mpb2c, a microtubule-associated plant protein binds … · mpb2c via the yeast sos recruitment...

MPB2C, a Microtubule-Associated Plant ProteinBinds to and Interferes with Cell-to-Cell Transport ofTobacco Mosaic Virus Movement Protein1

Friedrich Kragler2,3, Mirela Curin3, Kateryna Trutnyeva, Andreas Gansch, and Elisabeth Waigmann*

Max F. Perutz Laboratories, University Departments at the Vienna Biocenter, Institute of MedicalBiochemistry, University of Vienna, Dr. Bohr-Gasse. 9, A-1030 Vienna, Austria

The movement protein of tobacco mosaic virus, MP30, mediates viral cell-to-cell transport via plasmodesmata. The complexMP30 intra- and intercellular distribution pattern includes localization to the endoplasmic reticulum, cytoplasmic bodies,microtubules, and plasmodesmata and likely requires interaction with plant endogenous factors. We have identified andanalyzed an MP30-interacting protein, MPB2C, from the host plant Nicotiana tabacum. MPB2C constitutes a previouslyuncharacterized microtubule-associated protein that binds to and colocalizes with MP30 at microtubules. In vivo studiesindicate that MPB2C mediates accumulation of MP30 at microtubules and interferes with MP30 cell-to-cell movement. Incontrast, intercellular transport of a functionally enhanced MP30 mutant, which does not accumulate and colocalize withMP30 at microtubules, is not impaired by MPB2C. Together, these data support the concept that MPB2C is not required forMP30 cell-to-cell movement but may act as a negative effector of MP30 cell-to-cell transport activity.

Plant viruses produce movement proteins requiredto mediate cell-to-cell spread of the viral genomicinformation via plasmodesmata (Lucas, 1995; Laz-arowitz and Beachy, 1999; McLean and Zambryski,2000; Tzfira et al., 2000; Haywood et al., 2002; Hein-lein, 2002). Tobacco mosaic virus (TMV), a virus witha (�) single-stranded RNA genome encodes one ofthe most extensively studied movement proteins,MP30. Several biochemical and cell biological fea-tures of MP30 were revealed: MP30 binds coopera-tively single-stranded nucleic acids (Citovsky et al.,1990), interacts with the tubulin- and actin-basedcytoskeleton (Heinlein et al., 1995; McLean et al.,1995), associates to the endoplasmic reticulum (ER;Heinlein et al., 1998; Reichel and Beachy, 1998; Masand Beachy, 1999; Gillespie et al., 2002), and increasesthe size exclusion limit of plasmodesmata (Wolf etal., 1989; Waigmann et al., 1994). In the currentmodel, MP30 is proposed to form complexes withgenomic TMV-RNA (vRNA) and to move thesevRNA-protein complexes from ER-associated sites ofsynthesis throughout the cell via the cytoskeletal net-work toward plasmodesmata. There, MP30 induces

an increase in plasmodesmal size exclusion limit andfacilitates the transport of the vRNA-MP30 complexthrough the enlarged plasmodesmal channels intoadjacent cells (Lazarowitz and Beachy, 1999; Tzfira etal., 2000; Haywood et al., 2002; Heinlein, 2002). Inline with this model, microtubules have been func-tionally implicated in guiding vRNA-MP30 com-plexes toward plasmodesmata (Mas and Beachy,2000; Boyko et al., 2000a, 2000b). However, an alter-nate, non-movement-promoting role of microtubulesin providing a route toward an MP30 degradationsite was suggested (Padgett et al., 1996; Mas andBeachy, 1999; Reichel and Beachy, 2000). This modelis supported by the observation that TMV was able tospread in tissues with functionally impaired micro-tubules and by studies performed with a mutantvariant of MP30, MPR3 (Toth et al., 2002) withstrongly decreased affinity to microtubules but in-creased cell-to-cell movement capacity (Gillespie etal., 2002).

The complex intra- and intercellular transport pro-cesses likely require interaction of MP30 with cellularfactors that assist in subcellular distribution and cell-to-cell movement. So far, a cell wall-associated pectinmethyl esterase (PME; Dorokhov et al., 1999; Chen etal., 2000; Tzfira et al., 2000) was shown to specificallyinteract with TMV-MP30. Also, binding studies ofMP30 with cell fractions enriched in plasmodesmalcomponents indicate the presence of potential recep-tor molecules mediating protein movement via plas-modesmata (Kragler et al., 1998). Recently, Nt-NCAPP1, an ER-localized protein, was found to beinvolved in the plasmodesmal transport pathway ofMP30 (Lee et al., 2003). Furthermore, a cell wall-associated kinase is known to phosphorylate the

1 This work was supported by the Austrian Science Foundation(grant nos. P12614 –MOB and Sfb17, project part 08 to E.W.). E.W.was supported by an Austrian Programme for Advanced Researchand Technology fellowship (APART 441) from the Austrian Acad-emy of Science.

2 Present address: Section of Plant Science, University of Cali-fornia Davis, One Shields Ave, Davis, CA 95616.

3 These authors contributed equally to this paper.* Corresponding author; e-mail [email protected]; fax

43–1– 4277–9616.Article, publication date, and citation information can be found

at www.plantphysiol.org/cgi/doi/10.1104/pp.103.022269.

1870 Plant Physiology, August 2003, Vol. 132, pp. 1870–1883, www.plantphysiol.org © 2003 American Society of Plant Biologists www.plantphysiol.orgon August 29, 2018 - Published by Downloaded from Copyright © 2003 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

MP30 (Citovsky et al., 1993) thereby modulating itscell-to-cell movement activity (Waigmann et al.,2000).

For a number of movement proteins from otherviruses, interacting cellular factors such as a rabreceptor-like protein, Hsp40 (DnaK) family members,a homeodomain protein, and a transcriptional co-activator were recently identified (Soellick et al.,2000; Huang et al., 2001; Lin and Heaton, 2001; Des-voyes et al., 2002; Matsushita et al., 2002). The func-tions of these interesting movement protein interac-tion partners with respect to cell-to-cell movementhave yet to be elucidated.

Here, we report the identification of a previouslyuncharacterized MP30-interacting plant protein,MPB2C, from the TMV host plant Nicotiana tabacum.MPB2C associates with microtubules in a punctuatemanner and colocalizes with MP30. In functionalstudies, MPB2C mediates MP30 subcellular redistri-bution to microtubule-associated sites and selectivelyinhibits MP30 cell-to-cell movement. A potential roleof MPB2C as a negative effector of MP30 transportactivity is discussed.

RESULTS

Identification of MP30-Interacting FactorMPB2C via the Yeast Sos Recruitment System (SRS)

The partially hydrophobic nature of MP30(Citovsky et al., 1990; Brill et al., 2000) and its asso-ciation to cell compartments such as the ER mem-brane, cytoskeleton, and plasmodesmal receptors (forreview, see Haywood et al., 2002; Heinlein, 2002)indicate that a membrane-based interaction screeningsystem may be more suitable to identify MP30-interacting proteins than the conventional two-hybrid system. The yeast SRS (Aronheim et al., 1997)allows detection of interaction of proteins at theplasma membrane in Brewer’s yeast (Saccharomycescerevisiae; Aronheim et al., 1997; Yamanaka et al.,2000). Because the membranous environment pro-vided by the SRS might be more compatible with theobserved cellular distribution of MP30, this yeast-based interaction screening system was chosen toidentify plant-produced MP30-binding proteins.

In the SRS, the bait protein is fused to the human Rasguanyl nucleotide exchange factor (hSos), which com-plements the temperature-sensitive (ts) cdc25-2 yeastgrowth phenotype upon recruitment to the plasmamembrane. Prey library proteins are N-terminallylinked to a myristoylation signal providing plasmamembrane localization. Bait-prey interaction results inrecruitment of hSos to the plasma membrane and leadsto growth of yeast colonies at the restrictive tempera-ture (Aronheim et al., 1997). Library protein productionis regulated by a Gal-dependent promoter, which al-lows controlled expression of library protein and veri-fication of specific interactions by observing Gal-dependent growth of the yeast colonies.

First, a bait MP30-Sos fusion protein expressed bythe yeast shuttle vector ADNS-MP (Fig. 1A, topscheme) was constructed. To ensure that MP30-Sosby itself was not able to complement, or to interferewith complementation of, the cdc25-2 ts phenotype,cells were cotransformed with ADNS-MP30 andempty library vector YESMpolyA. No yeast colonieswere formed at the restrictive temperature (Fig. 1B,top panels). Thus, MP30 did not mediate hSos local-ization to the plasma membrane in yeast. SelectiveGal-dependent growth was observed when cellswere cotransformed with positive control vectorYESM7 encoding an hSos-interacting protein andMP30-Sos-expressing bait plasmid (Fig. 1B, panels insecond row). This functional test indicates that theMP30-hSos fusion protein is stable, can be translo-cated to the plasma membrane, and is able to activatethe Ras pathway. Thus, MP30 fused to hSos wassuitable to be used as bait in a SRS screen.

Next, directional cDNA derived from poly(A�)-RNA obtained from leaf, stem, shoot apex, andflower tissue of the natural TMV host plant N. taba-cum cv Turkish was used to construct a myristoyl-ation signal-tagged expression library (Fig. 1A, bot-tom scheme). To isolate cDNAs, which encodeinteraction partners of MP30, approximately 106 in-dependent cDNA clones were screened. A novelcDNA named movement protein binding 2C, library(MPB2Clib) coding for a 124-amino acid protein (Fig.3A) was selected for further analysis.

Specific interaction between MPB2Clib and MP30-Sos was apparent by Gal-dependent growth of cells

Figure 1. SRS screening for MP30 interaction partners. A, Schematicrepresentation of bait and library constructs. B, Complementation ofthe cdc25 ts phenotype. Bait and prey proteins expressed in trans-formed cells are indicated to left and right of panels, respectively.Growth of yeast is compared on Gal- and Glc-containing media atthe restrictive temperature. Prey proteins are expressed only onGal-containing medium.

MPB2C Interferes with MP30 Cell-to-Cell Transport

Plant Physiol. Vol. 132, 2003 1871 www.plantphysiol.orgon August 29, 2018 - Published by Downloaded from Copyright © 2003 American Society of Plant Biologists. All rights reserved.


co-expressing MPB2Clib and MP30-Sos at the restric-tive temperature (Fig. 1B, panels in third row). Incontrast, upon co-expression of MPB2Clib with a non-relevant bait protein, p110-Sos, no growth of yeastcolonies was observed (Fig. 1B, bottom panels).These complementation tests confirmed the specific-ity of interaction between MP30 and MPB2Clib. Inter-estingly, MPB2Clib interacted with �1-58MP30-Sos(Fig. 1B, panels in fourth row), a truncated form ofMP30 lacking the N-terminal 58 amino acids butretaining the region required for interaction with andmovement through plasmodesmata (Waigmann etal., 1994). The cDNA sequence of MPB2Clib was de-termined, and comparison of the predicted proteinagainst DNA and protein databases using theTBLAST algorithm (Pearson and Lipman, 1988) re-vealed that MPB2C is an uncharacterized plant pro-tein with structural properties similar to the myosin/kinesin super family (data not shown). Thus, we haveisolated the cDNA of a previously uncharacterizedtobacco protein, MPB2C, which specifically binds invivo to MP30 in the SRS interaction assay.

MPB2C and MP30 Interact in Vitro

Specificity of the yeast-based interaction studieswas confirmed in vitro by overlay assays. First, anAnti-Xpress-tag-MPB2Clib fusion protein of approxi-mately 20 kD, and, as a negative control, tag proteinalone with approximately 8 kD were produced inEscherichia coli and detected by western blotting us-ing Anti-Xpress-tag antibody (Fig. 2A, lanes 1 and 2).For overlay assays (Fig. 2B), purified MP30 was blot-ted onto nitrocellulose, and MP30 presence and sizewere verified with an antibody directed againstMP30 (Fig. 2A, lane 3). In parallel, part of the blotwas renatured and incubated with either total E. colilysate containing tag-MPB2Clib or tag protein alone.Addition of tagged MPB2Clib resulted in strong bind-

ing, which could be competed by the addition of anexcess of MP30 to the overlay mixture (Fig. 2B). Incontrast, no signal was detected in the control reac-tion with tag protein alone (Fig. 2B). To determinewhether MPB2C is capable of interacting with a viralmovement protein functionally related to MP30, animmobilized cucumber mosaic virus movement pro-tein (CMV3a; Li and Palukaitis, 1996) was used in aparallel overlay experiment. Here, no interaction ofMPB2Clib with CMV3a was observed (Fig. 2B). To-gether, these results underscore that MP30 andMPB2Clib specifically interact in vitro.

Sequence and Expression Analysis of MPB2C

By RACE techniques, full-length MPB2C cDNAwith a size of 1.2 kb was isolated. Upon northern-blotanalysis of poly(A�)-RNA from green tissue of N.tabacum with an MPB2C-specific probe (Fig. 3B, lane1), a band correlating to the calculated size of MPB2CcDNA was observed. As a positive control for RNAquality, a ubiquitin-specific probe was used (Fig. 3B,lane 2). Together, RACE and northern results indicatethat the full-length MPB2C cDNA was isolated.

Full-length MPB2C cDNA translates into a 321-amino acid protein (Fig. 3A) with a calculated molec-ular mass of 36 kD. MPB2C protein expression wasconfirmed by western blotting with an anti-MPB2Cantibody raised against purified E. coli-generatedMPB2C. When total protein extract derived from leaftissue was analyzed with the anti-MPB2C antibody, asingle MPB2C-specific band with a molecular mass ofapproximately 40 kD was detected (Fig. 3C, lane 1)that was not present upon incubation of the blot withpreimmune serum (Fig. 3C, lane 2), indicating that theMPB2C protein is constitutively expressed in leaf tis-sue. Note that the aberrantly high molecular mass ofthe MPB2C protein in denaturing polyacrylamide gelswas also observed for the recombinant E. coli MPB2Cprotein (data not shown).

Structural analysis of full-length MPB2C proteinrevealed a large coiled-coil region typical forcytoskeletal-associated factors (Lupas et al., 1991)stretching from amino acids 125 to 240 (Fig. 3A). Inaddition, a short hydrophobic region encompassingamino acids 44 to 52 was predicted (Kyte andDoolittle, 1982). When the predicted MPB2C proteinsequence was compared with the Arabidopsis data-base The Arabidopsis Information Resource byFASTA algorithm (Pearson and Lipman, 1988), onehomologous protein of similar size of 326 aminoacids (approximately 40% identity; gi:8346549), wasidentified. The carboxy-terminal part of the Arabi-dopsis MPB2C homolog was previously isolated in ascreen for plant cytoskeletal, cell cycle-related, andpolarity-related proteins (Xia et al., 1996). TBLASTNsearches (http://arabidopsis.org/Blast/) in nonre-dundant plant databases and subsequent analysiswith the ClustalW software (http://www.ebi.ac.uk/

Figure 2. MPB2C binds to MP30 in in vitro overlay assays. A,Verification of protein expression in E. coli: Tag-MPB2Clib (lane 1),tag protein alone (lane 2), and MP30 (lane 3) were detected bywestern blotting and labeling with Anti-Xpress-tag antibody (lanes 1and 2) or anti-MP30 antibody (lane 3). B, Detection of interactionbetween MPB2Clib and viral MPs by renatured blot overlay assay.Blotted MP30 and CMV3a were incubated with E. coli lysate con-taining tag-MPB2Clib, tag protein alone, and soluble MP30 as indi-cated in scheme. Binding was detected with Anti-Xpress-tagantibody.

Kragler et al.

1872 Plant Physiol. Vol. 132, 2003 www.plantphysiol.orgon August 29, 2018 - Published by Downloaded from Copyright © 2003 American Society of Plant Biologists. All rights reserved.


clustalw/; Thompson et al., 1994) highlighted a num-ber of uncharacterized cDNAs encoding significantlysimilar proteins from mono-and dicotyledonousplants with identity ranging from 75% to 37%. Theevolutionary relations between MPB2C from N. taba-cum and its various homologs from other plant specieswere compiled (Page, 1996) and depicted as a tree (Fig.3D). The related proteins fall into a dicotyledonousand a monocotyledonous clade, which reflects theevolutionary relation of the plant families. Interest-ingly, proteins with significant structural homology toMPB2C seem to be limited to the plant kingdom be-cause no related proteins with overall similarity above25% were found in databases of non-plant organismssuch as viruses, bacteria, yeasts, and animals.

MPB2C Localizes to Microtubules in aPunctuate Manner

To evaluate intracellular localization, full-lengthand N-terminally truncated variants of MPB2C lack-ing either the hydrophobic region (�1-66MPB2C), orin addition to the hydrophobic region, a large part of

the coiled-coil domain (�1-214MPB2C), were fused tored fluorescent protein (RFP; Matz et al., 1999) orgreen fluorescent protein (GFP; Chalfie et al., 1994;Table I), respectively. Fusion proteins were tran-siently expressed in epidermal cells of N. tabacumsource leaves and were analyzed for subcellular lo-calization patterns by confocal laser scanning micros-copy. The full-length MPB2C-RFP appears as smallpunctae and larger bodies (Fig. 4, A and D, arrowspoint to bodies, arrowheads to punctae; Table I).Variable relative amounts of MPB2C-RFP bodies andpunctae can be observed in individual cells rangingfrom primarily body-like phenotypes (Fig. 4A) overmixed phenotypes (Figs. 4D and 6A) to predomi-nantly punctuate phenotypes (Figs. 4G and 6D). Thetruncated �1-66MPB2C-RFP lacking the hydrophobicN-terminal domain localizes to filamentous struc-tures as well as to larger bodies and small punctae(Fig. 4B; Table I). Frequently, the bodies and punctaedecorated by the MPB2C-RFP fusion protein are inclose contact and appear to align along filamentousstructures (Figs. 4, B and G, and 6D). Similar local-ization patterns were observed for MPB2C-GFP and

Figure 3. Expression and sequence analysis ofMPB2C. A, Deduced amino acid sequence offull-length MPB2C. N-terminal hydrophobic re-gion (�), the coiled-coil domain (boxed),MPB2Clib sequence (underlined), and the startfor N-terminally truncated constructs (arrows)are indicated. B, Northern analysis of MPB2Cexpression. Tobacco poly(A�)-RNA was probedwith an MPB2C-specific probe (lane 1) or aubiquitin-specific probe as control (lane 2). C,Western-blot analysis of MPB2C expression. To-tal protein extracted from leaf tissue was incu-bated with anti-MPB2C-antibody (lane 1) or pre-immune antibody (lane 2). D, Phylogenetic treeobtained by MPB2C protein sequence compar-ison with translated cDNA sequences. The re-lated sequences fall into a dicotyledonous and amonocotyledonous clade, which reflects theevolutionary relation of the plant families. Notethat no significantly homologous sequence wasobserved in available non-flowering plants,fungi, animals, bacteria, and virus sequence da-tabanks. GenBank accession numbers of thecDNA sequences were At5g08120 (Arabidop-sis), AI898765 (tomato [Lycopersicon esculen-tum]), BE347146 (soybean [Glycine max]),BE186113 (maize [Zea mays]), BE356348 (sor-ghum [Sorghum bicolor]), AU094801 (Oryzasativum), and AL509409 (barley [Hordeum vul-gare]). Space bar represents 0.1 substitutions peramino acid.




�1-66MPB2C-GFP (data not shown). �1-214MPB2C-GFP lacks most of the coiled-coil structure in addi-tion to the hydrophobic domain and shows cytoplas-mic distribution equivalent to that observed for freesoluble GFP in epidermal cells (Fig. 4C; Table I).

The filamentous appearance of the �1-66MPB2C-RFP suggests association to microtubules. Therefore,colocalization experiments between MPB2C-RFP fu-sion proteins and microtubule-associated proteinMAP4-GFP (Marc et al., 1998) were performed. Thelarge bodies decorated by MPB2C-RFP (Fig. 4D, ar-rows) were also decorated by MAP4-GFP (Fig. 4E,arrows) and were in close contact with MAP4-GFP-decorated microtubules (Fig. 4F, arrows). Also,MPB2C-RFP appearing as small punctae (Fig. 4, D[arrowheads] and G) aligned at MAP4-GFP-decorated microtubules (Fig. 4, E and H) in themerged image (Fig. 4, F [arrowheads] and I). Simi-larly, the truncated �1-66MPB2C-RFP fusion proteinpredominantly decorated punctae and filaments thatcolocalized to filaments decorated by MAP4-GFP(Fig. 4, J–L).

Furthermore, to determine the subcellular localiza-tion of endogenous MPB2C, protoplasts derived fromleaf tissue were analyzed by immunofluorescence.Fixed protoplasts were incubated with anti-tubulinantibody and anti-MPB2C antibody (Fig. 5, A and B)or, as a negative control, with anti-tubulin antibodyand preimmune serum (Fig. 5C). Intracellular local-ization patterns were evaluated by confocal micros-copy. The MPB2C-specific signal was detected in thegreen channel and manifested as punctae that gener-ally aligned to or were in close contact with micro-tubular filaments detected in the red channel (Fig. 5,A and B, arrows). In contrast, in protoplasts incu-bated with preimmune serum, microtubule-associated punctae were not detected (Fig. 5C). Forcomparison, a high-resolution image of transientlyco-expressed MPB2C-RFP and MAP4-GFP is pro-

vided. Note that MPB2C-RFP punctae (Fig. 5D, ar-rows) align to MAP4-GFP-decorated microtubuleshighly similar to endogenous MPB2C.

Collectively, these studies indicate that endoge-nous MPB2C localizes in discrete punctae to micro-tubular filaments and that this punctuate localizationpattern is correctly reflected by transiently expressedMPB2C-RFP. However, in transient expression ex-periments, we additionally observe large MPB2C-RFP-decorated bodies that do not correctly reflect thelocalization of endogenous MPB2C and are likely aconsequence of the high expression level achieved inthe transient assay.

MPB2C Colocalizes to MP30 in Plant Cells

As a necessary prerequisite for interaction betweenMPB2C and MP30 in plant cells in vivo, both proteinsshould be present in the same subcellular compart-ment(s). In line, the filament and body-like appear-ance of MPB2C resembles described intracellular lo-calization patterns of MP30 (Heinlein et al., 1995,1998; McLean et al., 1995; Kahn et al., 1998; Mas andBeachy, 1999; Crawford and Zambryski, 2000; Boykoet al., 2000a, 2000b; Kotlizky et al., 2001).

Upon transient expression in epidermal cells of N.tabacum source leaves, MPB2C-RFP-decorated bodiesand punctae (Fig. 6, A and D) colocalize to MP30-GFP (Fig. 6, B and E) in the merged image (Fig. 6, Cand F; arrowheads in A, B, and C point to some of thecolocalizing punctae). In addition, an MP30-GFP sig-nal is also observed in small regular cortical punctae(Fig. 6B) reflecting ER-associated MP30-GFP(Gillespie et al., 2002), but these punctae are nottagged by MPB2C-RFP proteins (Fig. 6C). Upon co-expression with �1-66MPB2C-RFP, MP30-GFP is pre-dominantly detected at filaments that overlap withthose decorated by �1-66MPB2C-RFP (Fig. 6, G–I). Insummary, colocalization between MP30 and MPB2C

Table I. MPB2C constructs and intracellular localization pattern

Kragler et al.



Figure 4. Analysis of intracellular expression patterns of MPB2C. RFP and GFP fusion proteins were transiently expressedin epidermal cells of N. tabacum leaves by biolistic transformation and were analyzed by confocal microscopy after 48 hin the appropriate channels. Red color indicates RFP fluorescence; green color indicates GFP fluorescence. A, MPB2C-RFP;for better orientation, the cellular outline has been redrawn. B, �1-66MPB2C-RFP. C, �1-214MPB2C-GFP. D to I,Co-expression of MPB2C-RFP (D and G) with MAP4-GFP (E and H); F and I, merged image. J to L, Co-expression of�1-66MPB2C-RFP (J) with MAP4-GFP (K); L, merged image. Arrows point to bodies, arrowheads to punctae. Bars in A, G,and J � 20 �m; bars in B through D � 10 �m.




is observed at the level of microtubules but not whenMP30 is associated to the ER. This observation indi-cates that in vivo interaction between MP30 andMPB2C takes place at microtubules, in line with theendogenous localization of MPB2C.

Interestingly, in the presence of transiently ex-pressed MPB2C-RFP fusion proteins, MP30-GFPwas nearly exclusively localized to microtubule-associated sites and to the ER (Fig. 6, B, E, and H) andwas localized in only 4% of co-expressing cells to thecharacteristic cell wall-associated punctae, reflectingplasmodesmal association of MP30 (compare Fig. 7A;Crawford and Zambryski, 2000, 2001; Boyko et al.,2000a, 2000b; Kotlizky et al., 2001). In contrast, whenMP30-GFP was expressed in the absence of MPB2C,24% of expressing cells show the cell wall-associatedpunctuate expression pattern (compare Fig. 7A; K.Trutnyeva, R. Bachmaier, and E. Waigmann, unpub-lished data). A similar effect on MP30-GFP intracel-lular distribution was observed in Nicotiana benthami-ana: when MP30-GFP was expressed by itself, itlocalized to cell wall-associated punctae in 61% ofexpressing cells (pattern similar to Fig. 7A), whereasupon co-expression with MPB2C-RFP, the cell wall-associated pattern of MP30-GFP was detected in only26% of expressing cells. The reduction in cell wall-associated pattern was accompanied by a corre-sponding increase in the ER and microtubule-associated expression pattern of MP30-GFP in thepresence of MPB2C-RFP (pattern similar to Fig. 6, Band E). Thus, co-expression of MPB2C-RFP results inredistribution of MP30-GFP from the cell periphery

toward microtubule-associated sites, thereby furthersupporting the concept that MPB2C and MP30 inter-act at microtubules in vivo.

MPB2C Interferes with MP30 Cell-to-Cell Movement

Following biolistic bombardment of epidermalcells, MP30-GFP fusion proteins are known to movefrom cell to cell via plasmodesmata (Fig. 7A; Craw-ford and Zambryski, 2000, 2001; Kotlizky et al., 2001).Interestingly, upon co-expression of MP30-GFP andMPB2C-RFP or �1-66MPB2C-RFP constructs, noMP30-GFP cell-to-cell movement to adjacent cellscould be detected in more than 50 cells evaluated on10 independent leaves from different N. tabacumplants (Fig. 7B; compare also Fig. 6, B, E, and H).

Inhibition of intercellular transport of MP30-GFP incells expressing MPB2C-RFP proteins could be due tospecific interaction between MPB2C and MP30 orcould result from general down-regulation of plas-modesmal transport potentially induced by overex-pression of MPB2C proteins. To discriminate be-tween these possibilities, the viral movement proteinCMV3a, which is functionally related to MP30 butdoes not interact with MPB2C in vitro (compare Fig.2B), was employed. Like MP30-GFP, a GFP-taggedCMV3a fusion protein moves between cells (Fig. 7G;Itaya et al., 1997). Moreover, even in the presence ofthe MPB2C-RFP protein variants, CMV3a-GFP cell-to-cell movement was clearly detectable (Fig. 7, Hand I). Note that in co-expressing cells, CMV3a-GFPlocalization did not overlap with that of MPB2C-RFP

Figure 5. Endogenous MPB2C localizes inpunctae to microtubules. Leaf tissue-derivedprotoplasts were analyzed by immunofluores-cence and were observed with a confocal mi-croscope. A and B, Protoplasts incubated withanti-tubulin-antibody (red) and anti-MPB2C-antibody (green). Arrows point to MPB2C punc-tae. C, Protoplasts incubated with anti-tubulin-antibody (red) and preimmune serum (green). D,High-resolution image of MPB2C-RFP co-expressed with MAP4-GFP in an epidermal leafcell. Compare the punctuate MPB2C-RFP pat-tern (arrows) with that of endogenous MPB2C.Bar for A to D in A � 5 �m.

Kragler et al.



or �1-66MPB2C-RFP (Fig. 7, H and I, respectively).Together, these results imply that MPB2C selectivelyinterferes with intercellular transport of MP30.

The negative effect on cell-to-cell transport ofMP30-GFP might be due to improper amounts offluorescently tagged MPB2C proteins and/or theRFP tag which might mask a potential positive in-volvement of MPB2C in MP30 cell-to-cell transport.To further clarify the function of MPB2C in intercel-lular transport of MP30, we employed a L72V pointmutant of MP30 (MPR3; Gillespie et al., 2002; Toth etal., 2002) with enhanced cell-to-cell movement capac-ity in N. benthamiana. Because the MPR3 is primarilylocalized to ER-associated sites (Gillespie et al., 2002),we predicted that the MPR3 should not interact withthe microtubule-associated MPB2C protein. Conse-quently, transient expression of MPB2C-RFP proteinsshould not lead to accumulation of MPR3-GFP atmicrotubules and should not interfere with MPR3

cell-to-cell movement. Because the MPR3 mutant hasbeen characterized in N. benthamiana, we used thishost for quantitative, comparative movementstudies.

When MPR3-GFP was co-expressed with full-lengthMPB2C-RFP or �1-66MPB2C-RFP, the percentage ofcells with movement was only slightly reduced com-pared with the percentage observed in the controlexperiment where no MPB2C variants were present(Table II). In contrast, wild-type MP30-GFP move-ment was reduced to less than one-half in the pres-ence of MPB2C-RFP proteins (Table II). Furthermore,in the presence of co-expressed MPB2C-RFP proteins,the MPR3-GFP was not observed to accumulate atmicrotubules but retained its ER-associated localiza-

tion pattern (Fig. 7, D–F). Results indicate thatMPB2C and MPR3 are present in different subcellularcompartments, and thus do not interact in vivo. Con-sidering that the MPR3 protein is movement compe-tent, these data suggest that interaction with MPB2Cis not required for movement of MPR3.

We further tested whether in vitro binding be-tween MPB2C and MPR3 would be reduced as com-pared with binding between MPB2C and MP30. Forthis purpose, MPR3 and MP30 were overexpressedand purified from E. coli (Fig. 8A). Equal amounts ofpurified MPR3 and MP30 protein were blotted ontonitrocellulose, renatured, and overlayed with puri-fied MPB2C protein. Binding of MPB2C to MP30 orMPR3 was detected with an antibody directed againstthe MPB2C protein. Several molar ratios of MPB2C toMPR3 and MP30 were tested. When the molaramount of MPB2C exceeded that of MPR3 or MP302-fold, MPR3 showed slightly reduced binding toMPB2C as compared with MP30 (Fig. 8B, comparelanes MPR3 and MP30). When MPB2C was present ata 2.5-fold lower molar amount than MPR3 or MP30,binding of MPB2C to MPR3 was clearly reduced ascompared with MP30 (Fig. 8C, compare lanes MPR3

and MP30). Together, these results indicate that invitro MPB2C binds with lower affinity to the MPR3

protein than to the MP30 protein.

DISCUSSION

The frequently observed localization patterns ofMP30 comprising accumulation at ER-associatedpunctae, microtubules, large bodies, and cell wall-associated punctae have been combined in a model

Figure 6. MPB2C proteins colocalize withMP30-GFP. Analysis of expression and presen-tation of micrographs as described in Figure 4. Ato F, Co-expression of MPB2C-RFP (A and D)with MP30-GFP (B and E); C and F, mergedimage. G to I, Co-expression of �1-66MPB2C-RFP (G) and MP30-GFP (H); I, merged image.Arrowheads point to small punctae decoratedby MPB2C-RFP and MP30-GFP. Bars in A, D,and G � 20 �m.




for TMV cell-to-cell transport that involves transferof MP30 and viral RNA from the ER via a microtu-bular delivery mechanism toward and through plas-modesmata (Lazarowitz and Beachy, 1999; Tzfira etal., 2000; Haywood et al., 2002; Heinlein, 2002),thereby indicating a movement promoting role ofmicrotubules in transport of the MP30-vRNA com-

plex (Heinlein et al., 1995, 1998; McLean et al., 1995;Mas and Beachy, 2000; Boyko et al., 2000a, 2000b;Kotlizky et al., 2001). On the other hand, microtu-bules have also been implicated in MP30 degradation(Padgett et al., 1996; Mas and Beachy, 1999; Reicheland Beachy, 2000). In line, an MP30 mutant thatescapes accumulation at microtubules but displays

Figure 7. Cell-to-cell movement of viral MPs in epidermal leaf cells expressing MPB2C proteins. Analysis of expression andpresentation of micrographs as described in Figure 4, except that experiments depicted in C to F were done in epidermalleaf cells of N. benthamiana. A, Cell-to-cell movement of MP30-GFP in the absence of MPB2C-RFP. *, Expressing cell. B,Lack of cell-to-cell movement in cell co-expressing MP30-GFP and �1-66MPB2C-RFP. C, Cell-to-cell movement ofMPR3-GFP. D to F, Cell-to-cell movement of MPR3-GFP in cell co-expressing �1-66MPB2C-RFP. D, �1-66MPB2C-RFP; E,MPR3-GFP; F, merged image. G, Cell-to-cell movement of CMV3a-GFP. H and I, Cell-to-cell movement of CMV3a-GFP incells co-expressing MPB2C-RFP (H) or �1-66MPB2C-RFP (I). In B, H, and I, merged images are shown. Arrowheads pointto MP30-GFP, MPR3-GFP, or CMV3a-GFP that has moved into neighboring cell. Bar for A to C and H in H � 20 �m; barfor D to G and I in I � 40 �m.

Kragler et al.



enhanced cell-to-cell movement was recently de-scribed (MPR3; Gillespie et al., 2002), indicating thataccumulation at microtubules is detrimental forMP30 movement.

The complex MP30 localization pattern likely re-quires interaction with several plant endogenous fac-tors. Here, we describe the isolation and character-ization of a previously uncharacterized, plant-specific MP30-binding protein, MPB2C, that localizesin discrete punctae to microtubules. The C-terminalone-third of the 321-amino acid-long MPB2C is suf-ficient to bind MP30. The N-terminal part of MPB2Charbors a hydrophobic region and a coiled-coil do-main involved in association to microtubules. Asshown by northern and western analysis, MPB2C isconstitutively expressed in N. tabacum leaf tissue.Massively Parallel Signature Sequencing data on theArabidopsis homolog of MPB2C also demonstrateconstitutive expression at low levels throughoutplant tissues (http://allometra.com/mpss.shtml), in-dicating a general housekeeping function of MPB2C.

The localization of MPB2C to discrete punctaealong microtubules points to the presence of specifictarget sites on microtubules and is reminiscent ofpatterns observed for other microtubule-associatedproteins such as a kinesin-related motor protein (Caiet al., 2000) or microtubule- plus end-associated pro-teins, for example CLIP-170 (Perez et al., 1999). Inter-estingly, the localization of MPB2C to discrete punc-tae shifts toward a more continuous decoration ofmicrotubules when its 66 N-terminal amino acids aredeleted (compare Fig. 4, A and B). Similar observa-tions with CLIP-170 deletion mutants were inter-preted as evidence for regions in the CLIP-170 pro-tein that control and regulate its binding tomicrotubules (Pierre et al., 1994). In line, theN-terminal region of MPB2C might also exert a mod-ulating function on microtubule binding of MPB2C.

What may be the nature of large bodies that aredetected in addition to microtubule-associated punc-tae upon transient expression of MPB2C? In co-expression experiments, MPB2C bodies are alsotagged by MP30-GFP and MAP4-GFP and form inclose proximity to the microtubular network (com-pare Fig. 4, D–F), suggesting that these bodies maycontain a mixture of microtubule-associated proteins.Interestingly, microtubule-connected bodies are also

observed upon overexpression of CLIP-170-GFP(Perez et al., 1999) and tubulin-GFP (F. Kragler, un-published data) and upon expression of MP30-GFPeither by itself (Kotlizky et al., 2001; K. Trutnyevaand E. Waigmann, unpublished data) or in the con-text of the virus (Gillespie et al., 2002), suggestingthat their formation might represent a general re-sponse to high-level expression of microtubule-associated proteins.

In transient expression studies, both full-lengthMPB2C and �1-66MPB2C colocalize to MP30 at thelevel of microtubules and effectively interfere withMP30 localization to the characteristic punctuate pat-tern, indicating plasmodesmal localization (compareFig. 7A). In contrast, MPB2C neither colocalizes toER-associated MP30 nor was it observed to interferewith MP30 ER localization. This is in line with theview that MP30 is located at the ER before its transferonto microtubules or transport toward plasmodes-mata (Gillespie et al., 2002; K. Trutnyeva, R. Bach-maier, and E. Waigmann, unpublished data).

Functional analysis by transient expression studiesrevealed a specific, MPB2C-mediated decrease inMP30 cell-to-cell movement, suggesting a negativeimpact of MPB2C on MP30 transport activity. Thenegative effect of MPB2C on MP30 transport activitywas analyzed in two different host plants and variedfrom a complete block in N. tabacum to a more than50% reduction in N. benthamiana. This host-dependent difference may be correlated to the well-known high susceptibility of N. benthamiana for viralinfection (Gibbs et al., 1997) and the less tightly con-trolled plasmodesmal transport activity of N.benthamiana as compared with N. tabacum (Waig-mann et al., 2000).

Further support for a non-movement-promotingrole of MPB2C was derived from analyzing the im-pact of MPB2C proteins on intercellular transportactivity of MPR3, an L72V point mutant of MP30 withimproved cell-to-cell movement capacity and de-creased affinity for microtubules (Gillespie et al.,2002; Toth et al., 2002). Cell-to-cell transport of MPR3-GFP was not impaired by the presence of MPB2C

Figure 8. Comparison of MP30 and MPR3 binding to MPB2C. A,Western blot of purified MPR3 and MP30 detected by anti-MP30antibody. B and C, Renatured blot overlay assay. Twenty-three pi-comoles of MP30 and MPR3 was overlayed with 45 pmol of MPB2C(B) or 9 pmol of MPB2C (C) in the overlay mixture. Binding wasdetected by anti-MPB2C antibody. Arrow points to MP30 or MPR3,respectively.

Table II. Effect of MPB2C on movement of MP30-GFP andMPR3-GFP in N. benthamiana

Bombarded Plasmid DNACells withMovement

No. of AnalyzedCells

%

MP30-GFP 62.7 43MP30-GFP/MPB2C-RFP 28.6 42MP30-GFP/�1-66MPB2C-RFP 16.3 43MPR3-GFP 60.3 53MPR3-GFP/MPB2C-RFP 55.1 49MPR3-GFP/�1-66MPB2C-RFP 58.0 50




proteins (Table II). Moreover, even in the presence ofMPB2C proteins, MPR3-GFP was located at the ER(Fig. 7, D–F; Gillespie et al., 2002) and was not redis-tributed to microtubules. These results suggest thatMPR3 and MPB2C do not encounter each other inliving cells and because the MPR3 protein is highlymovement competent (Table II; Gillespie et al., 2002),that interaction with MPB2C is not required for MPR3

cell-to-cell movement. In support, in vitro binding ofMPR3 to MPB2C is reduced when compared withbinding between MP30 and MPB2C (Fig. 8). Thisreduction in binding affinity, likely together withchanges in binding affinity to other cellular factors,may contribute to the difference in subcellular local-ization of MPR3 as compared with MP30 and to thelack of inhibition of MPR3 cell-to-cell movementin the presence of transiently expressed MPB2Cproteins.

Considering that the MPR3 protein and wild-typeMP30 are distinguished only by a point mutation, itis likely that the two proteins move cell-to-cell via asimilar mechanism, and that interaction with MPB2Cmight also not be required for cell-to-cell movementof wild-type MP30. Because the MPR3 protein is char-acterized by enhanced transport activity and becausewe observed an MPB2C-mediated decrease of MP30transport activity in transient expression studies, wepresently favor the hypothesis that endogenousMPB2C serves a negative role in MP30 cell-to-celltransport. However, to further elucidate and definethe function of MPB2C in cell-to-cell transport ofTMV-MP and in intercellular spread of TMV, it willbe necessary to generate plants with reduced endog-enous MPB2C levels such as anti-sense or silencedplants.

In line with our hypothesis, we propose a modelwherein endogenous MPB2C encounters and bindsto MP30 at microtubules and acts in a pathway thatpromotes MP30 accumulation at microtubules. Thispathway is not involved in MP30 cell-to-cell trans-port but might be assigned for degradation as previ-ously suggested (Padgett et al., 1996; Mas andBeachy, 1999; Reichel and Beachy, 2000; Gillespie etal., 2002). When MPB2C is present at endogenouslevels, a number of MP30 molecules produced ininfected cells escape the MPB2C involving microtu-bular accumulation pathway and engage in cell-to-cell transport. When MPB2C is transiently overex-pressed, the equilibrium is shifted toward theMPB2C pathway, and directs MP30 away from thecell-to-cell transport pathway toward accumulationat microtubules. Hence we observed a negative effectof MPB2C on MP30 movement in the transient over-expression assay.

In summary, we report the identification andcharacterization of MPB2C, a plant endogenous,microtubule-localized MP30 interaction partner. Fu-ture studies will be directed toward a detailed eval-uation of the impact of MPB2C on local and systemic

spread of TMV and to reveal the as yet unknownendogenous function of the MPB2C protein.

MATERIALS AND METHODS

Plant Growth Conditions

Six- to 8-week-old Nicotiana tabacum cv Turkish or Nicotiana benthamianaplants were used. Plants were grown in a controlled environment with acycle of 16 h of light at 22°C and 8 h of darkness at 20°C.

Bait Plasmid and SRS cDNA Library Construction

Conventional PCR cloning techniques were employed to construct theMP30-Sos expressing bait plasmids ADNS-MP30 and ADNS �1-58MP30.pET-MP30 DNA was amplified with primers FK63 (5�-GATCCCAAGCTTAGGATCCAGAAGATGGCTCTAGTTGTTAAAGG-3�)and FK70 (5�-CGGAGCCCGGGAAACGAATCCGATTCGGCG-3�) and wasligated as HindIII-XbaI fragments (partial HindIII digest for full-lengthMP30 open reading frame) into HindIII-XbaI digested ADNS 5�SOS vector(Aronheim et al., 1997). After sequencing, cdc25-2 yeast was transformedwith ADNS-MP30 and ADNS �1-58MP30, and MP30 expression was veri-fied with polyclonal rabbit MP30 antibody (data not shown). Poly(A�)-RNAwas isolated from the following tissues of N. tabacum cv Turkish plantsaccording to information supplied by the manufacturer (RNeasy plant minikit and Oligotex mRNA kit, Qiagen USA, Valencia, CA): leaves (sink andsource), inflorescence and shoot apex including young leaves (� 1 cm),stem, and whole young plants (4 cm high). A directional cDNA synthesiswas carried out (cDNA synthesis kit, Stratagene, La Jolla, CA), and theresulting EcoRI-XhoI cDNA fragments with sizes �200 bp were ligated withEcoRI-XhoI-digested YES�polyA plasmid (Aronheim et al., 1997) and trans-formed into Epicurian Coli XL1-Blue MRF� supercompetent Escherichia colicells following the manufacturer’s instructions (Stratagene). Transformed E.coli cells were grown in LB-Amp media (Ausubel et al., 1987), and cDNAplasmids were isolated (Plasmid Midi kit, Qiagen USA).

SRS Screen

Compositions of yeast (Saccharomyces cerevisiae) Gal and Glc containinggrowth media and transformation reaction techniques and growth condi-tions were employed as described (Ausubel et al., 1987; Aronheim et al.,1997) with the following exception: The thermosensitive yeast cdc25-2(Mat�, ura3-5, his3-200, ade2-101, lys2-801, trp1-109, leu2-3, cdc25-2, andGal-) strains transformed with library plasmid and various bait plasmidswere first incubated in a 21°C incubation chamber on selectable Glc plates(non-restrictive conditions). Subsequently, colonies of transformed cellswere replica plated in parallel onto plates containing Gal- or Glc-selectivemedium. Plates were incubated at 36C° (restrictive conditions) for �5 d toevaluate bait-prey interaction-mediated growth complementation of thecdc25-2 ts phenotype (Aronheim et al., 1997). Standard control plasmidsADNS-5�p110-SOS, YESM7, YESMpolyA, and YESM�polyA are described(Aronheim et al., 1997; SRS protocol provided by A. Aronheim to F.K.).

RNA Expression Analysis and RACE

Poly(A�)-RNA was isolated as described above, and northern analysis wascarried out using the NorthernMax-Gly Kit according to the manufacturer’sprotocol (Ambion, Austin, TX). An MPB2C-specific probe was produced byPCR of the library plasmid YES-MPB2Clib with primers FK76 and FK94. Theubiquitin-specific probe was prepared as described (Sablowski and Meyer-owitz, 1998). Probes were radioactively labeled with Random primed DNAlabeling kit according to the manufacturer’s protocol (Roche Diagnostics,Mannheim, Germany). Note that the hybridized blot was exposed five timeslonger compared with the ubiquitin probed membrane to observe the MPB2Csignal. To obtain full-length MPB2C cDNA, 5�-RACE reactions were carriedout following the instructions supplied by the manufacturer (Marathona

cDNA amplification kit; BD Biosciences Clontech, Palo Alto, CA) using prim-ers FK100 (5�-TTATGTTCTCAGAACAAGTGATTTTGCAGG-3�) and FK110(5�-ATCTCATGAGGATCCCCTGTTCTCAGAAC-3�).

Kragler et al.



Production of MPB2C Protein andAnti-MPB2C Antibodies

A glutathionine S-transferase (GST)-MPB2C expression construct was ob-tained by PCR amplification of MPB2C from 35S::MPB2C-RFP using primers5�-GCAGGATCCGATGACGATGACAAGATGGCGCTTGCATTTG-3� and5�-GCAGAATTCTTATGTTCTCAGAACAAG-3�. The resulting fragmentwas digested with BamHI/EcoRI and was cloned into BamHI/EcoRI-digested vector pGEX-6P-1 encoding a GST moiety followed by a PreScis-sion Protease cleavage site (Amersham Biosciences, Uppsala) to yield con-struct pGEX-MPB2C. To obtain soluble GST-MPB2C, BL21-CodonPlusbacteria transformed with pGEX-MPB2C were grown at 30°C until opticaldensity � 0.6. Induction was performed with 0.1 mm isopropyl-�-d-thiogalactopyranoside for 1 h at 15°C. Bacteria were collected by centrifu-gation and were lysed by sonication in phosphate-buffered saline (PBS;Ausubel et al., 1987) containing 1% (v/v) Triton X-100 and Protease Inhib-itor Cocktail Tablets (Roche Diagnostics, Germany). To obtain purifiedMPB2C protein, cleared supernatant was incubated with GlutathioneSepharose 4B (Amersham Biosciences) and was subsequently released fromthe matrix with PreScission Protease according to the manufacturer’s in-structions (Amersham Biosciences). Because purified MPB2C protein mi-grated aberrantly high on denaturing polyacrylamide gels (observed mo-lecular mass, approximately 40 kD; calculated molecular mass, 36 kD) theidentity of MPB2C was confirmed by N-terminal sequencing and massspectroscopy before antibody production. Polyclonal antibodies againstMPB2C were obtained by immunization of rabbit with 200 �g of purifiedMPB2C and a boost with 150 �g of purified MPB2C after 8 weeks.

Western-Blot Analysis of MPB2C Protein

N. tabacum leaf tissue was frozen and ground to powder in liquid N2.Proteins were extracted by shaking the powder in the same volume of 2�Laemmli sample buffer (Ausubel et al., 1987) for 30 min and boiling for 10min. The extract was cleared from insoluble material by ultracentrifugationat 35,000g for 1 h. Proteins were separated on denaturing polyacrylamidegels and were blotted onto nitrocellulose membrane. The blot was blockedin 5% (w/v) milk powder, 10 mm Tris-HCl, pH 7.5, and 150 mm NaCl(M-TBS) for 30 min, incubated for 2 h with anti-MPB2C antibodies orpreimmune serum at a dilution of 1:1,000 in M-TBS, washed three timeswith M-TBS, and then incubated for 1 h with alkaline phosphatase-coupledgoat anti-rabbit antibody (Pierce, Rockford, IL) at a dilution of 1:2,000 inM-TBS. After washing three times with M-TBS and Tris-buffered saline,the blot was developed with nitroblue tetrazolium chloride and5-bromo-4-chloro-3-indolyl-phosphate.

In Vitro Interaction of MPB2C and MP30

The MPB2C cDNA fragment isolated in the SRS-screen was PCR amplifiedwith specific primers FK76 (5�-CCGGCTCGAGAGTTAAGTTAGCT-GACAAGAAAGCAGC-3�) and FK94 (5�-CGGAATTCCGTTAAGTTA-GCTGACAAGAAAGCAGCGGTAG-3�) and was cloned as an XhoI/EcoRIfragment into XhoI/EcoRI-digested pRSETA to obtain the expression plas-mid pRSETA-�1-197MPB2C giving rise to protein tag-MPB2Clib. The MPR3

coding sequence was amplified by PCR from plasmid MPR3:GFP (Gillespieet al., 2002) using primers R3start (5�-GCACATATGGCTCTAGTTGTTAA-AGG) and R3back (5�-GCAGGATCCTTAAAAGGAATCCGATTCGGC) andwas cloned as an NdeI/BamHI fragment into NdeI/BamHI digested pET3Avector giving rise to expression plasmid pET-MPR3. MP30, MPR3, andCMV3a proteins were overexpressed in E. coli from plasmids pET-MP30(Citovsky et al., 1990), pET-MPR3, and pET-CMV3a (Li and Palukaitis, 1996),respectively. Inclusion body isolation, PAGE, and western blotting weredone as described (Citovsky et al., 1990). Tag-MPB2Clib was overexpressedfrom plasmid pRSETA-�1-197MPB2C, and western-blot analysis of tag-MPB2Clib was done according to the manufacturer’s protocol (Invitrogen,Carlsbad, CA). Conditions for renatured blot overlay assay are based onpreviously described dot-blot assays (Kragler et al., 1998). In brief, MP30and CMV3a were blotted onto nitrocellulose membranes, treated twice for15 min at room temperature with incubation buffer (40 mm HEPES, pH 6.8,and 10% [v/v] glycerol) containing 0.2% (v/v) Triton X-100 and CompleteProtease Inhibitor Cocktail Tablets (Roche Diagnostics). After washing threetimes and blocking the membrane with overlay buffer (incubation buffer

containing 0.1% [v/v] Tween 20 and 2% [w/v] bovine serum albumin),nitrocellulose stripes were overlaid with overlay buffer containing total E.coli lysate from cells expressing tag-MPB2Clib protein or the Anti-Xpress-tagprotein alone. For competition experiments, an approximately 50� molaraccess of purified MP30 over MPB2Clib was added to the overlay buffer.After 48 h of incubation at 10°C, washing five times with overlay buffer, andtreatment with incubation buffer containing 2% (v/v) diglutaraldehyde for1 min at room temperature, bound MPB2Clib was detected with commer-cially available Anti-Xpress-tag antibody following the manufacturer’s pro-tocol (Invitrogen). In experiments designed to compare the affinity of MP30and that of MPR3 with MPB2C, 0.7 �g corresponding to 23 pmol of MP30and MPR3 protein, respectively, was blotted onto nitrocellulose. A smallstrip encompassing the MP30 and MPR3 bands was cut out and incubatedwith either 45 or 9 pmol of MPB2C protein in overlay buffer. The experi-mental procedure for renatured blot overlay assay was as described above.Binding of MPB2C was detected with anti-MPB2C antibody at a dilutionof 1:2,000.

Protoplast Preparation and Immunofluorescence

Protoplasts were prepared as described previously (Koop et al., 1996) andwere fixed in 4% (w/v) paraformaldehyde (Science Services, Munich, Ger-many) and 5% (v/v) dimethyl sulfoxide in PBS for 1 h at room temperature.Fixed protoplasts were washed five times with PBS and were incubatedwith anti-MPB2C antibody or preimmune serum at a dilution of 1:20 andmouse anti-�-tubulin antibody (Molecular Probes, Eugene, OR) at a dilutionof 1:1,000 in PBS for 1 h. After washing five times with PBS, protoplasts wereincubated for 1 h with Alexa Fluor 488-coupled goat anti-rabbit and AlexaFluor 568-coupled goat anti-mouse secondary antibodies (Molecular Probes)diluted 1:1,000 in PBS. After washing three times in PBS, protoplasts weremounted on slides and analyzed by confocal microscopy.

Construction and Transient Expression of GFP and RFPFusion Proteins

The pUC18-based vectors pZY226 and pZY260 containing a 10� Alalinker at the C or N terminus of a 35S driven mGFP6, respectively, were agenerous gift to F.K. by D. Jackson (Cold Spring Harbor Laboratories, ColdSpring Harbor, NY). GFP expression vectors producing soluble smGFP(CD3–326; Arabidopsis Biological Resource Center, Ohio State University,Columbus), ER-GFP (Haseloff et al., 1997), MAP4-GFP (Marc et al., 1998),CMV3a-GFP (Itaya et al., 1997), MP30-GFP (McLean et al., 1995), andMPR3-GFP (Gillespie et al., 2002) were described previously. Plasmid35S::RFP was a generous gift from V. Citovsky (State University of NewYork, Stony Brook).

To obtain MPB2C-RFP fusion proteins with a separating Ala linkerbetween the two protein moieties, a 35S::GFP-RFP plasmid was con-structed first by cloning a SalI/XbaI-digested RFP PCR fragment amplifiedfrom pDsRed1-N1 (BD Biosciences Clontech) using primers FK111 (5�-GCCGTCGACATGGTGCGCTCCTCC-3�) and FK112 (5�-CGCCTCTA-GACTACAGGAACAGGTGG-3�) into SalI/XbaI-digested pZY260. ABamHI/XbaI fragment of the 35S::GFP-RFP plasmid containing the Alalinker and RFP was ligated with BsaI(NcoI)/BamHI-digested MPB2C PCRfragments (primer combinations: FK131 (5�-GCCGGTCTCCCATGGCGCT-TGCATTTGAC-3�) or FK 114 (5�-GCCGGTCTCCCATGACCTATAC-CAAA-3�) with FK110) into NcoI/XbaI-digested pZY226 resulting in ex-pression plasmids 35S::MPB2C-RFP or 35S::�1-66MPB2C-RFP. Construct35S::�1-214MBP2C-GFP was generated by PCR amplification of MPB2Cwith primers FK132 (5�-AGGTCATGACTTCCAGCAAG-3�) and FK110 (5�-GATCTCATGAGGATCCCCTGTTCTCAGAAC-3�) and subsequent cloningof the resulting fragment into vector pZY260.

Transient Expression and Confocal Microscopy

DNA was coated onto 1-�m gold particles, and N. tabacum or N. benthami-ana source leaves (�20 cm) were bombarded using a Helios Gene Gunfollowing the manufacturer’s protocol (Bio-Rad Laboratories, Hercules,CA). Each bombardment experiment was repeated at least five times withthe total number of analyzed cells ranging from 40 to 100. Images wereobtained with a TCS-SP confocal microscope (Leica Microsystems, Heidel-berg). GFP and RFP were excited with a ArKr laser at 476 and 568 nm,




respectively. GFP was detected at 500 to 520 nm, and RFP was detected at600 to 620 nm. Micrographs were assembled from serial optical sections andwere prepared for presentation by Photoshop 5.1 software (Adobe Systems,San Jose, CA).

Accession Numbers

The cDNA sequence of MPB2C has been deposited at GenBank underaccession number AF326729.

Distribution of Materials

Upon request, all novel materials described in this publication will bemade available in a timely manner for noncommercial research purposeswith the exception of the anti-MPB2C antibodies, which are available onlyin limited amounts.

ACKNOWLEDGMENTS

We thank A. Aronheim for providing plasmids and the yeast strainrequired for the SRS screen. We are also indebted to Dave Jackson forproviding vectors pZY260 and pZY226, to Peter Palukaitis for providingvector pET-CMV3a, to Vitaly Citovsky for providing vector 35S::RFP, toBiao Ding for providing vector pRTL-CMV3a-GFP, to Richard Cyr forproviding a MAP4-GFP expression construct, and to Karl Oparka for pro-viding MPR3-GFP expression construct. E.W. is thankful to Andrea Barta forher invaluable support throughout the whole project.

Received March 5, 2003; returned for revision March 21, 2003; acceptedApril 24, 2003.

LITERATURE CITED

Aronheim A, Zandi E, Hennemann H, Elledge SJ, Karin M (1997) Isolationof an AP-1 repressor by a novel method for detecting protein-proteininteractions. Mol Cell Biol 17: 3094–3102

Ausubel FM, Brent R, Kingston RE, Moore DD, Smith JA, Seidman JG,Struhl K (1987) Current Protocols in Molecular Biology. GreenePublishing-Wiley Interscience, New York

Boyko V, Ferralli J, Ashby J, Schellenbaum P, Heinlein M (2000a) Functionof microtubules in intercellular transport of plant virus RNA. Nat CellBiol 2: 826–832

Boyko V, Ferralli J, Heinlein M (2000b) Cell-to-cell movement of TMVRNA is temperature-dependent and corresponds to the association ofmovement protein with microtubuli. Plant J 22: 315–325

Brill LM, Nunn SR, Kahn TW, Yeager M, Beachy RN (2000) Recombinanttobacco mosaic virus movement protein is an RNA-binding, �-helicalmembrane protein. Proc Natl Acad Sci USA 97: 7112–7117

Cai G, Romagnoli S, Moscatelli A, Ovidi E, Gambellini G, Tiezzi A, CrestiM (2000) Identification and characterization of a novel microtubule-based motor associated with membranous organelles in tobacco pollentubes. Plant Cell 12: 1719–1736

Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Greenfluorescent protein as a marker for gene expression. Science 263: 802–805

Chen M-H, Sheng J, Hind G, Handa AK, Citovsky V (2000) Interactionbetween the tobacco mosaic virus movement protein and host cell pectinmethylesterases is required for viral cell-to-cell movement. EMBO J 19:913–920

Citovsky V, Knorr D, Schuster G, Zambryski P (1990) The P30 movementprotein of tobacco mosaic virus is a single strand nucleic acid bindingprotein. Cell 60: 637–647

Citovsky V, McLean BG, Zupan J, Zambryski P (1993) Phosphorylation oftobacco mosaic virus cell-to-cell movement protein by adevelopmentally-regulated plant cell wall-associated protein kinase.Genes Dev 7: 904–910

Crawford KM, Zambryski PC (2000) Subcellular localization determines theavailability of non-targeted proteins to plasmodesmatal transport. CurrBiol 10: 1032–1040

Crawford KM, Zambryski PC (2001) Non-targeted and targeted proteinmovement through plasmodesmata in leaves in different developmentaland physiological states. Plant Physiol 125: 1802–1812

Desvoyes B, Faure-Rabasse S, Chen M-H, Park J-W, Scholthof HB (2002)A novel plant homeodomain protein interacts in a functionally relevantmanner with a virus movement protein. Plant Physiol 129: 1521–1532

Dorokhov YL, Makinen K, Frolova OY, Merits A, Saarinen J, KalkkinenN, Atabekov JG, Saarma M (1999) A novel function for a ubiquitousplant enzyme pectin methylesterase: the host-cell receptor for the tobaccomosaic virus movement protein. FEBS Lett 461: 223–228

Gibbs MJ, Armstrong J, Weiller GF, Gibbs AJ (1997) Virus evolution: thepast, a window on the future? In M Tepfer, E Balazs, eds, Virus-ResistantTransgenic Plants: Potential Ecological Impact. Springer-Verlag, Berlin,pp 1–17

Gillespie T, Boevink P, Haupt S, Roberts AG, Toth R, Valentine T,Chapman S, Oparka KJ (2002) Functional analysis of a DNA-shuffledmovement protein reveals that microtubules are dispensable for thecell-to-cell movement of tobacco mosaic virus. Plant Cell 14: 1207–1222

Haseloff J, Siemering KR, Prasher DC, Hodge SSO (1997) Removal of acryptic intron and subcellular localization of green fluorescent proteinare required to mark transgenic Arabidopsis plants brightly. Proc NatlAcad Sci USA 94: 2122–2127

Haywood V, Kragler F, Lucas W (2002) Plasmodesmata: pathways forprotein and ribonucleoprotein signaling. Plant Cell Suppl 14: S303–S325

Heinlein M (2002) The spread of tobacco mosaic virus infection: insightsinto the cellular mechanism of RNA transport. Cell Mol Life Sci 59: 58–82

Heinlein M, Epel BL, Padgett SH, Beachy RN (1995) Interaction of tobamo-virus movement proteins with the plant cytoskeleton. Science 270:1983–1985

Heinlein M, Padgett HS, Gens JS, Pickard BG, Caspar SJ, Epel BL, BeachyRN (1998) Changing patterns of localization of the tobacco mosaic virusmovement protein and replicase to the endoplasmic reticulum and mi-crotubules during infection. Plant Cell 10: 1107–1120

Huang Z, Andrianov VM, Han Y, Howell SH (2001) Identification ofArabidopsis proteins that interact with the cauliflower mosaic virus(CaMV) movement protein. Plant Mol Biol 47: 663–675

Itaya A, Hickman H, Bao Y, Nelson R, Ding B (1997) Cell-to-cell traffickingof cucumber mosaic virus movement protein:green fluorescent proteinfusion produced by biolistic bombardment in tobacco. Plant J 12:1223–1230

Kahn T, Lapidot M, Heinlein M, Reichel C, Cooper B, Gafny R, BeachyRN (1998) Domains of the TMV movement protein involved in subcel-lular localization. Plant J 15: 15–25

Koop H-U, Steinmuller K, Wagner H, Rossler C, Eibl C, Sacher L (1996)Integration of foreign sequences into the tobacco plastome via polyeth-ylene glycol-mediated protoplast transformation. Planta 199: 193–201

Kotlizky G, Katz A, van der Laak J, Boyko V, Lapidot M, Beachy RN,Heinlein M, Epel BL (2001) A dysfunctional movement protein of to-bacco mosaic virus interferes with targeting of wild-type movementprotein to microtubules. Mol Plant-Microbe Interact 14: 895–904

Kragler F, Monzer J, Shash K, Xonocostle-Cazares B, Lucas WJ (1998)Cell-to-cell transport of proteins: requirement for unfolding and charac-terization of binding to a putative plasmodesmal receptor. Plant J 15:367–381

Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathiccharacter of a protein. J Mol Biol 157: 105–142

Lazarowitz SG, Beachy RN (1999) Viral movement proteins as probes forintracellular and intercellular trafficking in plants. Plant Cell 11: 535–548

Lee J-Y, Yoo B-C, Rojas MR, Gomez-Ospina N, Staehelin LA, Lucas WJ(2003) Selective trafficking of non-cell-autonomous proteins mediated byNtNCAPP1. Science 299: 392–396

Li Q, Palukaitis P (1996) Comparison of the nucleic acid- and NTP-bindingproperties of the movement protein of cucumber mosaic cucumovirusand tobacco mosaic tobamovirus. Virology 216: 71–79

Lin B, Heaton LA (2001) An Arabidopsis thaliana protein interacts with amovement protein of turnip crinkle virus in yeast cells and in vitro. J GenVirol 82: 1245–1251

Lucas WJ (1995) Plasmodesmata: intercellular channels for macromoleculartransport in plants. Curr Opin Cell Biol 7: 673–680

Lupas A, van Dyke M, Stock J (1991) Predicting coiled coils from proteinsequences. Science 252: 1162–1164

Kragler et al.



Marc J, Granger CL, Brincat J, Fisher DD, Kao T, McCubbin AG, Cyr RJ(1998) A GFP-MAP4 reporter gene for visualizing cortical microtubulerearrangements in living epidermal cells. Plant Cell 10: 1927–1939

Mas P, Beachy R (1999) Replication of tobacco mosaic virus on endoplasmicreticulum and role of the cytoskeleton and virus movement protein inintracellular distribution of viral RNA. J Cell Biol 147: 945–958

Mas P, Beachy R (2000) Role of microtubules in the intracellular distributionof tobacco mosaic virus movement protein. Proc Natl Acad Sci USA 97:12345–12349

Matsushita Y, Miyakawa M, Nisiguchi M, Nyunoya H (2002) Cloning of atobacco cDNA coding for a putative transcriptional coactivator MBF1that interacts with the tomato mosaic virus movement protein. J Exp Bot53: 1531–1532

Matz MV, Fradkov AF, Labas YA, Savitsky AP, Zaraisky AG, MarkelovML, Lukyanov SA (1999) Fluorescent proteins from nonbioluminescentAnthozoa species. Nat Biotechnol 17: 969–973

McLean BG, Zambryski P (2000) Interactions between viral movementproteins and the cytoskeleton. In CJ Staiger, F Baluska, D Volkmann, PWBarlow, eds, Actin: A Dynamic Framework for Multiple Plant Cell Func-tions. Kluwer Academic Publishers, Dordrecht, The Netherlands

McLean BG, Zupan J, Zambryski PC (1995) TMV P30 movement proteinassociates with the cytoskeleton in tobacco cells. Plant Cell 7: 2101–2114

Padgett HS, Epel BL, Kahn TW, Heinlein M, Watanabe Y, Beachy RN(1996) Distribution of tobamovirus movement protein in infected cellsand implications for cell-to-cell spread of infection. Plant J 10: 1079–1088

Page RDM (1996) TREEVIEW: an application to display phylogenetic treeson personal computers. Comput Appl Biosci 12: 357–358

Pearson WR, Lipman DJ (1988) Improved tools for biological sequencecomparison. Proc Natl Acad Sci USA 85: 2444–2448

Perez F, Diamantopoulos GS, Stalder R, Kreis TE (1999) CLIP-170 high-lights growing microtubule ends in vivo. Cell 96: 517–527

Pierre P, Pepperkok R, Kreis TE (1994) Molecular characterization of twofunctional domains of CLIP-170 in vivo. J Cell Sci 107: 1909–1920

Reichel C, Beachy RN (1998) Tobacco mosaic virus infection induces severemorphological changes of the endoplasmic reticulum. Proc Natl Acad SciUSA 95: 11169–11174

Reichel C, Beachy RN (2000) Degradation of tobacco mosaic virus move-ment protein by the 26S proteasome. J Virol 74: 3330–3337

Sablowski RWM, Meyerowitz EM (1998) A homolog of NO APICALMERISTEM is an immediate target of the floral homeotic genesAPETALA3/PISTILLATA. Cell 92: 93–103

Soellick T-R, Uhrig JF, Bucher GL, Kellmann J-W, Schreier PH (2000) Themovement protein NSm of tomato spotted wilt tospovirus (TSWV): RNAbinding, interaction with the TSWV N protein, and identification ofinteracting plant proteins. Proc Natl Acad Sci USA 97: 2373–2378

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving thesensitivity of progressive multiple sequence alignment through sequenceweighting, position-specific gap penalties and weight matrix choice.Nucleic Acids Res 22: 4673–4680

Toth RL, Pogue GP, Chapman S (2002) Improvement of a plant virus basedvector through DNA shuffling. Plant J 30: 593–600

Tzfira T, Rhee Y, Chen M-H, Citovsky V (2000) Nucleic acid transport inplant-microbe interactions: the molecules that walk through the walls.Annu Rev Microbiol 54: 187–219

Waigmann E, Chen M-H, Bachmaier R, Goshroy S, Citovsky V (2000)Regulation of plasmodesmal transport by phosphorylation of tobaccomosaic virus cell-to-cell movement protein. EMBO J 19: 4875–4884

Waigmann E, Lucas W, Citovsky V, Zambryski P (1994) Direct functionalassay for tobacco mosaic virus cell-to-cell movement protein and identi-fication of a domain involved in increasing plasmodesmal permeability.Proc Natl Acad Sci USA 91: 1433–1437

Wolf S, Deom CM, Beachy RN, Lucas WJ (1989) Movement protein oftobacco mosaic virus modifies plasmodesmatal size exclusion limit. Sci-ence 246: 377–379

Xia G, Ramachandran S, Hong Y, Chan y, Simanis V, Chua N-H (1996)Identification of plant cytoskeletal, cell cycle-related and polarity-relatedproteins using Schizosaccharomyces pombe. Plant J 10: 761–769

Yamanaka T, Ohta T, Takahashi M, Meshi T, Schmidt R, Dean C, Naito S,Ishikawa M (2000) TOM1, an Arabidopsis gene required for efficientmultiplication of a tobamovirus, encodes a putative transmembrane pro-tein. Proc Natl Acad Sci USA 97: 10107–10112




mpb2c, a microtubule-associated plant protein binds … · mpb2c via the yeast sos recruitment...

Documents