The N-Terminal Portion of the 216-kDa Polyprotein of ... · VIROLOGY 232, 359–368 (1997) ARTICLE NO. VY978569 The N-Terminal Portion of the 216-kDa Polyprotein of Commelina Yellow

VIROLOGY 232, 359–368 (1997)ARTICLE NO. VY978569

The N-Terminal Portion of the 216-kDa Polyprotein of Commelina Yellow Mottle BadnavirusIs Required for Virus Movement but Not for Replication

Iris Tzafrir,* Ligia Ayala-Navarrete,†,1 Benham E. L. Lockhart,† and Neil E. Olszewski*,2

*Department of Plant Biology and †Department of Plant Pathology and *Plant Molecular Genetics Institute,University of Minnesota, St. Paul, Minnesota 55108

Received February 20, 1997; returned to author for revision March 26, 1997; accepted April 1, 1997

Commelina yellow mottle virus (CoYMV) is the type member of the badnaviruses, a genus of plant pararetroviruses. TheN-terminus of the polyprotein encoded by ORF III has limited similarity to known cell-to-cell movement proteins. To testthe hypothesis that the N-terminus is required for viral movement, the phenotypes caused by mutations constructed in thisregion were determined. Similar to mutants affected in the reverse transcriptase, mutants affected in the putative movementprotein were unable to cause a systemic infection. However, when the abilities of the mutated viral genomes to direct virionassembly and replication were tested using an in vitro stem-culture system, the mutants affected in the putative movementprotein were found to assemble virions, whereas the reverse transcriptase mutants were unable to do so. Moreover, theputative movement protein mutants were shown to be replication competent by detection and mapping of one of the genomicdiscontinuities that are the hallmark of replication by reverse transcription. Thus the N-terminal region of ORF III is requiredfor the systemic movement but not for the replication of CoYMV. q 1997 Academic Press

INTRODUCTION infection. This active movement process is potentiatedby virus-encoded movement proteins (reviewed by Hull,

Commelina yellow mottle virus (CoYMV) is the type1989; Atabekov and Taliansky, 1990; Maule, 1991; Deom

member of the badnavirus group of plant pararetrovi-et al., 1992). Plant viruses employ their movement pro-

ruses (Lockhart, 1990). Although the genomes of a num-tein(s) for spreading from infected to healthy cells by

ber of badnaviruses have been sequenced and charac-at least two different mechanisms that are not mutually

terized (Medberry et al., 1990; Dasgupta et al., 1991; Hayexclusive (reviewed by Deom et al., 1992; McLean et al.,

et al., 1991; Qu et al., 1991; Bouhida et al., 1993; Hagen et1993). Movement proteins of a number of plant viruses

al., 1993), the genomic region(s) required for badnavirushave been shown to bind nucleic acids (Citovsky et al.,

movement in their respective hosts has not been defini-1990, 1991, 1992; Giesman-Cookmeyer and Lommel,

tively identified.1993; Thomas and Maule, 1995) and to potentiate the

Open reading frame (ORF) III of CoYMV is believed tomovement of a movement protein–nucleic acid complex

encode a 216-kDa polyprotein that is processed by aacross the plasmodesmata to neighboring cells (Fuji-

virally encoded aspartic protease to yield a putative cell-wara et al., 1993; Waigmann et al., 1994; Noueiry et al.,to-cell movement protein, the coat protein, an aspartic1994; Ding et al., 1995). In addition, a number of move-protease, reverse transcriptase, and RNase H (Medberryment proteins have been demonstrated to be structuralet al., 1990). Comparisons with bona fide and putativecomponents of infection-specific virion-containing tubulemovement proteins from diverse groups of plant virusesstructures that extend across the plant cell wall and arehave identified limited homologies between these pro-believed to serve as conduits for cell-to-cell transport ofteins and the N-terminal region of the protein encodedvirus particles (van Lent et al., 1991; Kasteel et al., 1993;by ORF III, suggesting that this region is involved in theWellink et al., 1993; Perbal et al., 1993; Wieczorek et al.,cell-to-cell movement of CoYMV (Bouhida et al., 1993;1993; Storms et al., 1995; Ritzenthaler et al., 1995; C.-P.Mushegian and Koonin, 1993; Melcher, 1993).Cheng et al., submitted). The movement protein of cauli-Following their initial entry into the host, plant virusesflower mosaic virus (CaMV) has been shown to bothmust express gene product(s) to establish a systemicpossess nucleic-acid binding capabilities and to be astructural component of tubular structures (Citovsky et

1 Present address: INIAP, Departamento de Proteccion Vegetal, al., 1991; Thomas and Maule, 1995; Perbal et al., 1993).Apartado 340, Quito, Ecuador.

The model for pararetrovirus replication, based upon2 To whom correspondence and reprint requests should be ad-studies with caulimoviruses, is that reverse transcriptiondressed at 220 Biological Sciences Center, 1445 Gortner Ave., St. Paul,

MN 55108. Fax: (612) 625-1738. E-mail: [email protected]. occurs within virions or virion-like particles, suggesting

3590042-6822/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID VY 8569 / 6a36$$$341 05-09-97 21:57:24 viras AP: Virology

360 TZAFRIR ET AL.

that encapsidation occurs concomitantly with replication from New England Biolabs (Beverly, MA) or from Pro-mega (Madison, WI). All nucleotide and amino acid num-(Thomas et al., 1985; Marsh and Guilfoyle 1987; Fuetterer

and Hohn, 1987; Mesnard and Carriere, 1995). Therefore, bering used throughout this paper are according toMedberry et al. (1990). Oligonucleotide primers were syn-this linkage of replication to encapsidation is a useful

marker for analyzing putative movement protein-deficient thesized at the University of Minnesota or purchasedfrom IDT (Coralville, IA).mutants for their replication competency. Finding puta-

tive movement protein mutants which are encapsidation-Construction of mutant CoYMV genomescompetent, however, still leaves room for speculation

whether the observed virions contain viral DNA and Three deletion mutations, MP1, MP2, and MP3 (Fig. 1),whether this DNA is a product of replication by reverse which did not shift the reading frame, were constructed intranscription. the putative movement protein coding region of ORF III

For synthesis of the CoYMV genomic strands, an initia- by site-directed mutagenesis according to Kunkel (1985).tor methionine tRNA primes negative-strand synthesis, The oligonucleotides used for mutagenesis are listed inand a polypurine-rich oligonucleotide derived from the Table 1. To facilitate the identification of the mutants, anRNA template primes positive-strand synthesis (Med- NheI site was introduced in place of the deleted se-berry et al., 1990). At the genomic regions corresponding quences. All of the mutations were constructed in anto the priming sites, the encapsidated CoYMV genome is EcoRV–PstI (nucleotides 7124–7489/0–3261) fragmentinterrupted by discontinuities. The negative- and positive- of pCoYMV89 (Medberry et al., 1990) cloned into pBlue-strand discontinuities have been mapped to positions script II (SK0) (Stratagene, La Jolla, CA). Following muta-7488–2 and 4700–4701, respectively (Medberry et al., genesis, a SnaBI– SphI fragment (nucleotides 269–2752)1990). In the case of CaMV, the extreme 5* end of each was excised and cloned in place of the correspondingdiscontinuity has been shown to be composed of ribonu- wild-type region of the infectious clone pCoinf (Medberrycleotides that are suggested to be remnants of the prim- et al., 1990). Mutagenized pCoinf was linearized at aers for replication (Guilley et al., 1983, and references unique XhoI site and ligated to the unique SalI site oftherein). The discontinuities are not simply gaps in one pOCA28 (Fig. 1; Medberry et al., 1990) to produce thestrand; instead, they contain overlaps between the 3* infectious construct used for agroinoculation. All clonesand the 5* ends of the discontinuous strand. Hohn et used for agroinoculation were propagated in the A281al. (1985) have suggested that the short overlaps in the (Montoya et al., 1977) strain of Agrobacterium tumefa-discontinuities arise by the peeling off of limited ciens.stretches of the 3* of each strand. Thus, the site-specific The two deletion mutations, RT1 and RT2 (Fig. 1), werediscontinuities are the hallmark of pararetrovirus replica- constructed in the reverse transcriptase-coding regiontion by reverse transcription. of ORF III using a PCR-based method for site-directed

While a viral movement protein is required for virus mutagenesis by overlap extension (Ho et al., 1989). Thespread in the host, this protein has no predicted direct oligonucleotide primers used in this procedure are listedrole in virus replication. Therefore mutations affecting a in Table 1. To facilitate the identification of the mutants,movement protein are predicted to block movement but a NarI site was introduced in place of the deleted se-not replication. We have tested this hypothesis in whole quences. The nature of each mutation was confirmed byplants and organs and confirmed this prediction. We de- dideoxy-termination DNA sequencing (fmol DNA se-veloped tools to distinguish between mutants with de- quencing system; Promega).fects in movement from mutants with defects in replica-tion. These tools include testing for production of virions, Plant growth conditionsdetermining whether DNA is encapsidated in these viri-

Commelina diffusa plants were grown in a growthons, and demonstrating that encapsidated DNA containschamber with a 16 hr light:8 hr dark cycle, at 30 and 287,discontinuities resulting from replication by reverse tran-respectively.scription. Our results demonstrate that the movement-

deficiency phenotype of mutants in the N-terminus ofAgroinoculation of plants

ORF III is due to a defect in their movement rather thanthe replication function. A. tumefaciens-mediated inoculation (agroinoculation)

(Grimsley et al., 1986) of C. diffusa was performed asdescribed previously (Medberry et al., 1990). Briefly, aMATERIALS AND METHODSsingle colony of A. tumefaciens containing the recombi-nant CoYMV construct of interest was used to inoculateStandard molecular biological techniques were used

for plasmid construction, selection, and propagation a 5-ml LB culture containing 300 mg/ml streptomycin, 100mg/ml carbenicillin, and 75 mg/ml spectinomycin (Sigma(Sambrook et al., 1989; Ausubel et al., 1987), unless noted

otherwise. All restriction enzymes were purchased either Chemical, St. Louis, MO). Following growth for 48 hr,


361IDENTIFYING THE CoYMV MOVEMENT PROTEIN

FIG. 1. Map of pCoinf indicating the locations of the deletion mutations constructed in CoYMV ORF III. The top illustrates the viral portion of theinfectious construct pCoinf (Medberry et al., 1990), which includes 1.3 copies of the viral genome. The open box indicates viral DNA and the solidregions indicate vector DNA: X, XhoI; C, ClaI; Sn, SnaBI; Sp, SphI. The location of the full-length CoYMV transcript is illustrated by the arrow belowthis map. The solid bars below the transcript indicate the open reading frames and the lowest cross-hatched box is an enlargement of ORF III.The deletion mutations MP1, MP2, and MP3 were constructed in the putative movement protein-coding region of ORF III and result in the deletionof amino acids 165–175, 140– 156, and 300–321, respectively. The deletion mutations RT1 and RT2 were constructed in the reverse transcriptase-coding region of ORF III and result in the deletion of amino acids 1506–1512 and 1566–1573, respectively.

cells were collected by centrifugation and resuspended Young stems of C. diffusa were harvested, leaves wereremoved, and stems were cut in 6- to 10-cm pieces,in 1 ml of 10 mM MgSO4 . A sterile toothpick dipped into

the concentrated culture was used to wound the stem sterilized for 10 min in 10% bleach and 0.02% Tween20 (Sigma Chemical), and then rinsed three times withof C. diffusa plants at the second internode below the

apex. sterile distilled water. In a sterile environment, thestems were cut in 2- to 3-cm pieces and split length-wise. The half-stem segments were placed in a sus-Assays for systemic infection of plantspension of the A. tumefaciens containing the viral con-

Following agroinoculation, plants were observed for struct of interest in sterile water. Agroinoculated stems30 days. On plants agroinoculated with wild-type CoYMV were laid cut side up on a 0.8% agar medium con-constructs, systemic symptoms were routinely observed taining 0.51 MS salts (Murashige and Skoog, 1962)as yellow mottling, starting on the third leaf above the site and 0.51 B5 vitamins (Sigma Chemical). Stem seg-of inoculation. In ú99% of systemically infected plants, ments were incubated in a growth chamber under 16symptoms appeared by 14 days postinoculation (Ayala- hr light:8 hr dark regime at a constant temperature ofNavarrete, 1992). As a further test for systemic infection, 227. Following a 10-day incubation, virus extractiontotal DNA was extracted from uninoculated leaf material was carried out according to Bouhida et al. (1993) with(0.1 g) according to Reiter et al. (1992), and viral se-

the following modifications. Stem segments (Ç20 g)quences were detected by PCR amplification. Each PCR

were washed three times with water and then groundcycle included denaturation for 30 sec at 957, annealing

in a coffee grinder (Braun, Germany) with dry ice pel-for 30 sec at 507, and polymerization for 2 min at 727,

lets. The resulting powder was suspended in 40 ml offor a total of 25 cycles, using a PTC-100 programmable

an extraction buffer (200 mM NaH2PO4 adjusted to pHcontroller (MJ Research, Inc., Watertown, MA). The oligo-

6.0 with 200 mM K2HPO4 , 10 mM EDTA, 0.5% b-mer-nucleotide primers ORF I Forw and ORF II Rev (Table 1)

captoethanol, and 1% Na2SO3 ). The suspension waswere used for viral DNA detection. PCR products were

filtered through cheese cloth, clarified by extractionsubjected to electrophoresis on a 0.8% agarose gel andwith chloroform, and subjected to centrifugation in avisualized following staining with ethidium bromide.Beckman JA-20 at 10,000 RPM for 15 min. The superna-tant was then subjected to centrifugation in a BeckmanAssay of CoYMV replication in cultured C. diffusa50.2 Ti rotor at 35,000 rpm for 70 min. The resultingstem segmentspellet was resuspended in 75 ml of resuspension buffer(10 mM K2HPO4 adjusted to pH 7.4 with 10 mMAn in vitro stem culture system was developed andNaH2PO4 , 0.8% NaCl), shaken with an equal volume ofused to determine whether mutants that did not cause

a systemic infection were able to produce virions. chloroform, and subjected to centrifugation in a Beck-


362 TZAFRIR ET AL.

TABLE 1

Oligonucleotides Used in This Study

Mutation designation Primer name Annealing position a 5* r 3* Deleted nucleotides Deleted amino acids

MP1 Primer Del 1 b 2045–1983 2001–2027 165–175MP2 Primer Del 2 b 1989–1905 1926–1970 140–156MP3 Primer Del 3 b 2483–2379 2402–2459 300–321RT1 RT1 FORW c 6010–6057 6018–6038 1506–1512

RT1 REV c 6045–5998Primer 4630 FORW c 4630–4648Primer 6381 Rev c 6381–6365

RT2 RT2 FORW c 6193–6241 6201–6221 1566–1573RT2 REV c 6238–6182Primer 4630 FORW c 4630–4648Primer 6381 REV c 6381–6365

ORF I Forw d 489–510ORF II Rev d 1529–1505

a Positions according to Medberry et al. (1990).b Primers used for deletion mutagenesis according to Kunkel (1985).c Primers used for deletion mutagenesis by overlap extension PCR (Ho et al., 1989).d Primers used for virus detection by PCR amplification.

man JA-18.1 rotor at 12,000 rpm for 5 min. The clarified Taq buffer (Perkin–Elmer, Norwalk, CT), 1.5 mM MgCl2 ,with 1.25 units of AmpliTaq. Encapsidated sequencesaqueous phase constituted the partially purified virus

preparation. were released from virions and denatured by incubationat 947 for 4 min. Amplification was carried out in three

Immunosorbent electron microscopy stages: four cycles of 947/45 sec, 377/30 sec, 727/2 min,followed by 30 cycles of 947/45 sec, 507/30 sec, 727/2Immunosorbent electron microscopy (ISEM) analysesmin, and a final extension at 727/7 min. A 20-ml aliquotof partially purified virus preparations were carried outof each PCR was subjected to electrophoresis on a 0.8%as described by Bouhida et al. (1993), using antibodiesagarose gel and visualized following staining with ethid-prepared against purified CoYMV virions (Lockhart, un-ium bromide.published results).

PCR-based detection and mapping of genomicImmunocapture PCRdiscontinuities

Prior to determining whether virus particles detectedby ISEM contain DNA, it was necessary to remove resid- A modification of a ligation-mediated PCR procedure

(Mueller and Wold, 1989) was developed for detectionual A. tumefaciens and A. tumefaciens DNA. To this end,immunocapture PCR (Nolasco et al., 1993; Ducasse et al., and mapping of the site-specific genomic discontinuities

that are the hallmark of pararetrovirus replication by re-1995) was employed. Affinity-purified antibody directedagainst the CoYMV coat protein (Cheng et al., 1996) was verse transcription. Residual RNA and DNA from plants

and A. tumefaciens were removed from the partially puri-diluted 1:200 in carbonate coating buffer (60 mM Na2CO3 ,pH 9.5). Microcentrifuge tubes (0.5 ml; Fisher Scientific, fied virus preparations by digestion with RNase A (100

mg/ml) and DNase I (25 mg/ml) for 1 hr at 377. Virion DNAPittsburgh, PA) were coated with 25 ml of the dilutedantibody by incubation for 3 hr at room temperature. was deproteinized by digestion with protease K (100 mg/

ml), in 10 mM Tris–HCl, 50 mM EDTA, and 0.5% (w/v)Coated tubes were rinsed five times with phosphate-buffered saline–0.05% Tween 20 (PBST). Partially puri- sodium dodecyl sulfate for 1 hr at 557, followed by a

phenol:chloroform extraction and precipitation with iso-fied virus preparations from agroinoculated stem seg-ments were diluted 1:5 in 21 PBST–0.4% BSA to a total propanol. Following centrifugation, washing with 70%

ethanol, and drying, the DNA pellet was resuspended involume of 25 ml and added to the pretreated tubes. Fol-lowing incubation overnight at 47, tubes were rinsed five water and treated with piperidine (Sigma Chemical) as

described by Ausubel et al. (1992) to remove RNA tractstimes with PBST. For a 25-ml PCR, the following cocktailwas added to the tube containing the captured virions: presumed to be present at the discontinuities.

The piperidine-treated DNA was subjected to ligation-20 pmol each of ORF I Forw and ORF II Rev primers, 0.2mM each dNTP (Pharmacia, Piscataway, NJ), 11 Ampli- mediated PCR according to Hornstra and Yang (1993)



with some modifications. Primer 1 (annealing at positions in cell-to-cell movement. For construction of MP3, astretch of amino acids highly conserved among the bad-4848–4833 in the CoYMV genome) was extended to gen-

erate a blunt-ended DNA molecule terminating at the naviruses sugarcane bacilliform virus (ScBV), rice tungrobacilliform virus (RTBV), and CoYMV was deleted.site of the polypurine-rich discontinuity on the CoYMV

positive genomic strand (Medberry et al., 1990). Next, a C. diffusa plants were agroinoculated with the putativemovement protein mutants. In each experiment two tostaggered double-stranded DNA linker was ligated to

the blunt-ended DNA. This DNA linker was composed of five plants were inoculated with a single construct andmonitored for signs of a systemic infection for up to 30linker primer 1 (25 bp) and linker primer 2 (11 bp), both

designed according to Mueller and Wold (1989). Follow- days. At least three independent trials were performedwith each construct. In no instance did the mutants ap-ing ligation, an amplification reaction was carried out

with linker primer 1 and primer 2 (annealing at positions pear to cause a systemic infection, whereas wild-typeCoYMV constructs caused systemic symptoms in 100%4814–4790 in the CoYMV genome). The PCR products

were purified with QIAquick Spin kit (Qiagen, Chatsworth, of the test plants within 10–14 days postinoculation.Moreover, no viral DNA was detected by PCR analysisCA) and used as templates for an extension PCR protocol

using an end-labeled primer 3 (annealing at positions of the uninoculated leaves of plants inoculated with themutant constructs. The expected product was obtained4806–4779 in the CoYMV genome). Extension products

were precipitated with 1 volume isopropanol and 1/10 with DNA prepared from the uninoculated leaves ofplants infected with wild-type CoYMV (data not shown).volume 5 M NaCl at0207 overnight, collected by centrifu-

gation, washed twice with 80% ethanol, dried, and resus-pended in a 1:2 mix of loading dye and water. The final The putative movement protein mutants assembleextension products were denatured and resolved by virionselectrophoresis on a 6% sequencing gel. The size ofthe extension product was determined by comparing its Since failure to detect a systemic infection could be

due to either a defect in replication or a defect in virusmigration to that of the products of DNA sequencingreactions performed using end-labeled primer 3 and movement, we developed a stem-culture system in which

large number of cells could be inoculated, and used itpCoinf (Medberry et al., 1990). The gel was dried andexposed to X-ray film (Kodak, Rochester, NY) overnight to assess the replication competence of the putative

movement protein mutants. The principle of this systemat room temperature.is the same as that used by Thomas et al. (1993) to detectreplication of movement mutants of CaMV. Basically, ifRESULTSenough cells can be infected, it is possible to detect

Establishment of a systemic infection requires thereplication even in the absence of movement. Given the

putative movement proteinhigh efficiency of successful agroinoculation of C. diffusawith CoYMV, we reasoned that a quantity of cells suffi-To test the hypothesis that the N-terminus of ORF III

is required for virus movement, the effects of deletions cient to allow detection of replication products of move-ment-deficient mutants would be infected by inoculatingin this region were determined. Deletion mutations that

maintain the ORF III reading frame were created. This split stem segments.Two replication mutants, RT1 and RT2, which servedapproach both avoids disrupting the expression of the

remainder of the polyprotein and avoids the potential as negative controls for these experiments were alsoconstructed. The RT1 and RT2 mutants harbor deletionreversion that may occur with point mutations (Hohn et

al., 1985). mutations removing regions of the CoYMV reverse tran-scriptase that are conserved among retroviruses, retro-The regions deleted exhibit the highest sequence simi-

larity in comparisons between the putative cell-to-cell transposons, and pararetroviruses (McClure, 1993).These regions are implicated in template seating andmovement protein of CoYMV and putative and bona fide

cell-to-cell movement proteins of dsDNA and ssRNA polymerase activity (Prasad, 1993). The two aspartateresidues in the region deleted in RT2 are invariant amongplant viruses (Bouhida et al., 1993; Mushegian and Koo-

nin, 1993; Melcher, 1993). For construction of the MP1 all RNA-dependent DNA and RNA polymerases, and mu-tation analysis suggested that these residues are at theand MP2 mutants, the region deleted included either

aspartic acid (Asp 166) or histidine (His 140), respec- catalytic site (Prasad, 1993, and references therein; Kohl-staedt et al., 1993). When a total of 16 plants were inocu-tively, which were identified as highly conserved resi-

dues in the analyses by Bouhida et al. (1993), Mushegian lated with RT1 and when a total of 16 plants were inocu-lated with RT2, no symptoms or systemic viral DNA wereand Koonin (1993), and Melcher (1993). The amino acids

deleted for MP1 and MP2 may be part of an RNA-binding detected in any of the inoculated plants.Virions were detected by ISEM in partially purifieddomain (Mushegian and Koonin, 1993; Thomas and

Maule, 1995) which is believed to play an important role preparations from stem segments inoculated with the


364 TZAFRIR ET AL.

FIG. 2. Putative movement protein mutants assemble virions. (A–F) ISEM images of MP1, MP2, MP3, wild-type CoYMV, RT1, and RT2, respectively.The scale bar represents 500 nm in (A), (C), (D), and (E), 650 nm in (B), and 850 nm in (F).

MP1, MP2, MP3, or wild-type construct (Fig. 2 and Table ing that reversion of the mutation had not occurred (datanot shown).2). In contrast, no virions were detected in similar prepa-

rations from stem segments inoculated with either theThe putative movement protein mutants encapsidateRT1 or the RT2 replication mutant (Fig. 2 and Table 2).viral DNATo determine if the MP1, MP2, and MP3 mutations had

reverted to wild type, DNA was prepared from each stem- The results obtained with the ISEM analysis are con-culture treatment and subjected to a PCR in which the sistent with the hypothesis that the putative movementregion spanning the mutation was amplified and the protein-deletion mutants are able to replicate. This hy-product was analyzed for the presence of an NheI restric- pothesis also predicts that these virions will contain viraltion site that was introduced during construction of these DNA. As shown in Fig. 3, immunocapture PCR, usingmutants. In all cases, the PCR product was digested anti-CoYMV antiserum and CoYMV-specific primers (seecompletely by the appropriate restriction enzyme, indicat- Materials and Methods), amplified CoYMV DNA that was

presumably encapsidated, in virus preparations fromstem segments inoculated with the MP1, MP2, MP3, orTABLE 2wild-type construct. In contrast, no DNA was detected

Virion Counts Determined by Immunosorbent Electron Microscopy when this method was applied to virus preparations fromstem segments inoculated with either the RT1 or the RT2Mean number of virions per field { standard error*construct. However, a CoYMV PCR product was pro-

Construct Experiment 1 Experiment 2 Experiment 3 duced by all samples if the washing steps were not per-formed. The washing-dependent negative results ob-

WT 13.6 { 0.69 a 56.95 { 2.9 a 32.1 { 3.41 a

tained for the RT1 and RT2 mutants demonstrate thatMP1 2.85 { 0.37 b 7.5 { 0.9 b 2.8 { 0.005 b

residual inoculum DNA and A. tumefaciens were suc-MP2 2.25 { 0.36 b 7.1 { 0.7 b 1.05 { 0.18 c

cessfully eliminated by the washing steps of the immuno-MP3 0.65 { 0.195 c 0.45 { 0.17 c 0.45 { 0.153 d

RT1 0 0 0 capture procedure. These data suggest that the virionsRT2 0 0 0 produced by the MP1, MP2, and MP3 mutants contain

CoYMV DNA.* Mean number of virions per field based upon 20 randomly selected

fields examined at a magnification of 11,500 { standard error. ANOVAThe putative movement protein mutants areindicates that there is a construct 1 experiment interaction, thereforereplication competentaveraging across the three independent experiments is not appropriate.

a,b,c,d Different letters indicate values that are significantly differentThe hypothesis that the putative movement protein de-from each other within a given experimental day based on Student’s t

distribution (df Å 19, a Å 0.05). letion mutants are replication competent predicts not



reasoned that the residual RNA primers covalently boundto the DNA at the discontinuity were interfering with theligation step and must be removed. Therefore, themethod was modified (Fig. 4A) to include a piperidinehydrolysis step that removes the residual RNA primersat the discontinuities.

When wild-type DNA was used, the modified proce-dure detected a discontinuity at the expected location,suggesting that all, or nearly all, of the encapsidatedDNA molecules possess RNA at the 5*-end of the discon-tinuities, as has been observed previously for CaMV byGuilley et al. (1983). Using this procedure, DNAs ex-tracted from partially purified virions isolated from stemsegments infected with the wild-type, MP1, MP2, or MP3construct were all shown to contain the site-specific dis-continuity located near the polypurine-rich region on thepositive genomic strand of CoYMV (Fig. 4B). This discon-tinuity was not detected using DNA isolated from eitherFIG. 3. Immunocapture PCR detection of virion DNA of putative move-

ment protein mutants. Partially purified virus preparations of wild type RT1- or RT2-inoculated material.and mutants were subjected to immunocapture PCR. Reaction aliquotswere subjected to electrophoresis on a 0.8% agarose gel and photo-

DISCUSSIONgraphed following staining with ethidium bromide. The outer lanescontain 1 mg of l DNA digested with HindIII and EcoRI. Lanes 1–6

The N-terminus of ORF III has been hypothesized tocontain the immunocapture PCR products of MP1, MP2, MP3, RT1,RT2, and wild-type CoYMV, respectively. be required for the movement of CoYMV within its plant

host. This hypothesis predicts that mutations affectingthis region will affect the movement, but not the replica-

only that the virions of these mutants will contain DNA, tion of CoYMV. As deletion mutants in the N-terminus ofbut that this encapsidated DNA will contain the site- ORF III were found to be unable to carry out a systemicspecific discontinuities that are the hallmark of replica- infection, we have performed several tests which ruletion by reverse transcription (Pfeiffer and Hohn, 1983; out the possibility that this phenotype is due to a defectGuilley et al., 1983). Only after CoYMV replicates should in replication.it possess the discontinuities which have been mapped Mutants affected in a region that is important for move-previously (Medberry et al., 1990). Neither the binary vec- ment are expected to infect only the cells into whichtor construct used for the agroinoculation, nor its T-DNA, they are introduced. In accordance with this prediction,is expected to possess these discontinuities. Therefore, relative to wild type, fewer virions were recovered fromevidence for the existence of the site-specific discontinu- split stem segments that had been agroinoculated withities in a viral genome obtained from infected plant mate- the MP1, MP2, or MP3 mutants (Table 2). Wild type con-rial would provide direct evidence that the virus is repli- sistently had the highest titer, while MP3 stood out withcation competent and that any viral DNA detected is not the lowest (Table 2).from any other source, such as inoculum DNA. The region that is deleted from MP3 is conserved

By conventional primer extension methods, only the among badnaviruses but not among other movementdiscontinuities of wild-type viral DNA from agroinocu- proteins (Bouhida et al., 1993; Melcher, 1993). Althoughlated stem segments were detected (data not shown). not affecting a generally conserved region, this deletionThe inability to detect the discontinuities which were hy- may affect virus movement because, even though differ-pothesized to be present in DNA from partially purified ent movement proteins share common functions, se-virus preparations of the putative movement protein mu- quence conservation among viral movement proteins istants was possibly due to the low amounts of virions generally low (Melcher, 1990; McLean et al., 1993).(Table 2) and hence viral DNA which could be recovered Alternatively, the region deleted in MP3 may be im-from stem segments. portant for processing of the polyprotein. In Western blot

Since ligation-mediated PCR (Mueller and Wold, 1989) analyses of C. diffusa tissue infected by CoYMV engi-should be more sensitive than primer extension, we em- neered to encode an epitope-tagged movement protein,ployed it to detect the site-specific discontinuity located and in similar analyses using polyclonal antibodiesadjacent to the polypurine-rich region on the positive which recognize the wild-type CoYMV movement protein,strand of the CoYMV genome. Initial efforts to detect the we have detected an infection-specific protein, sized ap-

proximately 35 kDa (I. Tzafrir, B. E. L. Lockhart, and N. E.discontinuity of wild type CoYMV were unsuccessful. We


366 TZAFRIR ET AL.

Olszewski, manuscript in preparation). Based on this pre-dicted size for the movement protein, it is possible thatthe MP3 deletion is at the junction between the move-ment protein and the coat protein and therefore affectsthe proteolytic processing that liberates the maturemovement protein.

If it occurs, reduced polyprotein processing may resultin compromised packaging efficiency which could ac-count for the reduction in the number of particles de-tected from MP3-infected stem segments. This effect isnot likely to occur with MP1 and MP2, because they arenot affected at the putative processing region.

Thomas and Maule (1995) have shown the region be-tween amino acids 120 and 197 in the CaMV movementprotein to constitute the RNA-binding domain. These au-thors further suggest that the RNA-binding activity hasbiological significance since it is this region that is moststrongly conserved between caulimovirus movementproteins and the putative movement protein of badnavi-ruses (Bouhida et al., 1993; Thomas and Maule, 1995).The MP1 and MP2 mutants are affected in this putativeRNA-binding domain. Our data confirm that this domainis important for the ability of CoYMV to spread in itshost. Further studies of these mutants with respect tocharacteristics such as nucleic acid-binding capabilitywill contribute to uncovering the biological basis of themovement deficiency of these mutants.

The mutants harboring deletions in the putative move-ment protein-coding region were predicted to be replica-tion competent. For pararetroviruses, an indirect way totest their replication competency is to analyze their abilityto encapsidate DNA. Therefore, the detection of DNAin the virions of the putative movement protein-deficientmutants (Fig. 3) is consistent with replication compe-tency. In addition, the inability of reverse transcriptase-deficient mutants to form virions (Fig. 2 and Table 2) is inaccordance with earlier work on pararetroviruses whichindicates that reverse transcription is closely associatedwith virus encapsidation.

The virions assembled by the putative movement pro-tein mutants were found to contain virus DNA. Moreover,

FIG. 4. The putative movement protein-deficient mutants encapsidate the results obtained by the PCR-based method for detec-DNA containing the site-specific discontinuities that are the hallmark of

tion and mapping of the genomic discontinuities demon-replication by reverse transcription. (A) An outline of the detection proce-strate that this viral DNA is the result of de novo replica-dure: (1) the structure of the discontinuity present in encapsidated DNA;

R, ribonucleotides covalently attached to the viral DNA; (2) the product of tion by reverse transcription (Fig. 4B). Thus, the putativethe piperidine treatment; (3) an illustration of the annealing site for primer movement protein-deficient mutants replicate after being1 and the extension product it primes; (4) the product of ligating the stag- introduced to plant cells by agroinoculation. In contrast,gered linker to the blunt-ended extension product; (5) the location of the

no such evidence was obtained for the reverse tran-annealing site for the primers used to amplify the extension product; andscriptase-deficient mutants (Fig. 4B). These results con-(6) the annealing location of primer 3. (B) Lanes (G), (A), (T), and (C) contain

the dideoxy chain-termination products of a sequencing reaction set with firmed the prediction that the mutants MP1, MP2, andprimer 3, using a cloned CoYMV template. Lanes (1–6) contain the final MP3 are replication competent and that their inability toextension products generated using end-labeled primer 3: (1) wild-type move is due to a defect in a genomic region required forCoYMV and the mutants (2) MP1, (3) MP2, (4) MP3, (5) RT1, and (6) RT2.

movement but not for replication.As expected, the size of the final extension product is 25 bases longerIn summary, our investigation has genetically charac-than the product produced in primer extension reactions due to the 25-bp

linker sequences that have been ligated to the end of the DNA strand. terized the N-terminus region of ORF III of CoYMV and



plasmodesmata potentiated by the red clover necrotic mosaic virushas shown this region to be required for viral movementmovement protein. Plant Cell 5, 1783–1794.but not for replication. As the N-terminus of the ORF III

Giesman-Cookmeyer, D., and Lommel, S. A. (1993). Alanine scanningprotein of all sequenced badnaviruses (Medberry et al., mutagenesis of a plant virus movement protein identifies three func-1990—CoYMV; Hay et al., 1991, Qu et al., 1991—RTBV; tional domains. Plant Cell 5, 973–982.Bouhida et al., 1993—ScBV; Hagen et al., 1993—CSSV; Grimsley, N., Hohn, B., Hohn, T., and Walden, R. (1986). ‘‘Agroinfection,’’

an alternative route for viral infection of plants by using the Ti plas-Yang, 1995—KTSV; Harper et al., 1996—BSV andmid. Proc. Natl. Acad. Sci. USA 83, 3282–3286.DABV) is similar, this region is likely to be required for

Guilley, H., Richards, K. E., and Jonard, G. (1983). Observations concern-their movement function in their respective hosts.ing the discontinuous DNAs of cauliflower mosaic virus. EMBO J. 2,277–282.

Hagen, L. S., Jacquemond, M., Lepingle, A., Lot, H., and Tepfer, M.ACKNOWLEDGMENTS(1993). Nucleotide sequence and genomic organization of cacaoswollen shoot virus. Virology 196, 619–628.We acknowledge the partial support of I.T. by an NIH Predoctoral

Harper, G., Briddon, R. W., Phillips, S., Brunt, A. A., and Hull, R. (1996).Training Fellowship (5T32-GM07323) and a University of MinnesotaAnalysis of the sequences of two badnaviruses, banana streak virusGraduate School Doctoral Dissertation Fellowship. This work was sup-and Dioscorea alata badnavirus. Abstr. Xth Int. Congr. Virol., Jerusa-ported by grants from the U.S. Department of Agriculture National Re-lem, August 11–16.search Initiative Competitive Grants Program 94-37303-0465 and

Hay, J. M., Jones, M. C., Blakebrough, M. L., Dasgupta, I., Davies, J. W.,USAID Program in Science and Technology Cooperation 5600-G-00-and Hull, R. (1991). An analysis of the sequence of an infectious2017-0 to N.E.O. and B.E.L. We thank Drs. Sammy Sackey and Stephenclone of rice tungro bacilliform virus, a plant pararetrovirus. NucleicSwain for review of the manuscript.Acids Res. 19, 2615–2621.

Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K., and Pease, L. R. (1989).REFERENCES Site-directed mutagenesis by overlap extension using the polymer-

ase chain reaction. Gene 77, 51–59.Atabekov, J. G., and Taliansky, M. E. (1990). Expression of a plant virus- Hohn, T., Hohn, B., and Pfeiffer, P. (1985). Reverse transcription in

coded transport function by different viral genomes. Adv. Virus Res. CaMV. Trends Biochem. Sci. 10, 205–209.38, 201–248. Hornstra, I. K., and Yang, T. P. (1993). In vivo footprinting and genomic

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., sequencing by ligation-mediated PCR. Anal. Biochem. 213, 179–193.Smith, J. A., and Struhl, K. (1987, 1992). ‘‘Current Protocols in Molecu- Hull, R. (1989). The movement of viruses in plants. Annu. Rev. Phytopa-lar Biology.’’ Greene and Wiley–Interscience, New York. thol. 27, 213–240.

Ayala-Navarrete, L. I. (1992). ‘‘Genetic Analysis of Commelina Yellow Kasteel, D., Wellink, J., Verver, J., van Lent, J., Goldbach, R., and vanMottle Virus (CoYMV).’’ University of Minnesota, St. Paul. [M.S. thesis] Kammen, A. (1993). The involvement of cowpea mosaic virus M RNA-

Bouhida, M., Lockhart, B. E. L., and Olszewski, N. E. (1993). An analysis encoded proteins in tubule formation. J. Gen. Virol. 74, 1721–1724.of the complete sequence of a sugarcane bacilliform virus genome

Kohlstaedt, L. A., Wang, J., Rice, P. A., Friedman, J. M., and Steitz, T. A.infectious to banana and rice. J. Gen. Virol. 74, 15–22.

(1993). The structure of HIV-1 reverse transcriptase. In ‘‘Reverse Tran-Cheng, C.-P., Lockhart, B. E. L., and Olszewski, N. E. (1996). The ORF

scriptase’’ (A. M. Skalka and S. P. Goff, Eds.), pp. 223–249. ColdI and II proteins of Commelina yellow mottle virus are virion-associ-

Spring Harbor Laboratory Press, Plainview, NY.ated. Virology 223, 263–271.

Kunkel, T. A. (1985). Rapid and efficient site-specific mutagenesis with-Citovsky, V., Knorr, D., Schuster, G., and Zambryski, P. (1990). The P30

out phenotypic selection. Proc. Natl. Acad. Sci. USA 82, 488–492.movement protein of tobacco mosaic virus is a single-strand nucleic

Lockhart, B. E. L. (1990). Evidence for a double-stranded circular DNAacid binding protein. Cell 60, 637–647.genome in a second group of plant viruses. Phytopathology 80, 127–Citovsky, V., Knorr, D., and Zambryski, P. (1991). Gene I, a potential131.cell-to-cell movement locus of cauliflower mosaic virus, encodes an

Marsh, L. E., and Guilfoyle, T. J. (1987). Cauliflower mosaic virus replica-RNA-binding protein. Proc. Natl. Acad. Sci. USA 88, 2476–2480.tion intermediates are encapsidated into virion-like particles. VirologyCitovsky, V., Wong, M. L., Shaw, A. L., Prasad, B. V. V., and Zambryski,161, 129–137.P. (1992). Visualization and characterization of tobacco mosaic virus

Maule, A. J. (1991). Virus movement in infected plants. Crit. Rev. Plantmovement protein binding to single-stranded nucleic acids. PlantSci. 9, 457–473.Cell 4, 397–411.

McClure, M. A. (1993). Evolutionary history of reverse transcriptase. InDasgupta, I., Hull, R., Eastop, S., Poggi-Pollini, C., Blakebrough, M.,‘‘Reverse Transcriptase’’ (A. M. Skalka and S. P. Goff, Eds.), pp. 425–Boulton, M. I., and Davies, J. W. (1991). Rice tungro bacilliform virus444. Cold Spring Harbor Laboratory Press, Plainview, NY.DNA independently infects rice after Agrobacterium-mediated trans-

McLean, B. G., Waigmann, E., Citovsky, V., and Zambryski, P. (1993).fer. J. Gen. Virol. 72, 1215–1221.Cell-to-cell movement of plant viruses. Trends Microbiol. 1, 105–109.Deom, C. M., Lapidot, M., and Beachy, R. N. (1992). Plant virus move-

Medberry, S. L., Lockhart, B. E. L., and Olszewski, N. E. (1990). Proper-ment proteins. Cell 69, 221– 224.ties of Commelina yellow mottle virus’s complete DNA sequence,Ding, B., Li, Q., Nguyen, L., Palukaitis, P., and Lucas, W. J. (1995). Cu-genomic discontinuities and transcript suggest that it is a pararetrovi-cumber mosaic virus 3a protein potentiates cell-to-cell trafficking ofrus. Nucleic Acids Res. 18, 5505–5513.CMV RNA in tobacco plants. Virology 207, 345–353.

Melcher, U. (1990). Similarities between putative transport proteins ofDucasse, D. A., Mushegian, A. R., and Shepherd, R. J. (1995). Gene Iplant viruses. J. Gen. Virol. 71, 1009–1018.mutants of peanut chlorotic streak virus, a caulimovirus, replicate in

Melcher, U. (1993). HIV-1 proteinase as structural model of intercellularplants but do not move from cell to cell. J. Virol. 69, 5781–5786.transport proteins of plant viruses. J. Theor. Biol. 162, 61–74.Fuetterer, J., and Hohn, T. (1987). Involvement of nucleocapsids in re-

Mesnard, J.-M., and Carriere, C. (1995). Comparison of packaging strat-verse transcription: A general phenomenon? Trends Biochem. Sci.egy in retroviruses and pararetroviruses. Virology 213, 1–6.12, 92–95.

Fujiwara, T., Giesman-Cookmeyer, D., Ding, B., Lommel, S. A., and Lu- Montoya, A. L., Chilton M.-D., Gordon, M. P., Sciaky, D., and Nester,E. W. (1977). Octopine and nopaline metabolism in Agrobacteriumcas, W. J. (1993). Cell-to-cell trafficking of macromolecules through


368 TZAFRIR ET AL.

tumefaciens and crown gall tumor cells: Role of the plasmid genes. L. (1995). Grapevine fanleaf nepovirus P38 putative movement proteinis located on tubules in vivo. Mol. Plant–Microbe Interact. 8, 379–J. Bacteriol. 129, 101–107.

Mueller, P. R., and Wold, B. (1989). In vivo footprinting of the muscle 387.Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘Molecular Clon-specific enhancer by ligation mediated PCR. Science 246, 780–786.

Murashige, T., and Skoog, F. (1962). A revised medium for rapid growth ing—A Laboratory Manual,’’ 2nd ed. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY.and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473–

497. Storms, M. M. H., Kormelink, R., Peters, D., van Lent, J. W. M., and Gold-bach, R. W. (1995). The nonstructural Nsm protein of tomato spottedMushegian, A. R., and Koonin, E. V. (1993). Cell-to-cell movement of

plant viruses: Insights from amino-acid sequence comparisons of wilt virus induces tubular structures in plant and insect cells. Virology214, 485–493.movement proteins and from analogies with cellular transport sys-

tems. Arch. Virol. 133, 239– 257. Thomas, C. L., and Maule, A. J. (1995). Identification of the cauliflowermosaic virus movement protein RNA-binding domain. Virology 206,Nolasco, G., de Blas, C., Torres, V., and Ponz, F. (1993). A method

combining immunocapture and PCR amplification in a microtiter 1145–1149.Thomas, C. L., Perbal, C., and Maule, A. J. (1993). A mutation of cauli-plate for the detection of plant viruses and subviral pathogens. J.

Virol. Methods 45, 201–218. flower mosaic virus gene I interferes with virus movement but notvirus replication. Virology 192, 415–421.Noueiry, A. O., Lucas, W. J., and Gilbertson, R. L. (1994). Two proteins

of a plant DNA virus coordinate nuclear and plasmodesmal transport. Thomas, C. M., Hull, R., Bryant, J. A., and Maule, A. J. (1985). Isolationof a fraction from cauliflower mosaic virus-infected protoplasts whichCell 76, 925–932.

Perbal, M.-C., Thomas, C. L., and Maule, A. J. (1993). Cauliflower mosaic is active in the synthesis of (/) and (0) strand viral DNA and reversetranscription of primed RNA templates. Nucleic Acids Res. 13, 4557–virus gene I product (P1) forms tubular structures which extend from

the surface of infected protoplasts. Virology 195, 281–185. 4576.van Lent, J., Storms, M., van der Meer, F., Wellink, J., and Goldbach, R.Pfeiffer, P., and Hohn, T. (1983). Involvement of reverse transcription in

the replication of cauliflower mosaic virus: A detailed model and test (1991). Tubular structures involved in movement of cowpea mosaicvirus are also formed in infected cowpea protoplasts. J. Gen. Virol.of some aspects. Cell 33, 781 –789.

Prasad, V. R. (1993). Genetic analysis of retroviral reverse transcriptase 72, 2615–2623.Waigmann, E., Lucas, W. J., Citovsky, V., and Zambryski, P. (1994). Directstructure and function. In ‘‘Reverse Transcriptase’’ (A. M. Skalka and

S. P. Goff, Eds.), pp. 135–162. Cold Spring Harbor Laboratory Press, functional assay for tobacco mosaic virus cell-to-cell movement pro-tein and identification of a domain involved in increasing plasmo-Plainview, NY.

Qu, R., Bhattacharyya, M., Laco, G. S., De Kochko, A., Rao, B. L. S., desmal permeability. Proc. Natl. Acad. Sci. USA 91, 1433–1437.Wellink, J., van Lent, J. W. M., Verver, J., Sijen, T., Goldbach, R. W., andKaniewska, M. B., Elmer, J. S., Rochester, D. E., Smith, C. E., and

Beachy, R. N. (1991). Characterization of the genome of rice tungro van Kammen, A. (1993). The cowpea mosaic virus M RNA-encoded48-kilodalton protein is responsible for induction of tubular structuresbacilliform virus: Comparison with Commelina yellow mottle virus

and caulimoviruses. Virology 185, 354–364. in protoplasts. J. Virol. 67, 3660–3664.Wieczorek, A., and Sanfacon, H. (1993). Characterization and subcellu-Reiter, R. S., Young, R. M., and Scolnik, P. A. (1992). Genetic linkage of

the Arabidopsis genome: Methods for mapping with recombinant lar localization of tomato ringspot nepovirus putative movement pro-tein. Virology 194, 734–742.inbreds and random amplified polymorphic DNAs (RAPDs). In ‘‘Meth-

ods in Arabidopsis Research’’ (C. Koncz, N.-H. Chua, and J. Schell, Yang Z. (1995). ‘‘An Analysis of the Nucleotide Sequence and GenomicOrganization of Kalanchoe Top Spotting Virus.’’ University of Minne-Eds.), pp. 170–190. World Scientific, Singapore.

Ritzenthaler, C., Schmit, A.-C., Michler, P., Stussi-Garaud, C., and Pinck, sota, St. Paul. [M.S. thesis]


Documents

The N-Terminal Portion of the 216-kDa Polyprotein of ... · VIROLOGY 232, 359–368 (1997) ARTICLE NO. VY978569 The N-Terminal Portion of the 216-kDa Polyprotein of Commelina Yellow