic cells of trans-genic tobacco plants. Mol. Gen. Genet. 225, 209–216.

Lucas, W.J. and Gilbertson, R.L. (1994) Plasmodesmata relation toviral movement within leaf tissues. Annu. Rev. Phytopathol. 32,387–411.

Nagano, H., Mise, K., Furusawa, I. and Okuno, T. (2001) Conversionin the requirement of coat protein in cell-to-cell movementmediated by the cucumber mosaic virus movement protein. J.Virol. 75, 8045–8053.

Oparka, K.J. and Turgeon, R. (1999) Sieve element and companioncells – traffic control centers in the phloem. Plant Cell, 11, 739–750.

Oparka, K.J., Roberts, A.G., Boevink, P., Santa Crus, S., Roberts, I.,

Pradel, K.S., Imlau, A., Kotlizky, G., Sauer, N. and Epel, B. (1999)Simple, but not branched, plasmodesmata allow the nonspecifictrafficking of proteins in developing tobacco leaves. Cell, 97, 743–754.

Palukaitis, P., Roossinck, M.J., Dietzgen, R.G. and Francki, R.I.B.

(1992) Cucumber mosaic virus. Adv. Virus Res. 41, 281–348.Rojas, M.R., Jiang, H., Salati, R., Xoconostle-Cazares, B.,

Sudarshana, R., Lucas, W.J. and Gilbertson, R.L. (2001)Functional analysis of proteins involved in movement of themonopartite begomovirus, Tomato Yellow Leaf Curl Virus.Virology, 291, 110–125.

Scholthof, H.B. (2005) Plant virus transport: motion of functionalequivalence. Trends Plant Sci. 10, 376–382.

Taliansky, M.E. and Garcia-Arenal, F. (1995) Role of cucumoviruscapsid protein in long-distance movement within the infectedplant. J. Virol. 69, 916–922.

Thompson, J.R. and Garcıa-Arenal, F. (1998) The bundle sheath–phloem interface of Cucumis sativus is a boundary to systemicinfection by tomato aspermy virus. Mol. Plant–Microbe Interact.11, 109–114.

Walz, C., Giavalisco, P., Schad, M., Juenger, M., Klose, J. and Kehr,

J. (2004) Proteomics of cucurbit phloem exudate reveals a net-work of defence proteins. Phytochemistry, 65, 1795–1804.

Wolf, S. and Lucas, W.J. (1997) Plasmodesmal-mediated plantcommunication network: implications for controlling carbonmetabolism and resource allocation. In A Molecular Approach toPrimary Metabolism in Higher Plants (Foyer, C.H. and Quick, P.,eds). London: Taylor and Francis, pp. 219–236.

Viral infection enables phloem loading of GFP and long-distance trafficking of the protein: Viral...

Documents