UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Molecular characterization of the cucumber necrosis virus coat protein gene McLean, Morven A. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_fall_phd_mclean_morven.pdf [ 9.25MB ]
JSON: 831-1.0098844.json
JSON-LD: 831-1.0098844-ld.json
RDF/XML (Pretty): 831-1.0098844-rdf.xml
RDF/JSON: 831-1.0098844-rdf.json
Turtle: 831-1.0098844-turtle.txt
N-Triples: 831-1.0098844-rdf-ntriples.txt
Original Record: 831-1.0098844-source.json
Full Text

Full Text

MOLECULAR CHARACTERIZATION OF THECUCUMBER NECROSIS VIRUS COAT PROTEIN GENEbyMorven Anne McLeanB.Sc.(Agr.), McGill University, 1985M.Sc., University of Guelph, 1988A THESIS SUBMITTED IN PARTIAL FULFILMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Plant Science)We accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIAOctober, 1993© Morven Anne McLean, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of Q^Skj■QACQ___The University of British ColumbiaVancouver, CanadaDate Cipt^t iqcf)DE-6 (2/88)ABSTRACTA cucumber necrosis tombusvirus (CNV) mutant which lacks thecoding sequence for the coat protein protruding domain, PD(-), wasconstructed by site-directed mutagenesis of an infectious CNV cDNAclone, pK2/M5 (wild-type, 4701 nt). Transcripts of PD(-) wereinfectious in Nicotiana clevelandii, however, the resulting locallesions were significantly smaller than those on the correspondingleaves of plants inoculated with wild-type transcript. Inaddition, systemic symptoms took 8 to 12 days longer to developthan in wild type-inoculated plants. N. clevelandii inoculatedwith sap from systemically infected leaves of PD(-) transcript-inoculated plants showed local lesions the same as wild-type andthe delay in systemic symptom development was either reduced or,occasionally, lost. Systemic symptoms in both PD(-) transcript-and sap-inoculated N. clevelandii were either similar to thesystemic necrosis of wild-type or an oak-leaf necrosis wasobserved. High molecular weight RNA from mutant- and wild-type-infected plants was extracted and analyzed by Northern blotting.Full-length PD(-) RNA could be detected only rarely in RNApreparations from transcript-inoculated leaves; a further deleted,stable RNA species of approximately 3800 nucleotides was found inpreparations from systemically infected leaves of PD(-) transcript-and sap-inoculated plants. CNV coat protein could not be detectedby ELISA or ISEM in PD(-)-infected leaf material. The ca. 3.8 kbRNA, when cloned and sequenced, was found to have lost all but 74of the 1140 nucleotides of the CNV coat protein open reading frame.iiTranscripts from this coat-proteinless CNV cDNA clone producedsymptoms on N. clevelandii that were similar to those induced inPD(-)-sap-inoculated plants; wild-type and oak-leaf necroses wereobserved and there was a trend for systemic symptoms to be delayedwhen compared with wild-type. It would appear that CNV is able toreplicate and move systemically, in both transcript-inoculated andsap-inoculated N. clevelandii, in the absence of a functional coatprotein. Additionally, mechanical transmission of this virusoccurs in the absence of the coat protein, however, suchtransmission is less efficient when compared with wild-typeinfections.CNV and tomato bushy stunt tombusvirus (TBSV) have the samegenome organization and share high levels of nucleotide and aminoacid sequence similarity throughout their genomes, yet differbiologically. CNV, for which the natural vector is the fungusOlpidium radicale, has been found only in greenhouse-growncucumbers. TBSV, by contrast, has a relatively diverse naturalhost range and has no known vector. The role of the CNV coatprotein in determining transmission by 0. radicale was examined byexchanging the coat protein genes contained in full-length cDNAclones of CNV and TBSV. Transcripts from the resulting chimericclones (CNV with TBSV coat protein and vice versa) were infectiousin N. clevelandii. Virus particles, produced in infected plants,were purified and tested for in vitro acquisition and transmissionto roots of cucumber seedlings by zoospores of a cucumber strainof O. radicale. Particles produced by the TBSV chimeric cloneiiiencoding the CNV coat protein were transmitted, while thoseproduced by the CNV chimeric clone encoding the TBSV coat proteinwere not. These results demonstrate that the coat protein gene ofthe CNV genome is specifically involved in fungus transmissibility.ivTable of ContentsABSTRACT  ^iiLIST OF FIGURES  ^viiiLIST OF TABLES ^LIST OF ABBREVIATIONS  ^xiACKNOWLEDGEMENTS  ^xvCHAPTER I GENERAL INTRODUCTION  ^11.1 CHARACTERISTICS OF PLANT VIRUSES1.1.1 Biological characteristics  ^11.1.2 Physical characteristics  11.1.3 Virus-encoded proteins and their functions^ Structural proteins  ^ Non-structural proteins  31.1.4 Translation strategies of ssRNA plantviruses  ^ Segmentation of the genome^ Sub-genomic RNAs^ Proteolytic processing  ^ Suppression of read-throughtermination codons  ^ Translational frameshifting^ Internal initiation  ^ Leaky scanning  91.2 TOMBUSVIRUSES  ^101.2.1 Biological Characteristics  ^101.2.2 Cytopathology  ^111.2.3 Genome organization  ^111.2.4 Defective interfering and satellite RNAs^141.3 CUCUMBER NECROSIS VIRUS  ^151.3.1 Biological characteristics  ^151.3.2 Physical characteristics  ^161.3.3 Molecular biology of CNV  ^16CHAPTER II. MUTATIONAL ANALYSIS OF THE CNV COAT PROTEINGENE  ^19II.1 INTRODUCTION^ 1911.1.1 The capsid structure of small sphericalplant viruses  ^1911.1.2 The capsid structure of the tombusviruses^1911.1.2.1 The shell domain  ^2111.1.2.2 The protruding domain  ^2111.1.2.3 The amino-terminal arm ^ 2311.1.3 The movement of plant viruses within theirhosts  ^2311.1.3.1 Cell-to-cell movement  ^2411.1.3.2 Long distance movement  ^2611.1.3.3 The viral coat protein and its role invirus movement  ^2611.1.3.4 Other roles for the coat protein in thevirus life cycle  ^2711.1.4 Research objective  ^30^11.2 MATERIALS AND METHODS  ^3111.2.1 Basic molecular cloning techniques  ^3111.2.2 In vitro mutagenesis of the CNV coatprotein gene  ^3111.2.2.1 The Kunkel method of in vitromutagenesis  ^3111.2.2.2 Purification of single-stranded DNA fromCJ236  ^3211.2.2.3 Annealing of the mutagenicoligonucleotides  ^3411.2.2.4 Complementary strand synthesis^3411.2.2.5 Purification of syntheticoligonucleotides  ^3511.2.2.6 Phosphorylation of syntheticoligonucleotides  ^3611.2.3 Construction of CNV cDNA coat protein clonemutants  ^3611.2.4 Construction of CNV cDNA clones for in vitro translation  ^3811.2.5 In vitro transcription  ^4211.2.6 In vitro translation  ^4211.2.7 Inoculation of Nicotiana clevelandii  ^4311.2.8 Local lesion assay in Chenopodium quinoa .^4411.2.9 ELISA  ^4511.2.10 Immunoelectron microscopy  ^4611.2.11 Leaf RNA extraction  ^4611.2.12 Northern blots  ^4711.2.12.1 Denaturation and transfer ofRNA  ^4711.2.12.2 Nick-translated probes  ^4811.2.12.3 Random-primed probes  ^4911.2.13 Cloning of RNA extracted from PD(-)-infected plants  ^4911.2.13.1 Reverse transcription andpolymerase chain reactions  ^4911.2.13.2 Cloning of RT-PCR products  ^5111.3 RESULTS  ^5211.3.1 Symptomatology and infectivity of PD(-) inN. clevelandii  ^5211.3.2 Host range study  ^56TT 1^Tmmnroeler'tron microccopy  ^57vi11.3.4 ELISA  ^5811.3.5 In vitro translation  ^5811.3.6 Analysis of viral RNA species in infectedleaf extracts  ^6211.3.7 Local lesion assay of PD(-) in C. quinoa^.  ^ 6611.3.8 Cloning, sequencing, and infectivity of the ca.^3800 nt. RNA  ^6811.4 DISCUSSION  ^78CHAPTER III. FUNGAL TRANSMISSION OF CNV: A STUDY USINGCHIMERIC PLANT VIRUSES  ^89III.1 INTRODUCTION  ^89III.1.1 The life cycle of fungal vectors of plantviruses  ^89111.1.2 Plant viruses transmitted by fungi  ^91111.1.2.1 Viruses acquired in vivo  ^91111.1.2.2 Viruses acquired in vitro  ^93111.1.3 Transmission of CNV by 0. radicale  ^94111.1.4 The role of the CNV coat protein in fungaltransmission  ^95111.1.5 Research objective  ^97111.2 MATERIALS AND METHODS  ^98111.2.1 Production of chimeric CNV/TBSV cDNAclones  ^98III.2.1.1 Sub-cloning of the CNV and TBSVcoat protein genes  ^98111.2.1.2 Production of the chimeric clones .^100111.2.2 Purification of pMM100 and pMM200 virions . 102111.2.3 Fungal transmission of purified CNV wild-type, TBSV101, pMM100, and pMM200virions  ^103111.2.4 Local lesion assay in C. quinoa  104111.2.5 ELISA  ^104111.3 RESULTS  ^106111.3.1 Symptomatology and infectivity of pMM100and pMM200 in N.clevelandii  ^106111.3.2 Purification of pMM100 and pMM200virions  ^108111.3.3 ELISA  108111.3.4 Fungus transmissibility of pMM100 andpMM200 virions  ^111111.4 DISCUSSION  ^114CHAPTER IV. FUTURE DIRECTIONS  ^122BIBLIOGRAPHY ^ 124viiList of FiguresFig. 1.1. The genome organization of cucumber necrosisvirus ^12Fig. 2.1. The three-dimensional structure of the tombusviruscapsid.  ^20Fig. 2.2 Schematic diagram indicating the sequence of eventsleading to the creation of the CNV P-domainlessclone,PD(-) 37Fig. 2.3. Schematic diagram indicating the sequence of eventsleading to the creation of clones used for in vitro translation studies   41Fig. 2.4. Summary of the organization of the CNV genome andthe location of coat protein mutations  ^53Fig. 2.5. Inoculated leaves of N. clevelandii 6 days afterinoculation  ^54Fig. 2.6. Systemically infected N. clevelandii 22 days aftertranscript inoculation.  ^55Fig. 2.7. In vitro translation products of synthetic CNV coatprotein subgenomic RNA in wheat germ extracts ^61Fig. 2.8. The in vitro translation product of a syntheticPD(-) coat protein subgenomic RNA ^63Fig. 2.9. Northern blot analysis of high molecular weight RNAextracted from leaves of PD(-)-, XpK2M5- and wild-typetranscript- and sap-inoculated plants  64Fig. 2.10. Local lesion assay of PD(-), wild-type, and CNV RNAin C. quinoa.  ^67Fig. 2.11.^DNA products recovered after RT-PCR of high-molecular weight ssRNA extracted from N. clevelandii after the fourth passage of wild-type or PD(-)inoculum.   70Fig. 2.12. Nucleotide sequences of eight independent clonesof the RT-PCR products of the ca. 3800 nt RNA^71Fig. 2.13.^Location of CP(-) coat protein nucleotidesequences relative to wild-type and PD(-) coat proteinsequences ^72viiiFig. 2.14. Systemically infected N.transcript inoculation. ^clevelandii 7 days after75 Fig. 2.15. Systemically infected N.transcript inoculation. ^clevelandii 22 days after76Fig. 2.16. Northern blot analysis of high molecular weight RNAextracted from the inoculated- and systemically-infectedleaves of transcript-inoculated, CP(-)-infected N.clevelandii   77Fig. 3.1. Schematic diagram indicating the sequence of eventsleading to the creation of the CNV/TBSV chimeric cDNAclones, pMM100 and pMM200  99Fig. 3.2. Diagram illustrating the structure of the chimericcDNA CNV/TBSV clones, pMM100 and pMM200 ^107Fig. 3.3^Electron micrographs of the negatively stainedvirions purified from N. clevelandii infected with CNVwild-type, pTBSV101, pMM100, and pMM200 ^109Fig. 3.4 Diagram of a CNV/MNSV chimeric clone ^121ixList of TablesTable 2.1. DAS-ELISA A405 values for leaf sap extracted fromXpK2/M5-, PD(-)-, and buffer- inoculated N. clevelandii using a polyclonal antiserum raised against CNVvirions  60Table 3.1 Fungal vectors of plant viruses (Adams, 1988;Adams, 1991; Brunt, 1988)  ^92Table 3.2 DAS-ELISA A405 values for purified CNV, pTBSV101,pMM100, and pMM200 virions using polyclonal antiseraraised against CNV or TBSV virions.   110Table 3.3. Bio-assay to determine the fungus transmissibilityof the CNV/TBSV chimeric viruses, pMM100 and pMM200.^113xLIST OF ABBREVIATIONSa - armA260 - absorbance at 260 nmAMCV - artichoke mottled crinkle virusATP (dATP) - adenosine triphosphate (deoxyadenosine triphosphate)BE - borate/EDTABNYVV - beet necrotic yellow vein virusby - base pairBRL - Betheseda Research LaboratoriesBSA - bovine serum albuminC - Celsiusca. - approximatelyCaMV - cauliflower mosaic viruscDNA - complementary DNACi (1Ci) - Curie (microCurie)CIRV - carnation Italian ringspot virusCLSV - cucumber leaf spot virusCNV - cucumber necrosis virusCPMV - cowpea mosaic virusCsC1 - cesium chlorideCTP (dCTP) - cytidine triphosphate (deoxycytidine triphosphate)CyRSV - cymbidium ringspot virusDAS-ELISA - double antibody sandwich ELISADI-RNA - defective interfering RNADNA - deoxyribonucleic aciddpi - days post-inoculationxiDTT - dithiothreitolEDTA - ethylenediaminetetraacetic acidEGTA - ethyleneglycol-bis-(S-amino-ethyl ether) N,N'-tetra-aceticacidELISA - enzyme-linked immunosorbent assayg (mg, gg) - gram (milligram, microgram)g - centrifugal forceGTP (dGTP) - guanosine triphosphate (deoxyguanosine triphosphate)HC1 - hydrochloric acidHPLC - high pressure liquid chromatographyISEM - immunosorbent electron microscopykb - kilobasekbp - kilobase pairKC1 - potassium chloridekDa - kiloDalton1 (ml, gl) - litre (millilitre, microlitre)LB - Luria-BertaniLiC1 - lithium chlorideM-MLV - Moloney murine leukaemia virusMCS - multi-cloning siteMgC1 2 - magnesium chloridemin - minutemm - millimetreMNSV - melon necrotic spot virusmRNA - messenger RNANaAc - sodium acetateNaCl - sodium chloridcxiint - nucleotideOD - optical densityORF - open reading frameP - protruding domainPAGE - polyacrylamide gel electrophoresisPAMV - petunia asteroid mosaic virusPBS - phosphate-buffered salinePLCV - pelargonium leaf curl virusR - random domainRCNMV - red clover necrotic mosaic virusRNA - ribonucleic acidrpm - revolutions per minuteRT-PCR - reverse transcription polymerase chain reactionS - shell domains - sedimentation coefficientS-RNA - satellite RNASBMV - southern bean mosaic virusSBWMV - soil borne wheat mosaic virusSDS - sodium dodecyl sulfatess - single-strandedSSC - sodium chloride/trisodium acetateTAE - Tris/acetate/EDTATBSV - tomato bushy stunt virusTCV - turnip crinkle virusTE - Tris/EDTATMV - tobacco mosaic virusTHE - Tris/NaCl/EDTATNV - tobacco necrosis virusTomRSV - tomato ringspot virusTris - trishydroxymethylaminomethanedTTP - deoxythymidine triphosphateUBC - University of British ColumbiaUSB - United States BiochemicalVPg - genome-linked viral proteinxivACKNOWLEDGEMENTSI would like to extend sincere thanks to my graduate researchsupervisors, Dr. D.M. Rochon and Dr. R.I. Hamilton, for introducingme to the fascinating world of plant virology, and for theirguidance and support during the course of my studies. I would alsolike to thank the members of my graduate committee, Dr. R.J.Copeman, Dr. C. Douglas, and Dr. H.S. Pepin, for theircontributions to this work. I am indebted to Dr. R.N. Campbell,University of California, Davis, for his kind cooperation with thefungus transmission experiments, and Dr. T.J. Morris for thegenerous gift of pTBSV101. My appreciation is extended to Dr. M.Weintraub, Past-Director, and Dr. D.L. Struble, Director, of theVancouver Research Station, Agriculture Canada, for providing theopportunity to work at the Research Station. Financial assistancefrom the Natural Sciences and Engineering Research Council and theUniversity of British Columbia is gratefully acknowledged.XVCHAPTER I GENERAL INTRODUCTIONPlant viruses are important plant pathogens, second only tothe plant pathogenic fungi in terms of attributed global annualcrop losses. Diseases caused by plant viruses are of particularconcern in tropical and sub-tropical regions where continuouscropping is practised and where viral vectors are able to thriveyear round.I.1 CHARACTERISTICS OF PLANT VIRUSESI.1.1 Biological characteristicsPlant viruses are able to infect a very broad group of hosts:monocotyledonous or dicotyledonous plants; annuals, biennials, orperennials; and herbaceous or woody species. Plant viruses aretransmitted from one plant to another by a variety of means. Avirus may be vectored by an arthropod, fungus, or nematode or itmay move via seed, pollen, or tissue grafts. Viruses may also betransmitted mechanically, by rubbing extracts of diseased leavesor other tissues on healthy leaves, or by any of a number ofcultural methods currently used in crop production in whichextracts from diseased plants are deposited in wounds eg. bypruning, mowing, etc. Many viruses are transmitted by acombination of two or more of these types of transmission.1.1.2. Physical characteristicsA plant virus has been defined as "a set of one or morenucleic acid template molecules, normally encased in a protective1coat or coats of protein or lipoprotein, that is able to organizeits own replication only within suitable host cells" (Matthews,1991).Plant viruses may be spherical-, bacilliform-, or rod-shapedand they may be mono- or multi-partite, i.e. the genome may becontained in a single capsid structure (particle) or distributedamong two or more similar capsid structures (particles). Thecapsids of some plant viruses are surrounded by a lipoproteinenvelope. The viral nucleic acid may be single-stranded or double-stranded, RNA or DNA, and may have a positive or a negativepolarity. The majority of plant viruses have a single-stranded(ss), positive-sense, RNA genome that can be translated directlyupon infection of a host cell. The genome sizes of plant virusesare generally smaller than those of animal viruses, and theytypically code for 1 to 12 proteins (Matthews, 1991).Many ssRNA viruses have specialized structures at their 5' and3' termini that may have regulatory or recognition functions inviral replication. These include: 1) a 5'cap, similar to the capstructure present at the 5' termini of almost all eukaryotic mRNAs(the first two nucleotides of the cap in eukaryotic, but not plantvirus, mRNAs are methylated); 2) a small protein (VPg) that iscovalently linked to the 5' end of the genomic RNA; 3) apolyadenylate sequence, of variable length, at the 3' end of themRNA; 4) a tRNA-like structure bound to the 3' terminus of thegenomic RNA that may be specifically aminoacylated; and 5) a 3'hydroxyl group.21.1.3 Virus-encoded proteins and their functionsThe genomes of plant viruses encode both structural and non-structural proteins. The structural proteins are those thatcontribute to the capsid structure such as the coat protein, thecore proteins of the reoviruses, and the lipoprotein membranes ofthose viruses that are enveloped. Non-structural proteins includeviral polymerases, proteases, proteins involved in viral movement,transmission, and the VPg. Proteins produced by ssRNA plantviruses are described below. Structural proteinsThe functions of the plant virus coat protein are discussedin sections 11.1.1 and Non-structural proteins1) Viral polymerasesSingle-stranded RNA viruses encode their own RNA-dependent RNApolymerase, which associates with other viral- or host-encodedproteins to comprise the viral replicase, that is required by thevirus for replication of its genomic RNA. The (putative)polymerases of most plant viruses have been identified by thepresence of conserved amino acid sequence motifs, such as the Gly-Asp-Asp motif described by Kamer and Argos (1984), and a 14 to 15amino acid motif identified by Hodgman (1986). Host immunity toplant virus infection may involve a direct effect on the viralpolymerase; protoplasts from some plant species or cultivars have3been found to be resistant to virus replication, suggesting thatthe polymerase is not able to form a functional replicase in theseplant species (van Loon, 1987) Loss of virulence by plant viruseshas been associated with a decrease in virus multiplication,implying that changes to the viral polymerase that result inreduced virus multiplication can affect symptom expression (vanLoon, 1987).2) HelicasesHelicase proteins are involved in duplex unwinding duringnucleic acid replication, transcription, recombination, repair,and, possibly, mRNA translation and RNA splicing (Gorbalenya andKoonin, 1989; Habili and Symons, 1989), and comprise part of theviral replicase. A putative helicase may be assigned its functionon the basis of a series of motifs described by Hodgman (1988),Gorbalenya et al. (1988), Gorbalenya and Koonin (1989), and Habiliand Symons (1989). The helicase and polymerase motifs of positive-sense RNA plant viruses have been found in the same, or different,open reading frames, and on the same, or different, genomic RNAs(Habili and Symons, 1989).3) Viral proteasesThe replication of the poty-, como-, and nepoviruses, amongstothers, is dependent on proteolytic processing of a polyprotein torelease the mature virus proteins. The enzymes responsible for thecleavage are virus-encoded and are involved in the cleavage of4structural from non-structural proteins, the generation of matureenzymes, and the coordinated assembly of the mature virion(Krausslich and Wimmer, 1988). Cowpea mosaic comovirus (CPMV),which has a bi-partite genome, encodes a protease, the 24 kDaprotein, which is responsible for all the proteolytic cleavages ofthe CPMV M- and B- polyproteins (Krausslich and Wimmer, 1988). Thesingle polyprotein produced by tobacco etch potyvirus (TEV) iscleaved by a viral-encoded protease, the 49 kDa protein. The TEVprotease is autocatalytically released from the polyprotein andthen cleaves in trans at three other sites to yield five viral geneproducts (Dougherty et al., 1988). The amino acid residuessurrounding the polyprotein cleavage site are conserved amongst thepotyviruses. The amino acids on the amino-terminal side of thecleavage site are not absolutely required for cleavage but help todefine an optimal cleavage site (Dougherty et al.,1988). Doughertyet al. (1988) suggested that the ability of the protease torecognize similar but sub-optimal cleavage sites may be importantfor post-translational regulation of potyviral gene function.Cleavage at such sites late in the viral life cycle may inactivatean early function or activate a late one.4) Movement proteinsThe function of plant virus movement proteins is discussed insection Helper component proteins5Both the potyviruses and caulimoviruses encode proteins,referred to as helper components, that are required for non-persistent aphid transmission of these viruses (Pirone, 1981;Armour et al., 1983). Non-aphid transmissible strains of theseviruses can be successfully transmitted if the aphid feeds firston a plant infected with a strain encoding the helper component,and then on a plant infected with a non-aphid transmissible strain.6) Genome-linked viral protein (VPg)Members of the poty-, como-, sobemo-, luteo- and nepovirusgroups have a VPg at the 5' termini of their viral RNAs. The VPghas been speculated to function as a primer for RNA synthesis(Matthews, 1991) and is required by some viruses, such as tobaccoringspot virus, for viral infectivity (Mayo et al., 1982).7) Viral-encoded proteins involved in symptomatologyDisease induction and symptomatology in a virus-infected plantmay be influenced by changes to one or more of the viral-encodedgenes. For example, Nishiguchi et al. (1985) reported that asingle amino acid change in the putative polymerase gene of tobaccomosaic tobamovirus was involved in attenuation of symptoms. Beetnecrotic yellow vein furovirus RNA 3, while not being required forinfectivity, significantly affects symptom expression (Jupin etal., 1990). The role of coat protein in symptomatology is reviewedin sections and Translation strategies of ssRNA plant virusesMost eukaryotic mRNAs are monocistronic and their 80sribosomes are adapted to translate only the 5' proximal gene on anymRNA. A scanning model for initiation of translation in eukaryoteshas been proposed by Kozak (1978, 1989). In this model the 40sribosomal subunit binds to the cap at the 5' end of the mRNA andthen scans along the mRNA to the first downstream AUG codon. Ifthe AUG is in a favourable context for initiation, the 60sribosomal subunit binds to the 40s subunit and translation begins.Optimal consensus sequences for initiation of translation have beenestablished for animal and plant mRNAs (Kozak, 1986a; Lutke et al.,1987).Although plant viruses are genetically polycistronic, they aregenerally functionally monocistronic as they must depend on theprotein-synthesis machinery of their eukaryotic host fortranslation. Plant viruses have evolved a number of strategies,described below, to ensure that their downstream genes aretranslated. Any plant virus may employ one or more of these in thetranslation of its genome. Segmentation of the genomeThe genomes of plant viruses may be divided so that they arebipartite (nepoviruses), tripartite (bromoviruses), orquadripartite (beet necrotic yellow vein virus). Segmentation ofthe genome: 1) increases the number of 5' proximal genes that areaccessible to the host's translational machinery; 2) may be used7to physically or temporally separate various genes; 3) may reducethe probability that errors in replication will be deleterious tothe virus; and 4) may promote genetic reassortment duringreplication (Matthews, 1991). Subgenomic RNAsSeveral groups of plant viruses (eg. tobamoviruses,tombusviruses, sobemoviruses) are able to express downstream genesthrough the production of 3' co-terminal, subgenomic RNAs. The 5'open reading frame on each subgenomic mRNA can be translated byeukaryotic ribosomes. Because subgenomic mRNA synthesis is thoughtto occur later in viral replication, it may be an important meansfor viruses to regulate the timing of gene expression. Proteolytic processingThe mRNAs of the nepo-, poty-, and comoviruses, amongstothers, are translated as a single long polyprotein. Thepolyprotein is cleaved post-translationally at specific sites intotwo or more functional gene products by virus-encoded proteases. Suppression of read-through termination codonsPlant viruses such as tobamoviruses, tombusviruses,carmoviruses and furoviruses, have a termination codon at the endof the 5' open reading frame of the genomic mRNA that may besuppressed. This permits ribosomes to continue translation intothe region downstream of the "leaky" termination codon until8another termination codon is reached. Translational frameshiftingAs the ribosomal complex approaches the termination codon ofa 5' gene, it may switch reading frames and continue translationuntil a second termination codon is reached. Translationalframeshifting has been suggested for potato leaf roll and barleyyellow dwarf luteoviruses (Miller et al., 1988; Brault and Miller,1992) and reported for red clover necrotic mosaic dianthovirus(Xiong and Lommel, 1989). Internal initiationScanning-independent internal initiation, first reported forpoliovirus (Pelletier and Sonenberg, 1988), occurs when ribosomesinitiate translation at an internally located AUG. Internalinitiation has been reported for cowpea mosaic comovirus M RNA(Verver et al., 1991). Leaky scanningLeaky scanning (Kozak, 1986b) may occur if the first AUG codonat the 5' end of an mRNA is in a sub-optimal context fortranslation initiation. The scanning ribosome may pass over thefirst AUG and initiate translation at a second AUG that has astronger consensus sequence. Leaky scanning has been proposed fortranslation of the overlapping coat protein and 17K open readingframes (ORFs) of barley yellow dwarf virus subgenomic RNA 1 and the9bifunctional subgenomic mRNA of CNV (Dinesh-Kumar and Miller, 1992;Rochon and Johnston, 1991).1.2 TOMBUSVIRUSESMembers of the tombusvirus group of plant viruses are small,spherical viruses with a particle diameter of ca. 30 nm. Thetombusvirus capsid is constructed from 180 identical coat proteinsubunits (Mr 41000) arranged in a T=3 icosahedron. Their genome ismonopartite, and consists of a single-stranded, positive-sense RNAof approximately 4700 nt.The type member of the tombusvirus group, and the virus fromwhich the group name was derived, is tomato bushy stunt virus(TBSV). Other members include cymbidium ringspot virus (CyRSV),artichoke mottled crinkle virus (AMCV), carnation Italian ringspotvirus, pelargonium leaf curl virus, petunia asteroid mosaic virus(PAMV, which is the same as TBSV-Ch), eggplant mottled crinklevirus, and the serologically unrelated cucumber necrosis virus(CNV) (Martelli et al., 1988; Rochon and Tremaine, 1988).1.2.1 Biological CharacteristicsThe tombusviruses have a wide geographical distribution and,as a group, are able to infect a variety of woody and herbaceousdicotyledonous plants. Only TBSV and PAMV have a broad naturalhost range; the other members are restricted to infecting one ora few plant species only (Martelli et al., 1988). Thetombusviruses are soil-borne and it has been speculated that their10transmission is facilitated by fungi, however, only the vector ofCNV, Olpidium radicale Schwartz & Cook, has been identified to date(Dias, 1970a,b). The host ranges of the tombusvirus members canbe greatly extended by mechanical transmission.1.2.2 CytopathologyIn host cells infected with tombusviruses, three types ofcytoplasmic inclusion have been observed: 1) crystalline arrays ofvirus particles; 2) multi-vesicular bodies of peroxisomal ormitochondrial origin; and 3) dense granular inclusions (Martelliet al., 1988). Multi-vesicular bodies have been proposed as a siteof virus replication, and the dense granular inclusions, which donot occur in all tombusvirus-infected cells, are believed to be anaccumulation of excess virus coat protein (Appiano et al.,1986).1.2.3 Genome organizationThe complete nucleotide sequences of CNV, TBSV-Ch, and CyRSV,and a partial sequence of AMCV have been determined (Rochon andTremaine, 1989; Hearne et al., 1990; Greico et al., 1989; Greicoand Gallitelli, 1990).^All four share a similar genomeorganization, represented by CNV in Fig.1.1. The full-lengthtombusvirus genome has five long ORFs. The first, ORF 1, encodesa 33 kDa protein (p33). An amber termination codon at the end ofthe p33 ORF may be read through to produce a 92 kDa protein (p92),the putative polymerase. The 3 downstream ORFs are expressed111.1 Genome organization of cucumber necrosis virusthrough the production of two 3' co-terminal subgenomic messengermRNAs (Rochon and Tremaine, 1989). The larger of these, ca. 2.1kb in length, encodes the 41 kDa coat protein (p41) (Johnston andRochon, 1990). The smaller subgenomic mRNA, which is ca. 0.9 kb,is bifunctional (Rochon and Johnston, 1991) and encodes 2 proteins:the ORF for the 20 kDa protein (p20) is nested entirely within theORF for the 21 kDa protein (p21). The 21 kDa protein has beenproposed to function in cell-to-cell movement of the virus (Rochonand Johnston, 1991; Scholthof et al, 1992) and the 20 kDa proteinhas been suggested to be involved in viral RNA replication (Rochon,1991). Johnston and Rochon (1990) used hybrid-arrested translationand in vitro translation of synthetic CNV transcripts and CNVvirion RNA to identify the genomic locations of the 33, 41, 20 and21 kDa proteins. The 33 kDa protein was mapped to the 5' terminusof CNV RNA and the 41, 20 and 21 kDa proteins were mapped todownstream locations (Johnston and Rochon, 1990). A sixth shortORF, pX, has recently been proposed by Boyko and Karasev (1992) toexist within the previously identified 3'-terminal non-codingregions of TBSV, CyRSV, CNV, and AMCV. This ORF, if expressed,would encode a protein of 32 (CNV) to 69 (TBSV) amino acids. 5'and 3' terminal modifications have not been described for thetombusviruses. The 5' terminus of tombusvirus RNA may be capped(Rochon and Tremaine, 1989), although it has been found thataddition of cap analogue ('mGpppG) to transcription reactions doesnot enhance infectivity of in vitro-transcribed CNV or TBSV RNA(Rochon and Johnston, 1991; Hearne et al., 1990). Tombusviruses13have neither a poly(A) tail nor a tRNA-like structure at the 3'terminus, however, CNV, TBSV and CyRSV share extensive sequencesimilarity in their 5' and 3' non-coding regions. Hearne et al.(1990) speculate that these highly conserved regions may containreplication elements because the same sequences are also presentin the defective interfering RNAs associated with these viruses(section 1.2.4).1.2.4 Defective interfering and satellite RNAsIn addition to the genomic and subgenomic RNAs found intombusvirus infections, other small RNA species have been detected.These have been identified as either defective interfering RNAs (DIRNAs) or satellite RNAs (S-RNAs) (Hillman et al., 1987; Rochon,1991; Rubino et al., 1990).Defective interfering RNAs are deletion derivatives of thegenomic RNA that depend on the parental (helper) RNA forreplication and possibly for encapsidation as well. DI RNAs areoften associated with a decrease in both helper virus replicationand in the severity of symptoms normally induced by the helpervirus alone (Roux et al., 1991). Commonly associated with animalviruses, the first plant virus DI RNA was described by Hillman etal. (1987) for TBSV. DI RNAs have also been found in CNV- andCyRSV-infected plants (Rochon, 1991; Rubino et al., 1990), withseveral other plant viruses from different groups (Stanley andTownsend, 1985; Li et al., 1989; White et al., 1991; Resende etal., 1992; Romero et al., 1993)14Satellite RNAs, like DI RNAs, are dependant on a helper virusfor replication and encapsidation, but in contrast to DI RNAs, theyhave, by definition, no appreciable sequence similarity with theirhelper virus (Matthews, 1991). S-RNAs are associated with manytombusviruses including TBSV, CyRSV, and ACMV (Gallitelli and Hull,1985). Interestingly, the S-RNA of CyRSV was found to share twomajor regions of homology with CyRSV genomic RNA, which is unusualfor S-RNAs (Rubino et al., 1990). The S-RNAs of tombusviruses haveextensive sequence similarity with each other and are not strain-specific in terms of the helper virus (Gallitelli and Hull, 1985).Similar to DI RNAs, S-RNAs are associated with symptom attenuationin tombusvirus-infected plants.1.3 CUCUMBER NECROSIS VIRUS1.3.1 Biological characteristicsCNV was first observed as a pathogen of greenhouse-growncucumbers in Ontario in 1952 (McKeen, 1959). Infections, whichoriginated from a root invasion, were characterised by foliarchlorosis and malformation, and mottled and stunted fruit (McKeen1959). CNV was found to be transmitted by the soil-bornechytridiomycete, 0. radicale, [(formerly called O. cucurbitacearumBarr and Dias; Barr, 1968; Lange and Insunza, 1977); Dias, 1970a,b]. The natural host range of CNV is restricted to cucumber,but its host range can be greatly extended to a variety of locallesion hosts through mechanical transmission. Only Nicotiana clevelandii, N. benthamiana, and less consistently, cucumber, are15known to support systemic movement of this virus.A more comprehensive description of CNV transmission by 0.radicale is present in the introduction to Chapter III.1.3.2 Physical characteristicsA detailed description of the CNV particle structure is givenin the Introduction to Chapter II.1.3.3 Molecular biology of CNVCNV was originally considered as a possible member of themono-typic tobacco necrosis virus group (TNV), primarily becauseboth viruses are transmitted by Olpidium spp. (Teakle and Gold,1963; Dias, 1970a). Rochon and Tremaine (1988) determined throughnucleic acid hybridization studies that CNV RNA shared extensivenucleotide sequence similarity with TBSV RNA, but not with TNV RNA,and that the dsRNA profiles of TBSV and CNV are similar but differfrom that of TNV. Consequently, CNV was accepted as a definitivemember of the tombusvirus group although it is not serologicallyrelated to any of the other tombusviruses (Rochon and Tremaine,1988).The complete CNV RNA nucleotide sequence has been determinedand its genome organization deduced (Fig. 1.1; Rochon and Tremaine,1989). As described in section 1.2.3, CNV has the potential toencode proteins of molecular weight 33, 92, 41, 21, and 20 kDa.In order to assign functions to CNV proteins, comparisons were madebetween the amino acid sequences of the putative CNV proteins and16the amino acid sequences of proteins of known function (Rochon andTremaine, 1989).CNV p92 was found to share significant sequence similaritywith the putative replicases of carnation mottle carmovirus (CarMV)and barley yellow dwarf luteovirus. This observation by Rochon andTremaine (1989) led to the proposal that the tombus-, carmo-, andluteoviruses be classified as a third supergroup of viruses (theother two being the alpha- and picornavirus supergroups proposedby Goldbach, 1987). The proposition of a third viral supergroupwas further supported by Riviere and Rochon (1990) who reportedthat CNV, melon necrotic spot-, carnation mottle-, and turnipcrinkle carmoviruses, barley yellow dwarf luteovirus, and theunclassified virus, maize chlorotic mottle, lacked a proteincontaining the amino acid sequence motif that characterized thenucleotide-binding proteins of viruses from the alpha- andpicornavirus supergroups. Riviere and Rochon. (1990) suggested thatviruses within this third supergroup may share the same replicationstrategy as the members of the alpha- and picornavirus supergroups,but encode replicative proteins that differ significantly in theiramino acid composition, or may have a unique replication strategy.The third supergroup was later expanded to include thesobemoviruses, dianthoviruses, tobacco necrosis virus, and theanimal virus, Hepatitis C (Habili and Symons, 1989; Xiong andLommel, 1989; Rochon et al., 1991). CNV p92 was also found tocontain the glycine-aspartate-aspartate motif that ischaracteristic of the known and proposed replicases of many animal17and plant viruses (Kamer and Argos, 1984) and was consequentlyassigned a replicase function (Rochon and Tremaine, 1989). Acomparison of the deduced amino acid sequence of p41 and the aminoacid composition of the CNV coat protein determined by direct aminoacid analysis supported the assignment of p41 as the coat protein(Rochon and Tremaine, 1989; Tremaine, 1972). Subsequently, it wasfound that the p41 protein produced by in vitro translation of CNVRNA could be immunoprecipitated with CNV antiserum (Johnston andRochon, 1990). No functions for p20 and p21 were assigned althoughsubsequent work has provided evidence that p21 may function incell-to-cell movement and p20 may play a role in viral RNAreplication (Rochon and Johnston, 1991; Rochon, 1991).Rochon and Johnston (1991) provided evidence that the 20 and21 kDa proteins were produced in vivo and that they were bothexpressed from a single bifunctional subgenomic mRNA. Theyspeculated that translation of p20, whose ORF is nested entirelywithin that for p21, occurred by "leaky' scanning. Such expressionof p20 could occur if ribosomes scanned past the start codon forp21 and initiated translation instead at the p20 start codon(Rochon and Johnston, 1991).18CHAPTER II. MUTATIONAL ANALYSIS OF THE CNV COAT PROTEIN GENEII.1 INTRODUCTION11.1.1 The capsid structure of small spherical plant virusesThe three-dimensional structures of the capsids and coatprotein subunits of the small spherical plant viruses, tomato bushystunt tombusvirus (TBSV), turnip crinkle carmovirus (TCV), andsouthern bean mosaic sobemovirus (SBMV), are amongst the bestcharacterized of any of the plant or animal viruses studied todate. TBSV was the first virus whose complete capsid structure wasresolved to 2.9 A by high-resolution X-ray crystallography(Harrison et al., 1978) and, along with TCV and SBMV, hassubsequently provided much insight into the structural propertiesand associations of the individual coat protein subunits thatcomprise the capsid shells of these and, by extrapolation, otherrelated viruses.11.1.2 The capsid structure of the tombusvirusesThe capsid of the tombusviruses, and also the closely relatedcarmoviruses, is comprised of 180 identical coat protein subunitsunited to form a T=3 icosahedron (Harrison et al., 1978; Fig 2.1,A). Each coat protein subunit folds into three distinct domains:the random domain (R), the shell domain (S) and the protrudingdomain (P) (Harrison et al., 1978; Hopper et al., 1984) (Fig. 2.1,B). The R and S domains are joined by the arm (a) and the S andthe P domains are joined by a flexible, five amino acid hinge (h)that permits the conformational change required of the19Fig. 2.1. The three-dimensional structure of the tombusviruscapsid. A. A model showing the coat protein subunit arrangement.The protruding domain dimers (P) extend outward from the shelldomains (S). A, B, and C represent the three packagingenvironments of the subunit. S domains of A subunits pack aroundfive-fold axes, S domains of B and C subunits alternate aroundthree-fold axes. Subunit contacts which have true icosahedral two-fold symmetry are labelled s2, while those having local two-foldsymmetry are labelled q2. B. Coat protein sub-unit showing theS and P domains, the hinge (h), and the N-terminal arm (a). C. Thetwo conformational states of a coat protein subunit. Where thesubunit is in the C/C configuration, the hinge is "up" and theamino-terminal arms form a i3-annulus. Where the subunit is in theA/B conformation, the hinge is "down" and the amino-terminal armsare disordered. Reprinted from Harrison  e on•on , MacMillan Magazines Ltd.• - min •20coat protein to achieve icosahedral symmetry (Fig 2.1, C). The shell domainThe shell of the virus capsid is comprised of S domains thatassociate with each other non-covalently by trimer, pentamer, andhexamer subunit contacts (Harrison et al., 1978). Two classes ofCa2'-binding sites link neighbouring subunits together in a trimericinteraction. The Ca t'-binding sites are integral to maintainingcapsid conformation; if the Ca 2+ ions are removed, the particleswells in a pH-dependent fashion (Harrison et al., 1978). When thepH is raised above neutrality and the Ca t-' ions are removed, thearms of the N-terminal domain loop out of the expanded structure(Harrison, 1989; Fig 2.1, C). Subsequent sequestration of Ca 2+ fromvirus particles by the host cell and consequent swelling of thevirus particle may be a mechanism for release of viral RNA from thecapsid prior to the first round of replication, as speculated byLaasko and Heaton (1992) for TCV. The topology of the S domain hasbeen classified as a Class I "jellyroll" that is common to allother small, spherical, virus capsids so far analyzed. In manyanimal viruses the "jellyroll" domain functions both in forming thecapsid structure and also in recognition of host cell receptorsites (Gibson and Argos, 1990). The protruding domainThe P domain is unique to the coat proteins of the tombus-,carmo-, and diantho-, plant virus groups. The P domain of each21coat protein subunit exists as a dimer with the P domain of aneighbouring subunit, and the dimer clusters have been visualizedas projecting out from the shell of the capsid (Fig. 2.1, B). TheP domain contacts are nearly invariant; when virus particles expandthe P domain contacts remain unaltered and so continue to providestability to the capsid (Harrison, 1984). The P domain of TBSV hasa Class II "jellyroll" topology that is apparently unique among anyprotein structure so far examined (Gibson and Argos, 1990). It hasbeen speculated that P domain acquisition is a relatively recentevolutionary event and that an ancestral tombusvirus may havecaptured a cellular gene specifying this domain (Dolja and Koonin,1991; Gibson and Argos, 1990). The biological function(s) of theP domain have not been determined, however it could function bybinding to a virus receptor in a host cell or perhaps on the virusvector as suggested by Gibson and Argos (1990) and Riviere et al.(1989).The tombusvirus coat protein has two conformational statesthat are determined by the position of the hinge joining the P andS domains (Fig 2.1, A). Harrison et al. (1978) designate these twopositions, which differ by about 20°, as C/C or A/B, depending ifthe S contacts have true icosahedral (C/C), as opposed to local(A/B), two-fold symmetry (Fig. 2.1, A and C). Where the hinge isin the C/C position, a cleft exists between two S domains thatallows part of each of the amino-terminal arms to fold back alongthe edges of the S domains (Fig. 2.1, C). In the A/B position thetwo S domains abut together, the arms are disordered, and as such22are free to associate with the viral RNA (Fig 2.1, C).^Thismovement of the hinge provides the flexibility required forformation of the T=3 icosahedron. The amino-terminal armThe conformational state of the coat protein hinge has adirect influence on the conformation of the amino-terminal arm.Where the hinge is in a C/C configuration, the amino-terminal armsare ordered; three arms fold to form a small, globular clustercalled the 1-annulus at the three-fold axis of symmetry between Cshell domains (Harrison et al., 1987). Where the hinge is in theA/B configuration, the amino-terminal arms are disordered. Theyare not fixed to the main body of the subunit and are believed toassociate with the viral RNA. The R domain at the amino-terminusof the arm is rich in basic amino acids and it is believed thatthese amino acids, along with some in the shell, neutralize thepositively charged RNA, which is tightly packed within the capsid(Rossmann and Johnson, 1989).11.1.3 The movement of plant viruses within their hostsIt is generally accepted that plant virus movement within aplant can be characterized in two ways: as localized, or cell-to-cell movement; and as long distance or systemic movement throughthe vascular system (Maule, 1991).2311.1.3.1 Cell-to-cell movementPlant viruses are introduced into a plant by mechanical orvector-mediated inoculation of the epidermal, and sometimes,mesophyll cells (with the exception of those viruses that arephloem-limited). Once an infection locus has been established,progressive infection can only be achieved if the virus or itsmobile phase is able to move across the plant cell wall intoneighbouring cells. The cell-to-cell transport of plant virusesis an active process directed by the virus, but which involves bothhost and viral elements (Atabekov and Taliansky, 1990).Plasmodesmata, intercellular bridges that allow for continuoussymplastic connection throughout plant tissue, are generallyassumed to be the openings through which the virus, in the form ofa virion, naked RNA, or a virus-specific ribonucleoprotein(Atabekov and Dorokhov, 1984), moves from cell-to-cell.Plasmodesmata, however, are not sufficiently large enough to allowfor the movement of any of the three aforementioned entities, andso are thought to be modified by virus-encoded protein(s).Most, if not all, plant viruses are believed to encode amovement protein that functions to facilitate cell-to-cellmovement. The best studied of these is the 30K protein of tobaccomosaic tobamovirus (TMV). The 30K protein, which has beensuggested by in vitro studies to be a single-stranded nucleic acid-binding protein (Citovsky et al., 1990), is synthesized in theearly stages of TMV infection and has been found to be targeted tothe plasmodesmata. It binds cooperatively with the viral nucleic24acid, to form a long, thin, ribonucleoprotein complex that has beenproposed to be the form TMV RNA takes when moving from cell-to-cell(Citovsky et al., 1992).Citovsky and Zambryski (1991) proposed four steps in a modelfor TMV cell-to-cell movement: 1) the nucleic acid forms a complexwith the transport protein; 2) the transport complex isspecifically targeted to the plasmodesmata; 3) the movement proteininteracts with the plasmodesmata to increase its diameter; and 4)the viral RNA-protein complex is translocated through theplasmodesmatal channel into a neighbouring cell. The viral RNA mayremain associated with the movement protein and disassociate aftertranslocation, or it was also proposed that the movement proteinmay stay in the plasmodesmatal channel, which then remains in apermeable state, potentially allowing passage of largermacromolecules, including virions (Citovsky and Zambryski, 1991).An additional model has been proposed for the cell-to-cellmovement of members of the como-, nepo-, caulimo-, and geminivirusgroups (van Lent et al., 1990, 1991; Wieczorek and Sanfagon, 1993;Linstead et al., 1988; Kim and Lee, 1992). Cell-to-cell movementof these viruses has been associated with the formation of tubularstructures that appear in the plasmodesmata between neighbouringcells early in infection. The putative movement proteins of cowpeamosaic comovirus (CPMV) and tomato ringspot nepovirus (TomRSV) havebeen localized to the tubular structures (van Lent et al., 1991;Wieczorek and Sanfagon, 1993) and it appears that CPMV moves fromcell-to-cell through such tubules as intact virions (van Lent et25al., 1991). van Lent et al. (1991) demonstrated that the tubularstructures associated with cell-to-cell movement of CPMV wereproduced in CPMV-infected cowpea protoplasts, providing evidencethat the tubules were not a modification of the plasmodesmatadesmotubule, but that they arose de novo. It is not known whetherthe tubules consist strictly of virus-encoded protein or if hostfactors are also involved. Long distance movementLong distance, or systemic, movement of plant viruses isachieved when the virus enters the vascular system, which thentranslocates the virus throughout the plant. How the virus movesinto and out of the vascular tissue is not known. The viral coat protein and its role in virus movementThe role of the coat protein in virus movement differs withvirus group. Many plant viruses do not require coat protein orencapsidation to move efficiently from cell-to-cell (Maule, 1991),although, it has long been assumed that viruses must beencapsidated to move systemically (Matthews, 1991). Brome mosaicand cowpea chlorotic mottle bromoviruses, cowpea mosaic comovirus,cucumber mosaic cucumovirus, and turnip crinkle carmovirus allrequire their coat proteins for long distance spread (Sacher andAhlquist, 1989; Allison et al., 1990; Wellink and Van Kammen, 1989;Suzuki et al., 1991; Heaton et al., 1991). Mutants of TMV andmaize streak geminivirus that are defective for coat protein26production or assembly are able to move systemically, although atgreatly reduced efficiency (Dawson et al., 1988; Lazarowitz et al.,1989). There are now several plant viruses known that can movesystemically, under certain conditions and in certain hosts, in theabsence of a functional coat protein: barley stripe mosaichordeivirus, cymbidium ringspot tombusvirus, and red clovernecrotic mosaic dianthovirus (Petty and Jackson, 1990; Dalmay etal., 1992; Xiong et al., 1993) (see section 11.4 below). Theassumption for these latter viruses is that they move throughoutthe plant as a ribonucleoprotein complex. Other roles for the coat protein in the virus life cycleAs a means of maximizing their limited coding capacity, plantviruses often encode proteins that are multi-functional. The coatproteins of many plant viruses provide more than just protectionfor encapsidated RNA (as described in and are integralto the successful completion of the viral life cycle. Furtherexamples of the multi-functional nature of plant virus coatproteins are given below:1) Coat protein and virus replicationThe coat protein of alfalfa mosaic virus (A1MV) is requiredto initiate host infection by this virus (Bol et al., 1971). TheA1MV coat protein has been shown to regulate the balance betweenplus- and minus-strand RNA synthesis and it is believed that thecoat protein, which binds specifically to the 3' termini of the27four A1MV RNAs, is necessary for recognition of the positive senseRNAs by the replicase (Nassuth and Bol, 1983).2) Coat protein and symptomatologyMutations in the coat protein genes of TMV, TCV, A1MV, andcucumber mosaic virus affected symptom expression of these virusesin a variety of host plants (Dawson et al., 1988; Heaton et al.,1991; Neeleman et al., 1991; Shintaku and Palukaitis, 1990).Dawson et al. (1988) produced a series of TMV mutants withdeletions and/or insertions in the coat protein gene. The mutants,which were defective for virion assembly, produced distinctivesymptoms in tobacco, although there was no direct correlationbetween the sizes or positions of the coat protein gene mutationsand the phenotypic response elicited in transcript-inoculatedplants. Neeleman et al. (1991) found that a substitution of fouramino acids in the coat protein of A1MV strain YSMV was sufficientto change the typical symptoms of necrosis to chlorosis.3) Coat protein and host plant resistanceDawson et al. (1988) reported that mutations in the coatprotein gene of TMV were responsible for differential elicitationof resistance genes in tobacco. The induction of thehypersensitive response in plants carrying the n and N' genes, butnot the N gene, was a function of the TMV coat protein gene.284) Coat protein and vector transmissionChen and Francki (1990) reported that aphid transmission ofcucumoviruses by Myzus persicae was determined solely by the viruscoat protein. Nematode transmission of two nepoviruses wasdetermined to be a function of their coat proteins; the coatproteins of raspberry ringspot and tomato black ring nepoviruseswere found to control virus adsorption and release from theretention sites of their nematode vectors (Harrison and Murant,1977). Transmission of CNV by its fungal vector 0. radicale canbe eliminated if the virus is discharged into homologous antiserumbefore contact with fungal zoospores, implicating the coat proteinas the transmission determinant (Dias, 1970b) (see also section11.1.4, below).5. Coat protein and cross protectionCross protection occurs when the presence of one strain of aplant virus protects the plant against infection with a second.Several models have been proposed to explain this phenomenonincluding theories that invoke a role for the coat protein, perhapsby sequestering or encapsidating the RNA of the second virus, orby preventing its uncoating (Hamilton, 1980; Matthews, 1991). Coatprotein-mediated resistance to plant viruses in transgenic plantsexpressing either a gene encoding the coat protein, or thetranscriptional product of that gene, has been well documented(Powell Abel et al., 1986; Loesch-Fries et al., 1987; Hemenway etal., 1988).2911.1.4 Research objectiveThe objective of this portion of the study was to define thebiological function(s) of the CNV coat protein protruding domain.A CNV mutant, PD(-), which lacks the protruding domain codingsequence, was produced after in vitro mutagenesis of a CNV cDNAclone. The biological effects of deleting the protruding domaincoding sequence were assessed in a systemic host for CNV, Nicotiana clevelandii. PD(-) was able to replicate to a high level in N.clevelandii, and was able to move from cell-to-cell andsystemically within this host. The protruding domain mutant wasnot stable, and evidence is presented that the PD(-) coat proteincoding region was deleted in planta. The in planta deletion mutantwas cloned and synthetic RNA transcribed from this clone, CP(-),was able to systemically infect N. clevelandii, demonstratingconclusively that CNV does not require coat protein for systemicmovement. Finally, mechanical transmission of CNV can occur in theabsence of the CNV coat protein, although the efficiency of suchtransmission was reduced.3011.2 MATERIALS AND METHODS11.2.1 Basic molecular cloning techniquesProcedures commonly used in molecular biology and that wereutilized in this study, such as large scale and "mini-prep" plasmidDNA purifications, ligation of DNA fragments, transformation ofEscherichia coli, ethanol precipitation of DNA or RNA, gelelectrophoresis, and staining and photographing of gels, areessentially as described by Sambrook et al., (1989). Restrictionand DNA- or RNA-modifying enzymes were used according to themanufacturers' specifications for each enzyme unless otherwisestated.11.2.2 In vitro mutagenesis of the CNV coat protein geneIn vitro mutagenesis of the CNV coat protein gene, asdescribed in this section, is a modification of the mutagenesisprocedure of Kunkel et al. (1987), described in the InstructionManual accompanying the Muta-Genem Phagemid In Vitro MutagenesisKit (Bio-Rad). The Kunkel method of in vitro mutagenesisThe Kunkel method of in vitro mutagenesis takes advantage ofan E. coli strain, CJ236 (Kunkel et al., 1987), which is defectivefor dUTPase (dut) and uracil N-glycosylase (ung). As a consequenceof these mutations, DNA synthesized by CJ236 has numerous uracilbases in the place of thymine bases. The dut mutation inactivatesthe enzyme dUTPase, which results in high intracellular levels of31dUTP. The uracil remains incorporated in the DNA because the ungmutation, which inactivates uracil N-glycosylase, prevents theremoval of the uracil from the nascent DNA strand. The uracil-containing strand of DNA is purified from CJ236 and is used as thetemplate for complementary strand production. For in vitro mutagenesis, complementary strand synthesis is primed by anoligonucleotide that carries the desired mutations. The resultingdouble-stranded DNA is used to transform a chit + , ung+ strain of E.coli that produces active uracil N-glycosylase. The uracilcontaining strand (from CJ236) is consequently inactivated in thedut+, ung+ strain of E. coli, which allows for strong selectionagainst the non-mutagenized strand. Purification of single-stranded DNA from CJ236E. coli strain CJ236 (Bio-Rad) was transformed with 100 ng ofpK2/M5, a full-length, wild-type, CNV cDNA from which infectiousbacteriophage T7 RNA polymerase transcripts can be synthesized(Rochon and Johnston, 1991). The transformed colonies were platedonto Luria-Bertani medium (LB; 1% bacto-tryptone, 0.5% bacto-yeastextract, 1% NaCl, pH 7.5) 1% agar plates, amended with 50 µg/mlampicillin and 30 µg/ml chloramphenicol, and allowed to growovernight at 37°C. A single CJ236 colony was selected, used toinoculate 3 ml of LB containing 50 µg/ml ampicillin and 30 µg/mlchloramphenicol, and incubated with shaking overnight at 37°C.Fifty ml of 2xYT medium (1.6% bacto-tryptone, 1% bacto-yeastextract, 0.5% NaCl, pH 7.0) containing ampicillin and32chloramphenicol (as above) was inoculated with 1 ml of theovernight CJ236 culture, incubated with shaking at 37°C, andallowed to grow until an OD 600 of 0.3 was attained (corresponds toapproximately 1 x 10 7 colony-forming units/ml). Helper phageM13K07 (Viera and Messing, 1987; Bio-Rad) was added to obtain amultiplicity of infection of approximately 20 phages/cell. Shakingwas resumed and after 1 hour at 37°C kanamycin was added to a finalconcentration of 70 µg/ml, followed by further incubation for 4hours at 37°C. Thirty ml of the culture was centrifuged at 17,000x g for 15 minutes (min) at room temperature. The supernatant wascentrifuged again, as above but at 4°C. The supernatant wasremoved, mixed with 150 gg of RNase A (Pharmacia), and incubatedon ice for 30 min. A one-quarter volume of 3.5 M ammoniumacetate/20% PEG-6000 (BDH) was added to the supernatant which wasthen incubated on ice for a further 30 min. The phagemids werecollected by centrifugation at 17,000 x g for 15 min. The pelletwas resuspended in 200 gl of high salt buffer (100 mM Tris-HC1, pH8.0, 1 mM EDTA, 300 mM NaC1) and chilled on ice for 30 min. Thephagemid preparation was then centrifuged for 2 min to removeinsoluble materials. Phagemid DNA was extracted twice with phenol,once with phenol/chloroform/octanol (25:24:1), and several timeswith chloroform/octanol (24:1) until there was no visibleinterface. Each step was back-extracted with 100 gl of cold TEbuffer (10 mM Tris, 1 mM EDTA, pH 7.5). The aqueous phases werecombined and the DNA was precipitated in 2.5 volumes of absoluteethanol and 1/10 volume of 7.8 M ammonium acetate at -70°C for 3033min. The DNA was centrifuged for 20 min at 4°C, the pellet washedwith 70% ethanol, and resuspended in 20 gl of TE. The ssDNA wasstored at -20°C. Annealing of the mutagenic oligonucleotidesTwo mutagenic oligonucleotides (UBC Oligonucleotide SynthesisLaboratory), 5'GCCTCTTCTCGAGTCGTTGTTCC3' (CNV oligo #14) and5'CTTTGCGGCTCGAGCAATCAC3' (CNV oligo #13), each bearing twonucleotide changes from the published CNV sequence (the alterednucleotides are underlined) were purified, and phosphorylated (seesection, below). The mutations in these oligonucleotidesallowed for the introduction of two Xhol sites (CTCGAG) at theborders of the coding sequence for the coat protein P domain,starting at CNV nucleotides 3417 and 3733 for oligonucleotides 14and 13, respectively. The mutagenic oligonucleotides were annealedin separate reactions to the wild-type ssDNA purified from CJ236.Wild-type ssDNA (0.3 pmol) was mixed with 6-9 pmol of CNV oligo #13or CNV oligo #14 and annealing buffer (final concentration was 20mM Tris-HC1, pH 7.4, 2 mM MgC1 2 , 50 mM NaC1). The final reactionvolume was 10 gl. The annealing reaction mixtures were incubatedat 95°C for 1 min, 55°C for 5 min, 30°C for 5 min, and then cooledon ice. Complementary strand synthesisThe complementary DNA strand was synthesized by adjusting thecooled annealing reaction mixtures to 0.4 mM each dNTP, 0.75 mM34ATP, 17.5 mM Tris-HC1, pH 7.4, 3.75 mM MgC1 2 , 1.5 mM DTT and adding3 units of T4 DNA ligase (BRL) and 1.2 units T7 Sequenase Tm (USB)in a final reaction volume of 15 gl. The reactions were incubatedon ice for 5 min, at 25°C for 5 min, then at 37°C for 90 min.Reactions were stopped with 40 gl of TE (10 mM Tris-HC1, 1 mM EDTA,pH 7.5). After complementary strand synthesis, E. coli strain DH5a(BRL) was transformed with 10 gl of the reaction mixture. DH5acolonies were selected and inoculated to 3 ml of LB amended with50 µg/ml ampicillin and grown overnight at 37°C. Plasmid DNA wasextracted from a 1.5 ml culture by alkaline lysis. Plasmids werescreened for the desired mutation by Xhol digestion as describedin section Purification of synthetic oligonucleotidesPurification of synthetic oligonucleotides was according toa procedure developed at the UBC Oligonucleotide SynthesisLaboratory. A C 18 Sep-Pak Classic chromatography cartridge(Millipore) for solid phase extraction was prepared by firstpassing 10 ml of HPLC grade acetonitrile (BDH) through thecartridge, followed by 10 ml of sterile water. Crudeoligonucleotide, resuspended in 1.5 ml of freshly-prepared 0.5 Mammonium acetate, was loaded onto the Sep-Pak cartridge using a 3ml disposable syringe. The cartridge was washed with 10 ml ofsterile water, and then purged with air to displace all liquid.One ml of 40% acetonitrile was passed through the cartridge. Theeluant containing the oligonucleotide was collected, and this step35was repeated a further two times.^The concentration ofoligonucleotide in each of the three eluant fractions wasdetermined spectrophotometrically (1 A(1 260 unit of ssDNA = 33 µg/ml)and then each was evaporated to dryness in a Speed-Vac Concentrator(Savant). The oligonucleotide was resuspended in low-EDTA TEbuffer (10 mM Tris, 0.1 mM EDTA, pH 7.5) and was stored at -20°C. Phosphorylation of synthetic oligonucleotidesPhosphorylation of oligonucleotides was according to theInstruction Manual for the Muta-Gene Tm Phagemid In Vitro MutagenesisKit (Bio-Rad). Two hundred pmol of oligonucleotide was combinedwith 100 mM Tris, pH 8.0, 10 mM MgC1 2 , 5 mM DTT, 0.4 mM ATP, and4.5 units of T4 polynucleotide kinase (Pharmacia) in a finalreaction volume of 10 gl. The reaction was incubated at 37°C for45 min, and stopped by heating at 65°C for 10 min. Phosphorylatedoligonucleotides were diluted to a final concentration of 6 pmol/µlwith TE and were stored at -20°C.11.2.3 Construction of CNV cDNA coat protein clone mutantsTo determine if XhoI recognition sites were successfullyintroduced into the coat protein sequence of pK2/M5 byoligonucleotide-directed in vitro mutagenesis, the plasmid DNA wasscreened by digestion with XhoI (BRL). Two clones were selected.The first, pl4X (from CNV mutagenic oligonucleotide 14; Fig. 2.2),had a XhoI site at the 5' border of the P domain coding sequence.36B B 5'X Bs N B Bs= NAr'BstEll digestion,gel purification of BgIII-BstEll fragment from p14Xand BstEll-Ncol fragment from p13XB 5'X Bslir Bs3'X Nnt 1 ; ;^Xad 54 44701 nt 1^nt 4701 nt 1' B- ^+ 3% nt 4701NWILD-TYPE p14X p13XBgIII, Ncol digestion,gel purification of large fragment from wild-typeand small fragments from p14X and p13Xligation of large Ncol-Bgll fragment from wild-type,BgIII-BstEll fragment from p14X, and BstEll-Ncol fragment from p13X.TECC77-   Xp1<24.45Xhol digestiongel purifcation of large fragmentligationtransformation+ XPD(-)Fig.2.2^Schematic diagram indicating the sequence of eventsleading to the creation of the CNV P-domainless clone, PD(-).Figures within the box are of a larger scale than those outside ofthe box. Fragments from pl4X and pl3X have different hatching forease of recognition. Restriction endonuclease recognition sitesare abbreviated as follows: B=Bg/II, Bs= BstEII, N=NcoI, X= XhoI,5'X= XhoI site at the 5' end of the P domain (from pl4X), 3'X= 3'XhoI site at the 3' end of the P domain (from pl3X). Plasmid DNAis not shown.37The second, p13X (from CNVmutagenic oligonucleotide 13; Fig. 2.2),had a XhoI site at the 3' border of the P domain coding sequence.In order to produce a single clone carrying both XhoI sites, pl4Xand p13X were treated as follows (refer to Fig. 2.2). Both cloneswere digested with Bg/II (CNV nucleotide 3383; BRL) and NcoI (CNVnucleotide 3830; BRL). The resulting 447 by fragments wereseparated on a 1% TAE agarose gel, gel-purified using Gene Clean'(BIO 101), and subsequently digested with BstEII (CNV nucleotide3542). The Bg/II-BstEII fragment of p14X and the BstEII-NcoIfragment of p13X were ligated into the gel-purified, larger (7439bp) NcoI-BglII fragment of wild-type. The sequence betweennucleotides 3348 and 3843 of the resulting clone, XpK2/M5(Fig.2.2), was determined using the dideoxy-mediated chain-termination method of sequencing using T7 Sequenase tm (USB),essentially as described by the manufacturer. One spuriousnucleotide substitution (C to T that results in an amino acidchange of serine to leucine) at nucleotide 3500 within the P domaincoding sequence of XpK2/M5 was discovered. XpK2/M5 was thendigested with XhoI to excise the protruding domain coding sequence,and the large XhoI fragment was recircularized to produce clonePD(-) (Fig. 2.2).11.2.4 Construction of CNV cDNA clones for in vitro translationIn order to determine if the P domainless clone, PD(-), couldbe translated in vitro, it was necessary to construct a clone thatwould approximate the authentic CNV subgenomic mRNA that encodes38the coat protein. CNVtx2566/4166(+), a CNV cDNA clone thatcorresponds to CNV nucleotides 2566 - 4166 (prepared and describedby Johnston, 1989), contains the entire coat protein ORF, butattempts to translate synthetic RNA from this clone met withlimited success (Johnston, 1989). CNVtx2566/4166(+) contains twoAUG codons, AUG2574 and AUG2605 , upstream from the actual start codonfor the coat protein gene (AUG 2628 ). It is possible that one or bothof these two upstream codons could interfere with translation ofthe coat protein gene of synthetic RNA transcribed fromCNVtx2566/4166(+) as upstream AUG codons have been shown to reduceor eliminate protein synthesis from downstream coding regions(Kozak, 1984). In an attempt to enhance the translationalefficiency of RNA synthesized from CNVtx2566/4166(+), this cDNAclone was modified to remove AUG2574 and AUG2605 (see below).To delete AUG2574 , CNVtx2566/4166(+) (Fig. 2.3) was digestedwith BamHI (Bluescript multi-cloning site, upstream from the CNVinsert) and NcoI (CNV nucleotide 3830) to yield a 1272 by fragmentcontaining 69 and 61 by of extraneous sequence flanking the 5' and3' termini, respectively, of the CNV coat protein ORF. Thisfragment was treated with exonuclease Ba131 (0.25 U/gg DNA at 20°C,Boehringer Mannheim) for 3,5,7,9, and 11 min to make controlled 5'and 3' deletions. The reactions were stopped by adjusting themixture to a final concentration of 20 mM EGTA, pH 8.0. The endsof the Ba/31-treated DNA fragments were blunted using T4 DNApolymerase (5 U/gg DNA; BRL) at 37°C for 5 min and the reaction wasstopped with EDTA, pH 8.0 (final concentration 40 mM). The39reaction^mixtures^were^extracted^once^withphenol/chloroform/octanol (25:24:1) followed by extraction withchloroform/octanol (24:1), and then the DNA was precipitated at-70°C for 30 min by adding 0.1 volume 2M NaAc, pH 5.8 and twovolumes of absolute ethanol. DNA was resuspended in low-EDTA TEand a portion of the DNA was ligated with SmaI-digested Bluescript.Several plasmid DNA clones were sequenced in the regioncorresponding to the junction between the CNV coat protein gene 5'terminus and the Bluescript multicloning site, and one clone, pMM1(Fig. 2.3), which starts at CNV nucleotide 2591 was selected. pMM1was subsequently digested with Bailiff' (Bluescript multi-cloningsite) and BglII (CNV nucleotide 3383) and the 801 bp, gel-purifiedfragment was subcloned into the comparable region inCNVtx2566/4166(+) in order to reconstitute the 3' end of theoriginal clone. The resulting plasmid was named pMM2C (Fig. 2.3).pMM3A (Fig. 2.3) was prepared by replacing the BglII-NcoI fragmentof pMM2C with that from XpK2/M5. The XhoI-XhoI fragment from pMM2Cwas deleted to create pMM3PD(-) (Fig. 2.3), which lacks theprotruding domain coding sequence.A second series of clones, from which both AUG 2574 and AUG2605were removed, was created by Ba131 digestion of the BamHI-NcoIfragment of pMM2C in the same manner as described above. Afterdigestion with Ba131, the DNA was blunted with T4 DNA polymerase,and then blunt-end ligated into the Smal site of Bluescript. Aseries of clones was sequenced, and one, pMM6 (Fig. 2.3), whichstarts at CNV nucleotide 2612 (5 nucleotides downstream from the40nt 2566tint 2628 nt 4116 CNVtx2566/4116(+)BamHI (vector MCS)/Ncol digestiongel purificationnt 2566tint 2628^BBa131 exonudease digestionligation with Smal digested Bluescriptclones sequenced, pMM1 selectednt 2591nt 2628pMM1BamHI/Bg111 digestionnt 2591^+Bnt 26281ligation with BamHI/Bglil digested CNVtx2566/4116(+)nt 2591^ Bnt 2628 ti—nt 4116 pMM2CX N* f^-// nt ^pMM3AXhol digestionLigationnt 2591^ B X-  nt 2628 ++  I—/—nt 4116^pMM3PD(-)Fig. 2.3. Schematic diagram indicating the sequence of eventsleading to the creation of coat protein clones used for in vitro translation studies. Restriction endonuclease recognition sitesare abbreviated as follows: B= BamHI, E= EcoRI, H= HincII, K= KpnI,N= NcoI. The CNV coat protein ORF starts at nt. 2628. Plasmid DNAis indicated as a thick black line and is not drawn to scale.nt 2591nt 2628Bg111/Ncol digestionligation with BgIII-Ncol fragment of XpK2/M5rgif41coat protein subgenomic start site), was selected. The BamHI-Bg/IIfragment of CNVtx2566/4166(+) was replaced with the comparablefragment from pMM6 to produce pMM9 (Fig. 2.3). pMM10 (Fig. 2.3)carries the two XhoI sites of XpK2/M5, while pMM10PD(-) (Fig. 2.3),which was derived from pMM10 by XhoI digestion and religation,lacks the coding sequence for the coat protein P domain.11.2.5 In vitro transcriptionBacteriophage T7 RNA polymerase (BRL) run-off transcripts ofpK2/M5, XpK2/M5, PD(-) for plant inoculations, and of the clonesused for in vitro translation, were produced as follows. Fivemicrograms of Sinai linearized DNA was transcribed in 40 mM Tris,pH 7.9, 10 mM NaC1, 6 mM MgC1 2 , 2 mM spermidine, 10 mM DTT, 0.5 mMATP, CTP, GTP and UTP, 20 U RNAasin (Promega), and 100 U T7 RNApolymerase (BRL), in a 100 gl reaction, at 37°C for 45 min. Forplant inoculations, transcription reaction mixtures were adjustedto 10 mM sodium phosphate buffer, pH 7.0. RNA for in vitro translation reactions was purified with RNAaidTm (BIO 101) accordingto the manufacturer's protocol. The amount of synthetic RNAproduced was determined either spectrophotometrically after RNAaidTmpurification (1 mg/ml solution of RNA has an A26 0 value of 25) or bycomparison with CNV virion RNA after agarose gel electrophoresisand ethidium bromide staining. Transcripts were not capped.11.2.6 In vitro translationIn vitro translation of synthetic RNA transcribed from clones42pMM2C, pMM3A, pMM3PD(-), pMM9, pMM10 and pMM10PD(-) was carried outusing a wheat germ cell- free extract (Promega) programmed with 120gg/ml RNA. Translation reactions were carried out at 20°C for 60minutes in the presence of 35S-Met (specific activity approximately1000 Ci/mmol; New England Nuclear). Each 25 gl translation reactionmixture, in addition to RNA and 35S-Met, included 12.5 gl of wheatgerm extract, 80 gm amino acid mixture (minus methionine), and 75mM potassium acetate. Translation products were analyzed by SDS-PAGE using the discontinuous buffer system of Laemmli (1970).Translation products were electrophoresed through a 15% separatinggel (0.75 mm) using the manufacturer's recommended conditions forthe Mini-Protean II electrophoresis apparatus (Bio-Rad). Afterelectrophoresis, gels were fixed in three changes of 30%methanol/10 % acetic acid (with agitation for 30 minutes for eachchange) and fluorographed using Entensify (New England Nuclear)according to the manufacturer's instructions. Gels were dried for1 hour at 80°C on Whatmann 3MM filter paper under vacuum, andexposed to X-ray film (X-Omat K, Kodak) at -70°C. The sizes oftranslation products were estimated by comparison with thepublished sizes of the in vitro translation products of CNV(Johnston and Rochon, 1990) and brome mosaic virus (BMV) (Ahlquistet al., 1981).11.2.7 Inoculation of Nicotiana clevelandii Approximately 5 gg of in vitro transcribed RNA, in 100 gl oftranscription salts/10 mM sodium phosphate buffer, pH 7.0, was used43to rub-inoculate two leaves on each of two, four- to six- week oldCarborundum-dusted Nicotiana clevelandii Gray. For serial passageand host range determination of wild-type and the mutant viruses,1 cm2 discs were cut from systemically-infected leaves with asterile cork borer, ground in 1 ml of 10 mM potassium phosphatebuffer, pH 7.0, and used to rub-inoculate healthy N. clevelandii.Buffer-inoculated plants served as controls. All plants weremaintained in a greenhouse under a day/night temperature regime of22°C/17°C with a 14 hour photoperiod. Plants were monitored every1-2 days for symptom development.Sap from wild-type-infected plants was not used in either thehost range study and or in the ELISA experiments (section 11.2.9).Because the XhoI mutations in XpK2/M5 appeared to beinconsequential, XpK2/M5 was initially used as the controltreatment to which PD(-) was compared. In subsequent experiments,however, wild-type was added, or used to replace XpK2/M5, toprovide a more stringent control for comparative purposes.11.2.8 Local lesion assay in Chenopodium quinoa The following local lesion assay was used to determine theefficiency of sap transmission and also to assess if PD(-) viralRNA was protected by some means from degradative enzymes ininfected plants. Systemically infected leaves from second passage,PD(-)- and wild-type-infected N. clevelandii, and leaves frombuffer-inoculated N. clevelandii, were ground in 2 [PD(-) andhealthy] or 200 (wild-type) volumes (w/v) of 10 mM sodium phosphate44buffer, pH 7.0. The sap from wild-type-infected plants was diluted100-fold more than that from PD(-)-infected plants in order toinduce a local lesion response in C. quinoa that permitted countingof the lesions; more concentrated wild-type inoculum producedlesions too numerous to count. To determine stability of theinfectivity of viral RNAs in sap, 500 gl of sap from buffer-inoculated leaves was amended with CNV RNA (6.7 µg/ml sap) and allsamples were left on the bench at 30° C. After 1, 10, 30 and 60minutes, a 120 gl aliquot of each preparation was removed and usedto inoculate two leaves on each of two C. auinoa plants. Locallesions were counted 3 dpi.11.2.9 ELISAApproximately 0.1 g of inoculated or systemically infectedleaf tissue from transcript- and sap-inoculated plants was groundin 1 ml of 20 mM sodium phosphate buffer, pH 7.0/ 150 mM NaCl(PBS). Tissue from buffer-inoculated plants was used as a negativecontrol. A rabbit polyclonal antiserum to CNV, prepared usingintact virus particles as immunogen (Johnston, 1989), was used todetermine the presence of CNV capsid antigen in leaf sap by doubleantibody sandwich (DAS) ELISA (Clark and Adams, 1977). BackgroundA4 05 values were determined by taking the mean of A4 05 readings fromthe buffer-inoculated (negative) controls. A positive A405 valuewas conservatively set at ten times greater than background (Sutulaet al., 1986).4511.2.10 Immunoelectron microscopyTo determine if intact virions were present in inoculated andsystemically infected leaves of transcript-inoculated plants, gold-labelling experiments were conducted. Collodion-treated, carbon-coated, copper grids (400 mesh) were placed consecutively on 10 illdrops of 10 µg/ml Protein A for 10 min, CNV polyclonal antiserum(undiluted) for 10 min, leaf sap for 30 min, and CNV antiserum(1/100 dilution) pre-complexed with 5 nm gold for 30 min. Gridswere washed between each step with PBS. Grids received a finalwash with bacitracin (Calbiochem), were stained with 2%phosphotungstic acid, pH 6.8 and observed using a Hitachi 600transmission electron microscope.11.2.11 Leaf RNA extractionApproximately 0.1-0.5 g of leaf material was ground to a finepowder in liquid nitrogen, mixed vigorously with 400 glphenol/chloroform/octanol (25:24:1), 400 1.11 10X THE (100 mM Tris-HC1, pH 7.5, 100 mM NaC1, 10 mM EDTA), containing 0.1 % SDS and 5%2- mercaptoethanol, and centrifuged in an Eppendorf microcentrifugeat 14,000 rpm for two min. The aqueous phase was collected, re-extracted with an equal volume of phenol/chloroform/octanol, andthen extracted with chloroform/octanol (24:1). The aqueous phasewas combined with 0.1 volumes of 2 M NaAc, pH 5.8 and 2.5 volumesof absolute ethanol, the nucleic acids were precipitated at - 70°Cfor 10 min, and then pelleted by centrifugation for 10 min at 4°C.The pellet was resuspended in 300 gl TE, to which 100 gl of 8M LiC146was added to precipitate ssRNA. Samples were left on ice for 1-3hours, then centrifuged for 10 min at 4°C. The supernatant was re-precipitated with 2.5 volumes of absolute ethanol. The LiC1 pelletwas resuspended in 100 gl of water, and nucleic acids were re-precipitated in 0.1 volumes of 2M NaAc, pH 5.8, and 2.5 volumes ofabsolute ethanol. Samples were placed at -70°C for 0.5-1 hour,then centrifuged for 10 min at 4°C. Pellets were washed once with70% ethanol. The precipitated nucleic acids in the pellet from the2M LiC1 supernatant was resuspended in 50 gl of sterile, deionizedwater. The 2M LiC1 pellet of insoluble nucleic acids wasresuspended in 25 gl of sterile, deionized water. Nucleic acidconcentration was determined spectrophotometrically (1 A2 6 0 unit =40 gg RNA/ml).11.2.12 Northern blots11.2.12.1 Denaturation and transfer of RNATwo molar lithium chloride-insoluble, high molecular weightssRNA was denatured with either 5 mM methylmercuric hydroxide(Bailey and Davidson, 1976) or with glyoxal and dimethyl sulfoxide(Sambrook et al., 1989) and electrophoresed through 1% agarose gelsprepared with BE buffer (methyl mercury gels; 40 mM boric acid/ 1mM EDTA, pH 8.2) or 0.01 M sodium phosphate buffer, pH 7.0 (glyoxalgels). Methylmercuric hydroxide is highly toxic and volatile;consequently, all manipulations with this compound were carried outin a chemical fume hood. Gloves were worn at all times. Anymaterials coming in contact with methylmercuric hydroxide were47washed or soaked in a solution of 10 mM 2-mercaptoethanol prior todisposal. After electrophoresis, RNA was blotted onto Zeta-probemembranes (Bio-Rad) under alkaline conditions (10 mM NaOH) at roomtemperature for 3 - 16 hours (Reed and Mann, 1985). The membraneswere prehybridized in hybridization buffer (50% formamide, 0.12 MNa2HPO4 , pH 7.2) for 0.5 to 2 hours at 42°C. Hybridization with 32P-labelled, nick-translated (2 x 10 8 cpm/gg DNA; section random-primed probes (1 x 10 9 cpm/gg DNA; section wascarried out in a Hybaid Oven (Bio-Can Scientific) for 16-24 hr, at42°C. After hybridization, the membranes were rinsed briefly in2X SSC/0.1% SDS, and then washed successively in 2X SSC/0.1% SDS,0.5X SSC/0.1% SDS, and finally in 0.2X SSC. All washes were at60°C for 15 min each. Excess moisture was removed from themembranes, which were then wrapped in clear, plastic wrap. Themembranes were exposed to X-ray film (X-Omat K, Kodak), at -70°Cwith the aid of two Lightening Plus intensifying screens (DuPont). Nick-translated probesNick-translated 32P-labelled probes were prepared essentiallyas described by Rigby et al. (1977). Reactions were carried outin a 50 gl volume and contained: 50 mM Tris-HC1, pH 7.5, 10 mMMgC1 2 ; 10 mM DTT, 50 µg/ml nuclease-free BSA, 0.6 mM each dCTP,dGTP, dTTP; 30-50 gCi 00 2P-dATP (ICN, ca. 3000 Ci/mmole), 250 pgDNaseI (Sigma or Promega), 10 U E. coli DNA polymerase I holoenzyme(Pharmacia), and 50 - 100 ng of SmaI linearized cloned wild-typeCNV DNA. Reactions were incubated at room-temperature until a48specific activity of 2 x 10 8 cpm/ug DNA was attained. Reactionswere stopped using 1 gl of 20% SDS. Unincorporated nucleotideswere removed by passing the probe through a miniature Sephadex G-50(Sigma) column (Sambrook et al., 1989). The DNA probe in thecolumn eluant was denatured by adding NaOH to a final concentrationof 100 mM, followed by boiling for 5 min. The denatured probe wasquick-cooled on ice, then added directly to the hybridizationsolution and the prehybridized membrane. Random-primed probesRandom-primed 32P-labelled probes were prepared with the RandomPrimers DNA Labelling System (BRL) according to the manufacturer'sinstructions. Twelve nanograms of linearized plasmid DNA, whichcorresponded to the 3' terminus of the CNV coat protein ORF and theORFs for the 20/21 kDa proteins (CNV nt 3383 - 4447), was used astemplate. Specific activity of probes was approximately 1 x 10 9cpm/gg DNA.11.2.13 Cloning of RNA extracted from PD(-)-infected plants11.2.13.1 Reverse transcription and polymerase chain reactionsTo clone the ca. 3800 nt. RNA species from PD(-)-infectedplants, high molecular weight ssRNA from systemically-infectedleaves collected during the third and fourth passage in N.clevelandii was reverse transcribed and amplified by the polymerasechain reaction (RT-PCR), essentially as described by Sambrook etal. (1989). RNA (350 ng) was reverse transcribed in a 20 gl49reaction volume in the presence of Tag DNA polymerase buffer [50mM KC1, 10 mM Tris-HC1, pH 9.0, 0.01% gelatin (w/v), 0.1% TritonX-100; Promega], 1 mM dATP, dCTP, dGTP, dTTP, 50 pmol of 3' primer,20 Units RNAasin (Promega), 8 mM MgC1 2 , and 200 units of M-MLVreverse transcriptase (BRL). The 3' primer, RNA template, and waterwere mixed first and heated at 90°C for 2 min, then put on ice.The remaining reaction components were added, and the reactionmixture was then incubated for a further 30 min at 37°C. The M-MLVreverse transcriptase was inactivated by heating the reactionmixture at 95°C for 5 min. To amplify the DNA generated by reversetranscription, the transcription reaction mixtures were amendedwith 50 pmol each of the 5' and 3' primers, 2 units of Tag DNApolymerase (Promega) adjusted to 1.5 mM MgC1 2 and 1X Tag buffer(see above) to a final volume of 79 gl. The reaction mixtures wereoverlaid with 100 gl of autoclaved, light mineral oil. Theamplification reactions were performed in a thermo-cycler (Eri-Comp). For the first cycle, denaturation was at 94°C for 5 min,annealing was at 50°C for 2 min, and polymerization was at 72°C for3 min. This was followed with 30 cycles of denaturation at 94°Cfor 1 min, annealing at 50°C for 2 min, and polymerization at 72°Cfor 3 min. For the final cycle, the polymerization step wasextended to 10 min at 72°C. The mineral oil overlay was removedfrom samples by addition of 150 gl of chloroform/octanol (24:1)followed by brief centrifugation. The aqueous phase, containingthe amplified DNA, formed a micelle between the oil and thechloroform, and was removed to a new microcentrifuge tube.5011.2.13.2 Cloning of RT-PCR productsThe primers used for PCR, selected to amplify sequencesadjacent to and within the CNV coat protein open reading frame,were located approximately 181 nt upstream and approximately 133nt downstream from the start and stop codons of the CP ORF,respectively. The 5' primer, 5' ACGTGAATTCGTGACCCCTGAGGCAA 3', has10 non-CNV nucleotides (underlined), including an EcoRI site(GAATCC), as well as the unique CNV Bsu36I site (italics). The 3'primer, 5' GGGAGTAATGGTACCTCC 3', has a CNV KpnI site (underlined).The amplified DNA was digested with EcoRI and KpnI, and thenligated into EcoRI/KpnI-digested Bluescript. DNAs from four clonesfrom each of two independent RT-PCR experiments was sequenced (seesection The DNA from one clone was digested with Bsu361and NcoI and the 310 by fragment was ligated into the gel-purified,large fragment of similarly digested wild-type DNA. This clone,named CP(-), was linearized with SmaI, transcribed in vitro withT7 RNA polymerase and the run-off transcripts were used toinoculate N. clevelandii as described in sections 11.2.5 and11.2.7, above. Plants were maintained in a growth room with a meantemperature of 23°C (fluctuated between 21-25°C) and 16 hourphotoperiod. Total leaf RNA was extracted from the inoculated andsystemically infected leaves of CP(-) transcript-inoculated N.clevelandii, denatured, and analysed by Northern blotting (asdescribed above).5111.3 RESULTS11.3.1 Symptomatology and infectivity of PD(-) in N. clevelandii To examine the possible role that the CNV coat protein playsin various aspects of the CNV infection cycle, a systemic host ofCNV, N. clevelandii, was inoculated with wild-type synthetic CNVtranscripts, and transcripts from the two CNV coat protein mutants,XpK2/M5, and PD(-). The XpK2/M5 mutant contained two introduced,in-frame, XhoI restriction endonuclease sites in the CNV sequenceencoding the coat protein P domain (Fig. 2.4). The PD(-) mutantwas derived from XpK2/M5 by deletion of the XhoI fragmentencompassing most of the P domain coding sequence (Fig. 2.4). Eachinoculated plant developed necrotic local lesions 5 to 7 dpi. Thelesions on wild-type- and XpK2/M5-infected leaves were, at alltimes (9 and 12 replications, respectively), indistinguishable fromeach other (Fig. 2.5). At 8 dpi the local lesions were a buff-tancolour, 2-3 mm in diameter and were of an irregular shape. Bycontrast, the local lesions on PD(-)-inoculated leaves (16replicates) at 8 dpi were significantly smaller, not exceeding adiameter of 1 mm (Fig. 2.5). Systemic leaf necrosis was observedin wild-type- and XpK2/M5-infected plants approximately 8 - 12 dpiwhile systemic infection in PD(-)-infected plants was delayed afurther 8 - 12 days (Fig. 2.6). In transcript-inoculated plants,coat protein mutations in XpK2/M5 were apparently inconsequential,whereas the deletion of the coat protein P domain from PD(-)affected both the appearance of symptoms and virus movement.Sap from systemically infected leaves of wild-type-,522.1 kbcoat protein subgenomic RNAp20/21 sub enomic RNA cokbnt. 785 p20CNV GENOMIC RNA4.7 kbp33^p92putative polymerisep41coat proteinp21^pXh, 5aant. 2607ambercodonR a -^S 1 P58 as 34aa 167 as 116 aah1 .1►1■■wild-type TSPLLESLFREXpK2/M5 TSPLLESLFREPD(-)^TSPLP^/ / ^ LFAVRAITENAVQVVLFAARAITENAVQVV//4EQSLKMRCRLCKLGASIF . g. 2.4. Summary of the organization of the CNV genome and the location of coat protein mutations. (A) Thes i zes, locations and putative functions of CNV encoded proteins. The positions and sizes of CNV subgenomicAs are indicated above the genomic map. (B) The location and size in amino acids (aa) of the random (R),a (a), shell (S), hinge (h) and protruding (P) domains of the CNV coat protein. (C) Protruding domain borderse• ences of wild-type and XpK2/M5, and the amino acid sequence of PD(-) after deletion of the P domain. TheXp 2/M5 amino acids in bold type were those affected by mutagenesis. The boxed amino acids of PD(-) are dueto a frameshift mutation in the CNV coat protein ORF.Fig. 2.5. Inoculated leaves of N. clevelandii 6 days after mockinoculation with 0,01 M sodium phosphate buffer, pH 7.0 (A), orinoculation with in vitro transcribed RNA from wild-type (B),XpK2/M5 (C), or PD(-) (D) cDNA clones.54BUFFER CONTROL pK2/M5Fig. 2.6. Systemically infected N. clevelandii 22 days after mockinoculation with 0.01 M sodium phosphate buffer, pH 7.0 (A), orinoculation with in vitro transcribed RNA from wild-type (B) orPD(-) (C) cDNA clones.55XpK2/M5-, and PD(-)-inoculated plants was used to mechanicallyinoculate N. clevelandii. In independent transcript-inoculations,significant differences in symptoms between the different inoculawere no longer apparent on the inoculated leaves; all local lesionswere of the same appearance. All three inocula induced systemicsymptoms in 7-12 days, although, on closer inspection, there wasa trend for PD(-)-infected plants to induce systemic symptomsslightly later (1-3 days) than wild-type or XpK2/M5; almost two-thirds of the wild-type-inoculated plants exhibited systemicsymptoms before plants inoculated with PD(-). Approximately onein five systemically-infected, PD(-)-inoculated plants exhibitedan "oak leaf" pattern of necrosis that followed along the primaryand secondary veins before development of necrosis in the mesophylltissue (see plant C, Fig. 2.6). The necrosis induced in wild-type-and XpK2/M5-infected plants occurred systemically throughout theentire leaf, apparently affecting the veinal and mesophyll tissuesat the same time, and was usually lethal. Necrosis induced by sap-transmitted PD(-), whether systemic or oak-leaf, was also usuallylethal to N. clevelandii.11.3.2 Host range studyTo determine if PD(-) induced typical CNV-like symptoms onplants other than N. clevelandii, sap from systemically infectedleaves of either PD(-) or XpK2/M5 transcript-inoculated N.clevelandii was used to inoculate N. benthamiana Domin., Cucumis sativus L., Daturia stramonium L., Gomphrena globosa L.,56Chenopodium amaranticolour Coste & Reyn, and C. quinoa Willd. N.benthamiana and C. sativus are reported to be systemic hosts ofCNV, although in our hands C. sativus supports systemic infectiononly rarely. The other four plant species listed above are locallesion hosts for CNV. In N. benthamiana, grey, sunken locallesions were produced by both PD(-) and XpK2/M5. These appearedapproximately four dpi for both inocula and they eventuallycoalesced, causing leaf collapse. Systemic infection of N.benthamiana was established by PD(-) and XpK2/M5 eight dpi and waslethal. On all other hosts, including C. sativus, PD(-) inducedthe production of local lesions that were indistinguishable fromthose produced by XpK2/M5.11.3.3 Immunoeiectron microscopyCrude leaf sap preparations of the inoculated and systemicallyinfected leaves of transcript-inoculated plants were examined byimmunosorbent electron microscopy for the presence of CNV coatprotein and/or virions, using polyclonal antisera raised againstCNV virions. Gold-labelled virus particles were observed in allsap preparations from wild-type- and XpK2/M5- infected plants, butwere never observed in sap from PD(-)-infected plants. The Xholmutations in XpK2/M5 did not appear to affect coat proteinsynthesis or virion assembly. That fact that no coat protein orvirions were detected in PD(-) leaf sap preparations indicates: 1)that the truncated coat protein gene of PD(-) may not have beentranslated to produce P-domainless coat protein; 2) if coat protein57was produced, lack of the P domain may have compromised particleassembly and/or stability; and 3) if P-domainless virus particleswere produced they may not have been detectable by ISEM using CNVantiserum raised against intact virions. The coat protein P domainis considered to be highly antigenic and consequently a CNVpolyclonal antiserum raised against intact virions may not have hadsufficient specificity to label P-domainless virus particles, hadthey been produced.11.3.4 ELISAEvidence for lack of coat protein synthesis in PD(-)- infectedplants was further supported by double antibody sandwich ELISAanalysis. Coat protein was detected in leaf sap preparations fromtranscript- and sap-inoculated XpK2/M5-infected plants, while nocoat protein could be detected in samples from any of the PD(-)-infected plants assayed (Table 2.1). The ELISA results for thetranscript-inoculated, first and second passage plants indicatethat, as with the ISEM experiments (see section 11.3.3), the coatprotein gene of PD(-) may not have been translated in vivo or that,if translated, may not have been detected due to the nature of thepolyclonal antiserum used.11.3.5 In vitro translationGiven that neither coat protein nor virus particles weredetected in PD(-)-infected leaves by ISEM or ELISA, it wasimportant to determine whether or not the PD(-) coat protein gene58could, in fact, be translated in the absence of the P domain codingsequence. In vitro translation studies of CNV coat proteinsubgenomic RNA were previously described by Johnston and Rochon(1990). A CNV cDNA, CNVtx2566/4166(+), encompassing the ORF forthe coat protein, was successfully translated, but the level ofcoat protein was low. The clones used for the in vitro translationstudy described below were derivatives of CNVtx2566/4166(+), andwere produced in an attempt to enhance the translational efficiencyof in vitro transcribed coat protein subgenomic mRNA.Transcripts of clone CNVtx2566/4166(+) have two AUG codons(CNV nucleotides 2574 and 2605) lying upstream of the authenticsubgenomic start site (CNV nucleotide 2607) that could potentiallyinterfere with efficient translation of the subgenomic messengertranscribed from this clone. In an attempt to enhance translationof this synthetic coat protein subgenomic mRNA, AUG 2574 and AUG2605were sequentially deleted from CNVtx2566/4116(+) cDNA clones usingBa131 exonuclease and the resulting clones, pMM2C from which AUG 2574was removed, and pMM9, lacking both upstream AUG codons, weretranslated in vitro. Deletion of the first of these two AUG codonsdid not appear to greatly increase translational efficiency(compare lanes 3 and 4, Fig. 2.7), while removal of the second AUGsignificantly enhanced coat protein production (lane 5, Fig. 2.7).To determine if a putative PD(-) coat protein subgenomic RNAproduced in PD(-)-infected plants could direct synthesis of P-domainless coat protein, synthetic RNA transcribed from the CNV59Table 2.1. DAS-ELISA A4 05 values' for leaf sap extracted from XpK2/M5-, PD(-)-, and buffer-inoculated N. clevelandii using a polyclonal antiserum raised against CNV virions.Plant e transcript inoc. first passage second passageInocula inoc. 3 syst.4 inoc. syst. inoc. syst.XpK2/M5 ND6 1.978 ND 1.850 1.664 1.672PD(-) 0.047 0.032 ND 0.033 0.046 0.014 mock s ND 0.033 ND 0.013 0.011 0.009The A4 05 value for background was 0.015. Any value greater than ten times background (mock)as considered a positive result.Refers to the transcript-inoculated, or sap-inoculated first or second passage N.levelandii from which the leaf sap was extracted.Sap was taken from an inoculated leaf.4 Sap was taken from a systemically infected leaf.5 Leaf sap was taken from a plant inoculated with 0.01 M sodium phosphate buffer, pH 7.0.ND = not determinedsubgenomic^coat proteinstart site start codonnt 2607 nt 2628CNVtx2566/4116(+) S'GACGCGCalQAGCCCAGCATCCTTGACTCCGCCGTAGCAMACCAAGCAAACACAAACCAC 3'pMM2C^ S'GACTCCGCCGTAGCAMACCAAGCAAACACAAACACA8M0.3.pMM9 S'AGCAAACACAAACACAAMPFig. 2.7. In vitro translation products of synthetic CNV coatprotein subgenomic RNA in wheat germ extracts. A. Sequences of the5' terminal regions of the cDNA clones CNVtx2566/4116(+), pMM2C,and pMM9 which were transcribed to produce the synthetic RNA usedfor in vitro translation. B. The 35S-labelled translation productswere analyzed by electrophoresis through a 15% SDS-polyacrylamidegel followed by fluorography. In vitro translation reactionvolumes were 25 gl of which 6 gl were loaded in each well (2 gl forBMV). Lane 1, no added RNA; lane 2, CNV virion RNA (6.0 gg); lane3, CNVtx2566/4116(+) RNA; lane 4, pMM2C; lane 5, pMM9. 3.0 gg ofRNA was loaded in lanes 3-5. The numbers on the left indicate thesizes in kDa of CNV in vitro translation products. The 41 kDaprotein is the coat protein.61cDNA clone pMM10PD(-), was tested for its ability to producetruncated CNV coat protein in vitro in wheat germ extracts.pMM10PD(-) corresponds to a CNV coat protein subgenomic mRNAlacking the coding sequence for the coat protein P domain. Fig.2.8, lane 2, shows that pMM10PD(-) RNA was successfully translatedto produce a protruding-domainless coat protein of the predictedmolecular weight of 29 kDa.^The 41 kDa translation productsproduced from pMM9 and pMM10 synthetic transcripts (correspondingto the coat protein subgenomic RNA from wild-type and XpK2/M5,respectively) are also shown in Fig. 2.8 (lanes 3 and 4).Translation of pMM10 to produce a 41 kDa protein demonstrates thatthe introduction of the two XhoI sites in XpK2/M5 (see Fig. 2.1)did not affect coat protein production in vitro. A 29 kDa proteinwas also observed when translation of pMM10PD(-) was carried outusing rabbit reticulocyte lysates (data not shown).11.3.6 Analysis of viral RNA species in infected leaf extractsIn order to compare the replication products from wild-type,XpK2/M5, and PD(-) infected plants, high molecular weight, single-stranded RNA extracted from the inoculated and systemicallyinfected leaves of transcript- and sap-inoculated plants wasanalyzed by Northern blotting using a 32P-labelled random-primedprobe. The plasmid DNA used for the random-primed probe was arestriction fragment specific to the p21 and p20 ORFs, and alsocontained the 3' portion of the CNV coat protein ORF. Fig. 2.9shows that all transcripts replicated well in N. clevelandii,62Bnt 2612VII^R^I aln12612R^I aln1 2612R^I al ^41161 2^3^4^5^6ApMM10PD(-)II^P^1----11—nt 4116 pMM9X^X1r^P V 1.---„--nt 4116 pMM10X 41kDa-H■33 kDa21 kDa20 kDaFig. 2.8. The in vitro translation product of a synthetic PD(-)coat protein subgenomic RNA. A. Diagram showing the cDNA clonespMM9, pMM10, and pMM10PD(-) which were transcribed to produce thesynthetic RNA used for in vitro translation. B. 35S-labelled, invitro translation products were analyzed by electrophoresis througha 15% SDS-polyacrylamide gel followed by fluorography. In vitro translation reaction volumes were 25 gl of which 6 gl were loadedin each well (2 gl for BMV). Lane 1 shows the in vitro translationproducts of CNV virion RNA (6.0 gg ); lane 2, pMM10PD(-) RNA (3.0gg); lane 3, p112410 RNA (3.0 gg); lane 4, pMM9 RNA (3.0 gg); lane5, BMV RNA (0.5 gg); lane 6, no added RNA. The numbers on the leftindicate the sizes in kDa of CNV in vitro translation products.The 41 kDa protein is the coat protein (CP).631 2 3 4 5 6 7 8 9 10 11 12 13 144.4 kb^4W 4-4.7 kb4-2.1 kb1- 0.9 kb Fig. 2.9. Northern blot analysis of high molecular weight RNAextracted from leaves of PD(-)-, XpK2M5-, and wild-type transcript-and sap-inoculated plants. RNA samples were denatured withglyoxal/DMSO, electrophoresed through a 1% agarose gel, blotted andhybridized with a 32P-labelled random-primed CNV DNA probecorresponding to the 3' terminal region of the coat protein ORF andthe p20/21 ORFs. Lane 1 contained 50 ng of PD(-) in vitro transcribed RNA (4.4 kb). Lanes 2-14 each contained 1 gg of 2Mlithium chloride insoluble RNA extracted from: PD(-) transcriptinoculated leaf (lane 2); PD(-) systemically infected leaf from atranscript inoculated plant (lane 3); PD(-) systemically infectedleaf after first passage (lane 4); PD(-) systemically infected leafafter fourth passage (lane 5); XpK2/M5 transcript inoculated leaf(lane 6); XpK2/M5 systemically infected leaf from a transcriptinoculated plant (lane 7); XpK2/M5 systemically infected leaf afterfirst passage (lane 8); XpK2/M5 systemically infected leaf afterfourth passage (lane 9); wild-type transcript inoculated leaf (lane10); wild-type systemically infected leaf from a transcriptinoculated plant (lane 11); wild-type systemically infected leafafter first passage (lane 12); wild-type systemically infected leafafter fourth passage (lane 13); buffer inoculated leaf (lane 14).The numbers on the right correspond to the sizes in kb of genomiclength CNV RNA (4.7) and the subgenomic RNAs for the coat protein(2.1) and the p21/20 kDa proteins (0.9) (molecular size markers notshown).64however, replication of the complete PD(-) RNA (4.4 kb) was rarelydetected in transcript-inoculated leaves.In most inoculated leaves as well as in systemically infectedleaves of transcript-, and subsequently sap-inoculated plants, RNAcorresponding to full-length PD(-) (ca. 4385 nt; see Fig. 2.9, lane1) could no longer be detected. Instead, an RNA species of about3800 nt. (Fig. 2.9, lanes 2-5) became dominant and was found to bestable and to replicate to very high levels through five passagesin N. clevelandii [see passages 1 and 4 in lanes 4 and 5,respectively, in Fig. 2.9].In Fig. 2.9 (lanes 6-13) two anomalous RNA species are shownthat migrated slightly faster than the coat protein subgenomic RNA(2.1 kb). The identity of these RNAs is presently unknown;however, RNA species of this size have been noted previously in CNVvirion RNA preparations (Johnston and Rochon, 1990). Additionally,the 2.1 Kb coat protein subgenomic RNA was detected in RNA extractsfrom wild-type- (Fig. 2.9, lanes 10-13) and XpK2/M5-infected plants(Fig. 2.9, lanes 6-9) but an RNA species of this size was notdetected in PD(-)-infected leaves (Fig. 2.9, lanes 2-5).An RNA of approximately 1000 nucleotides was detected inPD(-)-infected plant extracts (Fig. 2.9, lanes 3-5), but not inwild-type or XpK2/M5 infected leaf material (Fig. 2.9, lanes 6-13).While appearing as a large diffuse band of 900-1000 nt in Fig. 2.9,two bands were distinguishable on the original autoradiograph. Thelower band is the previously identified 0.9 kb, p21/20 subgenomicRNA (Johnston and Rochon, 1990) and, as expected, is present in all65PD(-), XpK2/M5, and wild-type lanes (Fig. 2.9, lanes 2-13). Theupper band, specific to lanes 3-5, may be a coat protein subgenomicRNA from which most or all of the coat protein coding sequence hasbeen deleted (predicted size 1.1 kb).11.3.7 Local lesion assay of PD(-) on C. quinoa CNV coat protein could not be detected by ISEM or DAS-ELISA(sections 11.3.3, 11.3.4) nor could a PD(-) coat protein subgenomicRNA (predicted size, 1.8 kb) be detected by Northern blot analysisof ssRNA from PD(-)-infected plant material (Fig. 2.9, lanes 2-5).It was therefore surprising that each attempt to sap-transmit PD(-)was successful. A local lesion assay was used to determine theefficiency of PD(-) sap transmission and also to assess if PD(-)viral RNA was protected by some means from degradative enzymes ininfected plants. Sap from second passage-, wild-type- or PD(-)-infected N. clevelandii or sap from buffer-inoculated N.clevelandii amended with CNV virion RNA (6.7 µg/ml) was incubatedat 30 C for varying times and then inoculated to C. quinoa. Sapfrom both PD(-)- and wild-type-infected plants induced locallesions for all time points, but the number of lesions for wild-type was consistently greater than that for PD(-) (Fig. 2.10).Given that wild-type sap was diluted 100-fold more than that ofPD(-), it appears that the wild-type inoculum is significantly moreinfectious than PD(-) inoculum. No lesions were observed on plantsinoculated with CNV virion RNA mixed with healthy plant sap. CNVRNA (6.7 µg/ml) in 10 mM sodium phosphate buffer, pH 7.0, and66U)Z 40OU)w -—I 30C.)020CC1112 100^ 0^ 0(V)TIME OF INCUBATION IN VITRO (MINUTES)Fig. 2.10. Local lesion assays of PD(-), wild-type, and CNV RNA onC. quinoa. Sap from second-passage, PD(-)- or wild-type-systemically infected plants, and from healthy N. clevelandii amended with CNV RNA (6.7 µg/ml), was used to inoculate leaves ofC. quinoa, a local lesion host for CNV. PD(-)- and healthy leaftissues were ground in 2 volumes, and wild-type in 200 volumes, of10 mM sodium phosphate buffer, pH 7.0. The samples were left at30° C and aliquots were removed to inoculate C. quinoa after 1, 10, JU anti 60 minutes. Local lesions were counted 3 dpi.67retained on ice for the duration of the experiment, was used toinoculate C. quinoa. More than 100 local lesions/inoculated leafwere produced (data not shown), demonstrating that the CNV RNA usedin this experiment was infectious. The local lesion assay wasrepeated a second time, however, the time course of incubation wasextended to 180 minutes. A similar trend was observed as in thefirst assay. Local lesions were induced for all time points bywild-type- sap and at each time point more local lesions wereinduced by wild-type than by PD(-). Sap from PD(-) -infected plantsinduced local lesions for all times points but the number of locallesions induced declined over time and at 180 minutes only a singlelocal lesion was observed on one leaf of one plant. One lesion wasobserved on one leaf of one plant inoculated with CNV virion RNAmixed with healthy plant sap (1 min incubation) but no lesions wereobserved on any of the other plants inoculated with this treatment.The fact that PD(-) sap retained its infectivity under conditionsthat led to the rapid degradation of infectious CNV virion RNAsuggests that PD(-) RNA is protected in some manner from hostribonucleases.11.3.8 Cloning, sequencing, and infectivity of the ca. 3800 nt. RNAAs shown in the Northern blot experiment in Fig. 2.9, the mostprominent RNA species in systemically infected leaves of PD(-)transcript- or sap-inoculated plants, is approximately 3800nt, about 900 nucleotides smaller than wild-type CNV genomic RNA(and approximately 600 nt smaller than PD(-) transcript). It is68possible that the reduction in the size of PD(-) RNA is due todeletions in the CNV coat protein ORF that may have occurred inplanta during PD(-) RNA replication. This may also explain theinability to detect CNV virions or CNV coat protein in PD(-)-infected plants.Reverse transcription of ssRNA from PD(-)-infected plants,followed by DNA amplification using the polymerase chain reaction(RT-PCR), was used to determine which, if any, region of the CNVcoat protein ORF remained in passaged PD(-)-infected plants. Theoligonucleotide primers used were complementary to regionsapproximately 180 nt and 130 nt upstream and downstream,respectively, of the CNV coat protein ORF. The DNA recovered afterRT-PCR of ssRNA extracted from third and fourth passage, PD(-)-infected N. clevelandii, was approximately 380 by in length, ascompared to the expected, and observed, 1471 by fragment recoveredwhen wild-type RNA was similarly treated (Fig. 2.11). From thesequences of eight independent clones (Fig. 2.12) it was determinedthat the deletion within the PD(-) genome that resulted in theproduction of the smaller ca. 3800 nt species corresponded to aloss of almost the entire coat protein coding sequence. The first27 nucleotides of the coding region for the coat protein R domainwere retained, followed by 14 nucleotides corresponding to, but outof frame with, the 5 amino acid hinge, and, finally, the terminal33 nucleotides of the coat protein P domain (Fig. 2.13). It isimportant to note that the region deleted from the PD(-) coatprotein ORF was apparently specific; the eight clones sequenced,691^2 3 4 5 6.4-- 21.8 kbrrir^ I/ 5.05 kb1.98 kb4=1.90kb41-- 1.57 kb.4--1.32 kb4— 0.93 kb•—• 0.84 kb4— 0.58 kb1.55 kb0.68 kb0.59 kb0.25 kb0.13 kb —10.Fig. 2.11. DNA products recovered after RT-PCR of high-molecularweight ssRNA extracted from N. clevelandii after the fourth passageof wild-type or PD(-) inoculum. Lane 1, PvuII/RsaI-digestedBluescribe (300 ng; used as a molecular size standard); lane 2, DNAfrom RT-PCR of PD(-)-infected plants; lane 3, DNA from RT-PCR ofwild-type-infected plants; lane 4, DNA after RT-PCR of CNV virionRNA; lane 5, DNA after PCR amplification of cloned, wild-type, CNVplasmid DNA; lane 6, EcoRI/HindIII-digested lambda DNA molecularsize markers (300 ng). Lanes 2-5 were loaded with 2 gl aliquotsfrom each RT-PCR reaction. After RT-PCR, DNA samples wereelectrophoresed through a 1% TAE agarose gel, and stained with 0.5ug/ml ethidium bromide. The sizes of the molecular markers in kbpare indicated on the left (PvuII/RsaI-digested Bluescribe) andright (EcoRI/HindII-digested lambda) of the gel.70CNV TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 1 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 2 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 3 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 4 TGGACAGATGGGAAATGGACCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 5 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 6 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 7 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGclone 8 TGGACAGATGGGAAATGGATCTTTTTGGAGAGGAGGGTGTTGACGGCGATGCNV AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 1 AGCCCAGCATCCTTGACTCCGCCGTAGCACGACCAAGCAAACACAAACACAclone 2 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 3 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 4 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 5 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 6 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 7 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACAclone 8 AGCCCAGCATCCTTGACTCCGCCGTAGCATGACCAAGCAAACACAAACACACNV ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 1 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 2 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 3 ACGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 4 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 5 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 6 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCclone 7 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATTclone 8 ATGGCACTCGTAAGCAGGAACAACAATAACGTCGCCTCTTCTCGAGCAATCCNV ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGG-CTTCTTGAATCTAACCclone 1 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGG-CTTCTTGAATCTAACCclone 2 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGG-CTTCTTGAATCTAACCclone 3 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGG-CTTCTTGAATCTAACCclone 4 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGG-CTTCTTGAATCTAACCclone 5 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGGCTTCTTGAATCTAACCclone 6 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGGCTTCTTGAATCTAACCclone 7 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGGCTTCTTGAATCTAACCclone 8 ACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGGCTTCTTGAATCTAACCFig. 2.12. Nucleotide sequences of eight independent clones of theRT-PCR products of the ca. 3800 nt RNA isolated from the third(clones 5-8), or fourth (clones 1-4) passages of PD(-) sap-inoculated, N. clevelandii. The sequenced portion starts at CNVnucleotide 2527 and ends at CNV nucleotide 3795. The CNV sequenceis shown as DNA and is presented for comparative purposes; areasof deleted sequence are shown diagramatically in Fig. 2.12 B.Nucleotides that differ from the published CNV sequence (Rochon andTremaine, 1989) are written in bold-face. The start and stopcodons for the CNV coat protein gene are underlined.71ACNVPD(-)CP(-)coat proteinsubgenomic RNAir start site^R domain1 (2606 nt) GACCAAGCAAACACAAACACAATGGCACTCGTAAGCAGGAACAACAATA (749 nt)1 (2606 nt) GACCAAGCAAACACAAACACAATGGCACTCGTAAGCAGGAACAACAATA (749 nt)1 (2606 nt) GACCAAGCAAACACAAACACAATGGCACTCGTAAGCAGGAACAACAATA ir carboxy-terminal portion of P domainCNV 3405 ACGTCGCCTCTTCTG (308 nt) GAGCAATCACTGAAAATGCGGTGCAGGTTGTGTAA (931 nt)PD(-) 3405 ACGTCGCCTCTTCTC^GAGCAATCACTGAAAATGCGGTGCAGGTTGTGTAA (931 nt)CP(-) 3405 ACGTCGCCTCTTCTC GAGCAATCACTGAAAATGCGGTGCAGGTTGTGTAA (931 nt)f hingeBCNV coatprotein ORFCP(-)R^a- 111111111rS^hF g. 2.13. Location of CP(-) coat protein nucleotide sequences relative to wild-type andP (-) coat protein sequences. (A) The relevant nucleotide sequences of wild-type, PD(-),a d CP(-) (clone 2, Fig 2.11) are compared beginning with the coat protein subgenomic starts te and ending with the coat protein TAA (UAA) termination codon (initiation and terminationc•dons are underlined). For reference, the nucleotide positions corresponding to theb ginning of the R (random) domain and hinge coding sequences are indicated. The remainingc rboxy-terminal portion of the protruding (P) domain is also indicated. Dashed linesr present areas of deletion in PD(-) and CP(-) and the numbers in brackets indicaten cleotides present in each genome, but not shown here. (B) A schematic diagram showing whicha eas of the CNV coat protein coding sequence are retained in CP(-).representing RNA extracted from two different passages of PD(-),shared the same start and stop nucleotides for the deletions (Fig.2.13 A).To determine if CNV RNA can in fact replicate and spread fromcell-to-cell and systemically in the complete absence of the coatprotein and/or particle assembly, as is suggested by the studiesreported with the PD(-) mutant, the RT-PCR product correspondingto clone 2 (Fig 2.12) was used to replace the corresponding regionof wild-type CNV cDNA clone. In vitro transcripts of one suchclone, designated CP(-) (Fig 2.13), were used to inoculate N.clevelandii. The symptoms produced were similar to those observedin sap-inoculated, PD(-)-infected plants in that local lesionsproduced on the inoculated leaves were of the same size in eachinoculated plant and were indistinguishable from wild-type.Systemic symptoms appeared 7-12 d.p.i (Fig. 2.14), although, aswith PD(-)-sap-inoculated N. clevelandii, CP(-) sometimes inducedsystemic symptoms after those of wild-type appeared. In two of sixindependent transcript-inoculations of N. clevelandii, an oak-leafnecrosis was induced as opposed to a wild-type necrosis (Fig. 2.15D). It should be noted that under different environmentalconditions (in a greenhouse under a day/night temperature regimeof 22°C/17°C with a 14 hour photoperiod), CP(-)-transcriptinoculated N. clevelandii consistently produced an oak leafnecrosis (Dr. T. Sit, personal communication). Northern blotanalysis of RNA extracted from CP(-)-infected plants demonstratedthat CP(-) replicates to high levels in plants (comparable to that73observed with wild-type) and does not appear to undergo furtherdeletion (Fig. 2.16). These results demonstrate that CNV RNA canindeed replicate in the absence of the complete coat protein andthat the coat protein is not obligatory for cell-to-cell orsystemic movement of this virus.74SUFFER INOCULFig. 2.14. Systemically infected N. clevelandii 7 days after transcript inoculation. Plantswere mock inoculated with 0.01 M sodium phosphate buffer, pH 7.0 (A), or were inoculated withequal amounts of in vitro transcribed RNA from wild-type (B), or CP(-) (C) cDNA clones.Fig. 2.15. Systemically infected N. clevelandii 22 days aftertranscript inoculation. Plants were mock inoculated with 0.01 Msodium phosphate buffer, pH 7.0 (A), or were inoculated with equalamounts of in vitro transcribed RNA from wild-type (B), PD(-) (C),or CP(-) (D) cDNA clones.761 2 3 4 5 6 74— 4.7 kbFig. 2.16. Northern blot analysis of high molecular weight RNAextracted from the inoculated- and systemically-infected leaves oftranscript inoculated CP(-)-infected N. clevelandii. RNA sampleswere denatured with methylmercuric hydroxide, electrophoresedthrough a 1% agarose gel containing methylmercuric hydroxide,blotted and hybridized with a 'P-labelled nick-translated full-length, CNV cDNA probe. Lane 1 contained 50 ng of CP(-) in vitro transcribed RNA (3636 nt). Lanes 2-6 each contained 500 ng of 2M LiCl-insoluble RNA extracted from: CP(-) transcript-inoculatedleaf (lane 2); CP(-) systemically infected leaf from a transcript-inoculated plant (lane 3); PD(-) systemically infected leaf (lane4); wild-type systemically infected leaf (lane 5); buffer-inoculated leaf (lane 6). Lane 7 contained 50 ng of CNV virion RNA(4701 nt). The upper band in lane 1 corresponds to linearizedplasmid DNA template used in the transcription reaction. The smearin lane 7 is due to degradation of the virion RNA.7711.4 DISCUSSIONA mutant of CNV lacking the coding sequence for the P domainof the coat protein was found to replicate in transcript-inoculatedleaves of N. clevelandii and to move systemically, although withreduced efficiency, in transcript-inoculated plants.Interestingly, the initial reduction in lesion size in thetranscript-inoculated leaves was lost upon subsequent sapinoculations of N. clevelandii. Systemic symptoms were inducedin wild-type, XpK2/M5, and PD(-) sap-inoculated N. clevelandii in7 to 12 days, but there was a trend for PD(-)-infected plants toinduce systemic symptoms slightly later than wild-type or XpK2/M5.Systemically infected leaves of PD(-)-inoculated plantsaccumulated high levels of a stable RNA of ca. 3800 nt. [PD(-) RNAis 4385 nt in length]. There is some precedence among the plantviruses for the occurrence of such spontaneous deletion events.Petty and Jackson (1990) found that coat protein deletion mutantsof barley stripe mosaic virus gave rise to further deletionderivatives that could be detected in systemically infected leavesof transcript-inoculated plants. The CNV 3800 nt RNA, whichremained the dominant viral RNA species for all passages of PD(-),when cloned and sequenced, was found to contain only 75 of the 1140nt of the coat protein ORF. The deletion within the coat proteinORF did not appear to be a random event, as all eight of the clonesthat were sequenced, although differing internally by one or twonucleotides, had precisely the same sequences removed. The factthat all eight clones retained the 5' terminal 27 nucleotides and78the 3' terminal 33 nucleotides of the PD(-) coat protein ORF, mayindicate that these regions have alternative function(s), perhapsin gene regulation, that are independent of their constitutive roleas part of the coat protein gene. The further deletion of PD(-)RNA appears to render it better able to replicate and/or move fromcell to cell into the vascular system, as after sap transmission,the 8-12 day delay in the development of systemic symptoms observedin transcript-inoculated plants was lost or significantly reducedin PD(-) sap-inoculated plants.It is unlikely that the deleted PD(-)-template moves withinthe plant and from plant-to-plant as naked RNA, a form which wouldleave it susceptible to host ribonucleases. The possible stableassociation of this RNA with a protein is supported by the findingthat sap from second passage, PD(-)-infected N. clevelandii remained infectious under conditions that led to the rapiddegradation of otherwise highly infectious CNV virion RNA. It ispossible that one of the CNV non-structural proteins, such as p21,which has already been implicated in cell-to-cell movement (Rochonet al., in press) or a host protein, could function to protect CNVRNA in a manner similar to that suggested for the tobacco mosaictobamovirus p30 movement protein or the cauliflower mosaiccaulimovirus gene I product (Citovsky et al., 1990; Citovsky etal., 1991).Several plant viruses are known to move systemically in theabsence of a functional coat protein. Coat protein mutants of thesingle-stranded DNA geminiviruses African cassava mosaic, maize79streak, tomato golden mosaic, or the RNA viruses, red clovernecrotic mosaic dianthovirus, barley stripe mosaic hordeivirus andtobacco mosaic tobamovirus have been found to move systemically,although in some cases erratically or inefficiently, in systemichost species (Stanley and Townsend, 1986; Lazarowitz et al., 1989;Gardiner et al., 1988; Xiong et al., 1991; Petty and Jackson, 1990;Dawson et al., 1988). There are also several plant viruses thatcannot move systemically at all without an intact coat protein,examples being cucumber mosaic cucumovirus, cowpea mosaic comovirusand brome mosaic bromovirus (Suzuki et al., 1991; Wellink and VanKammen, 1989; Sacher and Ahlquist, 1989).The CNV coat protein mutants, PD(-) and CP(-), share somecharacteristics with other coat protein mutants described for othertombusviruses, and also the related carmoviruses anddianthoviruses. Scholthof et al. (1993) determined that the coatprotein of TBSV was not required for replication, subgenomic RNAsynthesis, translation of viral mRNAs, and, as with CNV, for bothlocalized and systemic movement of this virus. Transcripts of aTBSV coat protein mutant, pHS7 (Scholthof et al., 1993), caused adelay in local lesion development when compared with wild-type TBSVtranscript in N. clevelandii and N. benthamiana. pHS7 was derivedfrom a full-length TBSV cDNA clone in which the first 72nucleotides of the coat protein gene were followed by a 50nucleotide deletion, placing the remainder of the coat protein geneout of frame. The host response to pHS7 infection differed fromthat induced by the CNV mutants PD(-) and CP(-), in that no delay80in local lesion development was observed for either CNV mutant intranscript-inoculated leaves. Similar to PD(-) and, often CP(-),systemic symptom development was delayed in pHS7 transcript-inoculated N. clevelandii. However, unlike PD(-), furtherdeletions in the pHS7 coat protein ORF did not occur since full-length pHS7 genomic RNA was recovered from systemically-infectedN. clevelandii (Scholthof et al., 1993).TBSV coat protein mutants differed from PD(-) and CP(-) inthat the TBSV mutants rarely induced the lethal necrosis thattypifies wild-type TBSV infections (Scholthof et al., 1993). BothCNV coat protein mutants, PD(-) and CP(-), in either transcript-or sap-inoculated plants, were usually lethal to N. clevelandii,as was wild-type transcript. Scholthof et al. (1993) proposed thatthe non-lethal necrosis produced in response to pHS7 infectionmight have been a function of the mutated TBSV coat protein, inthat the coat protein itself could influence symptom severitydirectly. This is not the case with CNV, however, as the coatproteinless mutant CP(-) was lethal to N. clevelandii.The movement of CNV in a systemic host also differs from thatreported for another tombusvirus, cymbidium ringspot virus (CyRSV;Dalmay et al., 1992). A coat protein mutant of CyRSV, CP1, whichresulted in a deletion of two amino acids from the coat proteinshell domain, moved systemically in N. benthamiana, but not in N.clevelandii (Dalmay et al., 1992). No CyRSV particles wereobserved in CP1-infected N. clevelandii or N. benthamiana tissue,yet CyRSV coat protein subgenomic mRNA and CyRSV coat protein could81be detected in the inoculated- and systemically-infected leaves ofN. benthamiana and in the transcript-, but not systemically-,infected leaves of N. clevelandii. Additionally, the coat proteinmutation was stable and the mutant genome underwent no furthermodifications as seen with the CNV mutant PD(-). The authorsconcluded that CyRSV RNA does not have to be encapsidated to movefrom cell-to-cell, but, depending on the host, encapsidation maybe required for long distance movement. CNV, in contrast, can movesystemically in either N. clevelandii or N. benthamiana in theabsence of RNA encapsidation (see above). Whether virion assemblyis required for systemic movement of CNV in some hosts is difficultto ascertain as consistent systemic movement of this virus isrestricted to N. clevelandii and N. benthamiana, both of whichsupport the long distance movement of sap-transmitted PD(-).Turnip crinkle carmovirus (TCV) is considered to be a closerelative of the tombusviruses as the coat protein subunit andparticle structures of the tombus- and carmo- viruses are highlysimilar (Hogle et al., 1986) as are their genomic organizations andamino acid sequences (Carrington et al., 1987; Rochon and Tremaine,1989; Riviere and Rochon, 1990). A TCV mutant that carried tandem,in-frame, translation terminators within the five amino acid hingelocated directly upstream of the protruding domain sequence of theTCV coat protein gene, was found to replicate well in Brassica campestris protoplasts. Symptoms were produced in transcript-inoculated C. amaranticolor, a local lesion host for TCV, but notin transcript-inoculated B. campestris or N. benthamiana, both82systemic hosts (Heaton et al., 1991). Nucleic acid analysisrevealed detectable levels of full-length TCV coat protein mutantRNA only in the transcript-inoculated leaves of C. amaranticolor and B. campestris, while analysis of total proteins from thesehosts was negative for coat protein. It would appear that TCV mayrequire coat protein for systemic movement, but not for cell-to-cell movement. CNV, unlike TCV, does not require coat protein forsystemic movement, suggesting that coat protein is not required forsuch movement of wild-type CNV. Coat protein, however, mayfacilitate more rapid systemic movement of wild-type CNV RNA.The differences between the TCV and CNV coat protein mutantsand the responses they induce in their respective hosts may be due,in part, to two of the smaller, non-structural genes encoded bythese viruses. CNV has a bifunctional, 3' co-terminal, subgenomicmRNA that codes for p21 and p20, the functions of which are notcertain, although Rochon and Johnston (1991) speculate that p21 maybe involved in virus movement. TCV also encodes two proteins froman internally located bicistronic messenger RNA; p8 and p9 are non-structural proteins that have been shown to be necessary for, andto function together, in cell-to-cell movement (Hacker et al.,1992). It is possible that p21 and p20 (or a host-encoded protein)may provide the necessary interaction that permits CP(-) RNA tomove systemically. p8 and p9 of TCV may function in an entirelydifferent way to assist cell-to-cell transport of this virus, andconsequently, may provide no protection for the TCV P-domainlessRNA.83Similarly, coat protein was not required for cell-to-cellmovement of red clover necrotic dianthovirus (RCNMV) in thesystemic hosts N. benthamiana and N. clevelandii (Xiong et al.,1993). Symptoms typical of a wild-type infection appeared on thetranscript-inoculated leaves of either host in one week. However,only N. benthamiana maintained at 15°C developed typical wild-typesystemic symptoms. At the elevated temperature of 25°C, systemicsymptom development in N. benthamiana was delayed by two weeks atwhich time an atypical "oak leaf" pattern of necrosis developed ineither the first, or sometimes the second, non-inoculated leavesdirectly above that inoculated with transcript. No other leavesbecame infected. Xiong et al. (1993) proposed that this limitedsystemic movement is not true long distance transport through thevascular tissue, but instead is a form of cell-to-cell movementthrough stem cells. Systemic symptoms did not develop ontranscript-inoculated N. clevelandii at 15°C or 25°C. Theobservations of Xiong et al. (1993) differ significantly from thestudies described here for the CNV coat protein mutants, PD(-) andCP(-), on N. clevelandii and N. benthamiana. In N. clevelandii,PD(-) produced smaller local lesions on transcript-inoculatedleaves than wild-type (see section 11.3.1) which may indicate someinitial restriction in cell-to-cell movement. There was an 8 to12 day delay in the development of systemic symptoms in PD(-)-transcript-inoculated N. clevelandii, similar to RCNMV on N.benthamiana at 15°C; but when PD(-) was sap transmitted, the delayin PD(-)-induced systemic infection, when it occurred, was84significantly reduced. Similar to the RCNMV coat protein mutantin N. benthamiana, an oak leaf necrosis sometimes developed in PD(-) transcript and sap-inoculated, and CP(-) transcript-inoculatedN. clevelandii, but this necrosis was observed throughout the plantand was not restricted to one or two leaves as with the RCNMVmutant.A comparison of the results reported for the various coatprotein mutants described above allows for some interestingobservations. CNV, TBSV, CyRSV, TCV, and RCNMV can all move cell-to-cell in the absence of a functional coat protein in at least oneof their host species. This is consistent with the model proposedby Xiong et al. (1993), and others (Atabekov and Dorokhov, 1984;Citovsky and Zambryski, 1991), that some plant viruses (eg. RCNMV,TMV, CaMV) may move from cell-to-cell in an unencapsidated form,probably in association with a movement protein that protects theRNA from degradation, modifies its form to allow passage from cell-to-cell, and targets it to the plasmodesmata. Additionally, asreported by others (Xiong et al., 1993; Dalmay et al., 1992;Scholthof et al., 1993), it would appear that N. clevelandii is amore restrictive host than N. benthamiana in terms of permittingsystemic movement of these plant viruses. Neither the RCNMV norCyRSV coat protein mutants described above, nor two TBSV coatprotein mutants in which the coat protein gene was replaced witha reporter gene (Scholthof et al., 1993), were able to movesystemically in N. clevelandii, while long distance movement ofthese mutants was permitted in N. benthamiana. Movement of the CNV85mutants PD(-) and CP(-) in N. clevelandii was not restricted asboth mutants were able to move systemically in this host. The factthat CNV, TBSV, CyRSV, TCV and RCNMV all vary in their local andsystemic movement in N. clevelandii and N. benthamiana emphasizesthe importance of host factors, as well as virus-encoded factors,in determining whether a virus can move from cell-to-cell orsystemically in a certain host species.The mechanism by which the remaining coat protein sequencesof PD(-) were deleted to produce the ca. 3800 nt RNA species inPD(-)-infected plants is unknown. However, some insight may begained from studies of the defective interfering (DI) RNAs producedby the tombusviruses CNV, TBSV and CyRSV (Rochon, 1991; Hillman etal., 1987; Rubino et al., 1990). These sub-viral RNAs that areassociated with some animal, fungal, and plant viruses, aredeletion mutants of genomic RNA that require the virus from whichthey were derived for their replication and, usually,encapsidation. DI RNAs affect the parental virus in a manner thateither exacerbates or attenuates the viral infection. CNV DI RNAsattenuate infection in systemic host species such as N. clevelandii so that the lethal necrosis associated with wild-type transcriptbecomes, instead, a persistent mild necrosis (Rochon, 1991; Finnenand Rochon, 1993).The manner in which plant viral DI RNAs are produced is notdefinitively known, but it has been speculated that DI RNAs.areproduced by template switching of the viral replicase (Rochon,1991; Roux et al., 1991). A model for such "copy-choice" RNA86recombination is described by Kirkegaard and Baltimore (1986) andLai (1992). Briefly, the viral replicase pauses at a site ofsecondary structure on the template RNA and then dissociates fromthe template. The nascent RNA strand, along with the boundreplicase, then binds at a different site on the same or adifferent template and nascent RNA transcription resumes. It ispossible that the ca. 3800 nt RNA species that becomes dominantafter PD(-)-inoculation of N. clevelandii is produced by amechanism such as RNA recombination. Deletion of the P domaincoding sequence could alter the secondary structure of the PD(-)template such that new stem loops are formed that are not normallyassociated with wild-type RNA and that promote replicase switching.The ca. 3800 nt RNA species is preferentially selected in planta,at the expense of the full-length PD(-) template; it may be thatthe smaller RNA species is able to replicate more efficiently thanthe original PD(-) in a manner similar to DI RNAs that replicateat the expense of their parental helper RNAs (Jones et al., 1990).It is unlikely that the ca. 3800 RNA species is a product of RNAsplicing, a process which is normally restricted to the cellnucleus. Tombusvirus replication is believed to occur in the cellcytoplasm; Appiano et al. (1986) identified multivesicular bodiesand, when present, adjacent chloroplast invaginations, as the sitesof TBSV replication in TBSV-infected Gomphrena qlobosa cells.If the PD(-) coat protein gene is translated in vivo, as itcan be in vitro (section 11.3.5), it is possible that expressionof only the shell portion of the coat protein is deleterious to the87mutant, and therefore subsequent selection is for further coatprotein deletion mutants. PD(-) coat protein could be producedwhich, although lacking the capacity to encapsidate RNA, is ableto bind and sequester PD(-) RNA, preventing its subsequentreplication. RNA templates that are produced (perhaps as a resultof RNA recombination) but that lack the coat protein ORF wouldtherefore have a selective advantage in that they would not producecoat protein and/or virions and so would not be self-limiting.Deletion of the P domain coding sequence to produce PD(-) mayhave interfered with the promoter sequence for the p21/20subgenomic RNA and consequently production of the p21 (movement)and p20 proteins. The core promoter sequences required for in vivo amplification of brome mosaic virus (BMV) and alfalfa mosaic virus(A1MV) coat protein subgenomic RNAs have been mapped to sequencesextending upstream to positions -96 (BMV) and -136 (A1MV) relativeto their subgenomic start sites (designated +1) (French andAhlquist, 1988; van der Kuyl, 1991). If promoters for the two CNVsubgenomic RNAs extend as far upstream from their start sites asthose for BMV and A1MV, the p21/20 promoter would be expected tobe affected by the P domain deletion. The in planta deletion ofthe PD(-) coat protein ORF to produce the 3.8 kb RNA positions theputative coat protein promoter next to the p21/20 ORFs. Such adeletion would be likely be selected for during PD(-) replicationas it could restore p21/20 production thus permiting cell-to-cellmovement of the PD(-) deletion derivative, and consequently, thewild-type symptoms observed on PD(-) sap-inoculated leaves.88CHAPTER III. FUNGAL TRANSMISSION OF CNV: A STUDY USING CHIMERICPLANT VIRUSESIII.1 INTRODUCTIONThe role of fungi in the transmission of plant viruses isoften overlooked in vector relationship studies as the overwhelmingmajority of plant pathogenic viruses are spread by arthropods.Nevertheless, virus transmission by fungi warrants study as manyof these viruses cause economically important diseases of staplefood crops.111.1.1 The life cycle of fungal vectors of plant virusesThere are, at present, two classes of fungi that containgenera known to transmit viruses: the Chytridiomycetes and thePlasmodiophoromycetes. Both are considered to be primitive classesas they are characterized as zoosporic, obligate biotrophsdependent on water for their mobility. The life cycles of thechytrid and plasmodiophorid fungi are generally similar. Thedescription below is derived from descriptions by Karling (1968),Thouvenel and Fauquet (1980), Webster (1980), Barr (1988), andAdams (1991).A thick walled cyst, that functions as a resting structure,germinates to produce a single, heterokont zoospore referred to asthe primary zoospore. The zoospore propels itself through the soilwater by means of one (chytrids) or two (plasmodiophorids) flagellauntil it contacts a suitable host root. The flagella retract andthe zoospore, in contact with a root hair or an epidermal cell inproximity to a root hair, encysts. The zoospore adhprs to the 89wall of a host cell. The protoplasmic contents of the zoospore areforcibly injected into the host root cell and are present withinthe host cell as naked thalli or plasmodia.The zoosporangial plasmodia move within the host root bycytoplasmic streaming. Once in an epidermal cell the zoosporeprotoplast divides to become a multinucleate plasmodium. Commonly,many plasmodia will be found within a single host cell and thesewill aggregate together. A zoosporangium is formed when wallmaterial is laid down over the membranes of adjoined plasmodia.As the zoosporangium matures the walls between the individualsporangia breakdown to produce a single, multinucleate body whilethe peripheral walls remain intact. The protoplasm within thezoosporangium then cleaves to produce secondary zoospores ready fordischarge into the soil water.Small plasmodial cells not involved in the formation of thezoosporangia are believed to produce an exit in the host cell wall,possibly by enzymatic degradation, for release of the zoospores.The zoospores are expelled into the rhizosphere, their flagella areactivated and they swim away to contact other root cells. In theabsence of these exit-producing plasmodial cells, the escape porein the host cell wall is not produced and the zoospores remainentrapped within the host. Most of these zoospores will die.Some, however, may move farther into the host to give rise to newzoosporangia by secondary infection or, more commonly, cystogenousplasmodia and resting spores. Resting spores may also be formedfrom infection of plants by the primary zoospore or, more90frequently, from secondary zoospores released through the host cellwall.111.1.2 Plant viruses transmitted by fungiThose viruses known to be transmitted by fungal vectors varyin their taxonomic classification (Table 3.1). To date all arepositive-sense, single-stranded RNA viruses with the exception oftobacco stunt virus, that has a double-stranded RNA genome. Theymay be mono- or multi-partite, rod-shaped or icosahedral. Thoseviruses transmitted by the chytridiomycetes Olpidium brassicae (Wor.) Dang and 0. radicale Schwartz & Cook are members of thetombusvirus, carmovirus, dianthovirus, and necrovirus groups or areunclassified. Viruses transmitted by the plasmodiophorids Polymyxa betae Keskin, P. qraminis Ledingham, and Sponcrospora subterranea (Wallr.) Lagerh, are furovirses, baymoviruses, or have yet to beclassified.The mechanism by which plant viruses are transmitted to hostplants by their fungal vectors is not well understood, but theassociation of a virus with its vector is a specific one and is nota simple matter of passive acquisition. The transmission of avirus by its fungal vector has been categorized as in vitro or invivo (Campbell, 1993). Viruses acquired in vivoViruses that are acquired in vivo, such as beet necroticyellow vein virus (BNYVV), soil-borne wheat mosaic virus (SBWMV)91Table 3.1 Fungal vectors of plant viruses (Adams, 1988; Adams,1991; Brunt, 1988).Virus VectorBAYMOVIRUSESbarley yellow mosaic virus Polymyxa qraminisoat mosaic virus P. qraminisrice necrosis mosaic virus P. qraminiswheat spindle streak mosaic virus P. qraminiswheat yellow mosaic virus P. qraminisFUROVIRUSESbeet necrotic yellow vein virus P. betaebeet soilborne mosaic virus P. betaebroad bean necrosis virus P. qraminisoat golden stripe virus P. qraminispeanut clump virus P. qraminispotato mop-top virus Spongospora subterranearice stripe necrosis virus P. graminissoilborne wheat mosaic virus P. graminisTOMBUSVIRUSEScucumber necrosis virus Olpidium radicaleCARMOVIRUSESmelon necrotic spot virus 0. radicaleDIANTHOVIRUSESred clover necrotic mosaic virus O. radicaleNECROVIRUSES AND SATELLITEStobacco necrosis virus O. brassicaetobacco necrosis satellite O. brassicaeUNCLASSIFIED VIRUSESfreesia leaf necrosis virus O. brassicaelettuce big vein virus O. brassicaelettuce ring necrosis virus O. brassicaepepper yellow vein virus O. brassicaetobacco stunt virus O. brassicaewatercress chlorotic leaf spot virus S. subterraneawatercress yellow spot virus S. subterranea 92and tobacco necrosis virus (TNV), are those that persist for longperiods of time (years) in the resting spores of the fungal vector(Jones and Harrison, 1969; Slykhuis and Barr, 1978; Schlosser,1987; Campbell, 1985; Hiruki, 1987). They cannot be acquired bytheir vectors after both virus and vector have been released fromthe host plant. The acquisition occurs in planta, however, howthis occurs is not known. Attempts to transmit viruses such astobacco stunt virus, lettuce big vein virus, wheat spindle streakmosaic virus and barley yellow mosaic virus, by mixing virus-freezoospores of the vector species with the virus have failed (Hiruki,1965; Campbell and Grogan, 1964; Slykhuis and Barr, 1978; Adams etal., 1988). Viruses acquired in vivo are not believed to replicatein vector zoospores or resting spores. Viruses acquired in vitroViruses that are acquired in vitro, such as CNV and melonnecrotic spot carmovirus (MNSV), are those that do not have a long-term association with the resting structures of their vectorspecies. The virus is adsorbed onto the plasmalemma of the vectorzoospores after release of both the virus and the vector into therhizosphere of the host plant and it is assumed that the virusenters the zoospore when the flagellum is retracted prior toencystment on the root cell. The virus is subsequently transferredinto the plant when the fungal protoplasm is injected into the hostcell.Viruses that are acquired in vitro do not remain associated93with their vectors in planta. This was demonstrated in severalstudies in which acquisition of the virus by vector zoospores wasreduced or eliminated if the virus was mixed with its homologousantiserum (Teakle and Gold, 1963; Campbell and Fry, 1966; Dias,1970b). Had the virus been associated with its vector upon releasefrom the host plant, transmission would not have been blocked bythe antiserum. Viruses acquired in vivo are not neutralised byantisera in this way (Rao and Brakke, 1969).111.1.3 Transmission of CNV by 0. radicale Natural transmission of CNV is facilitated by the zoosporesof 0. radicale. Dias (1970a) demonstrated that CNV transmissionwas dependent on the presence of viable 0. radicale zoospores;transmission was prevented when zoospores were removed byfiltration or were heat-killed. Zoospores discharged from theroots of CNV-infected cucumber were found to be free of the virusupon their release and subsequent virus acquisition was rapid,occurring 5 to 10 min after zoospore discharge (Dias, 1970 b). CNVis not carried internally in the resting spores of 0. radicale andis not transmissible by 0. brassicae, a vector of other viruses(Dias, 1970 a, b). 0. radicale is also known to transmit thecarmovirus MNSV, and is the putative vector of RCNMV (Campbell etal., 1991; Bowen and Plumb, 1979), but does not transmit the cherrystrain of TBSV [TBSV-Ch (also referred to as PAMV); Campbell etal., 1991].The mechanism by which CNV is transmitted to cucumber by 0.94radicale is not fully understood. Stobbs et al. (1982) observedCNV virions adsorbed to the plasmalemma and axonemal sheath of 0.radicale zoospores and speculated that the virus entered thezoospore when the flagellum was retracted. Stobbs et al. (1982)reported that after flagellar retraction the axonemal sheath, whichsurrounds the flagellum and is continuous with the plasmalemma, wastaken into the zoospore. They observed a "whorl of membranes" inthe periphery of the zoospore that appeared to be covered withvirus particles. However, it remains unclear if the axonemalsheath is in fact taken into the plasmalemma when the flagellum isretracted and if the virus particles associated with the sheath arethose that are transmitted to and infect the host plant (Adams,1991).111.1.4 The role of the CNV coat protein in fungal transmissionThe mechanism by which CNV is transmitted by 0. radicale hasreceived little attention since 0. radicale was first identifiedas the vector of CNV (Dias, 1970 a,b). There is some evidence,however, that the coat protein of CNV plays a role in facilitatingthis transmission. Dias (1970b) reported that transmission of CNVby O. radicale was abolished if CNV was discharged into CNV-specific antiserum prior to incubation of the virus with vectorzoospores. The antiserum may have interfered with transmission bybinding to antigenic sites on the virus capsid that were alsorequired by CNV for adsorption to the plasmalemma of 0. radicale zoospores. Stobbs et al. (1982) concluded that specific adsorption95of CNV virions by 0. radicale zoospores was associated with thevirus capsid as attempts to transmit CNV virion RNA with O.radicale to cucumber failed.Amino acid sequence similarity exists between carmoviruses(MNSV, CarMV and TCV), and tombusviruses (CNV, TBSV-Ch, and TBSV-Bs-3) (see Riviere et al., 1989, and references therein). Thedifferent domains of the coat proteins of these viruses shareddifferent degrees of similarity; the S domains were most similar,the R domain and arm less so, and the P domains least similar.Interestingly, Riviere et al. (1989) discovered a region ofsequence similarity in the P domains of CNV and MNSV that was notshared in the other tombus- or carmo-virus sequences analyzed. Ofthe six viruses whose sequences were compared, only CNV and MNSVare naturally restricted to infecting cucurbits and are transmittedby O. radicale. Riviere et al. (1989) speculated that the regionof conservation in the P domains of CNV and MNSV may play a rolein determining host range and/or vector transmission of theseviruses.CNV and TBSV are genetically very similar yet areserologically unrelated and have very different biologicalcharacteristics. CNV and TBSV share extensive amino acid sequencesimilarity throughout their genomes (89.2%, 92.6%, 89.5%, 73.6%,for the p33, p92, p21, and p20 genes, respectively) with the coatprotein genes being the least conserved (37.5%) (Hearne et al.,1990). TBSV has a broad natural host range that includes woody andherbaceous host species while natural infections of CNV are96restricted to greenhouse-grown cucumbers and are dependent ontransmission by 0. radicale. No vector has been identified forTBSV although it is assumed that this virus is transmitted by asoil-borne microorganism, perhaps a fungus (Campbell, 1968). Itis possible that biological differences between CNV and TBSV, suchas transmission by 0. radicale, may be a consequence of thedivergence of their coat protein genes and resulting coat proteins.111.1.5 Research objectiveThe objective of this study was to determine if CNV coatprotein is the specific determinant of CNV transmission by O.radicale. This was achieved by exchanging the coat protein genesof full-length clones of CNV and TBSV-Ch. Synthetic RNAtranscribed from the chimeric mutants was infectious on N.clevelandii and virus particles were produced and purified. Thevirions were then used in a transmission experiment with zoosporesof 0. radicale. Evidence is presented that the CNV coat proteinis the only CNV-encoded protein necessary to confertransmissibility by 0. radicale.97111.2 MATERIALS AND METHODS111.2.1 Production of chimeric CNV/TBSV cDNA clones111.2.1.1 Sub-cloning of the MTV and TBSV coat protein genesTo produce a CNV sub-clone encoding the CNV coat protein gene,CNV wild-type DNA (pK2/M5) was digested with EcoRI (nt 2156) andKpnI (nt 3904). The resulting 1748 by EcoRI -KpnI fragment,encompassing the final one-fifth of the p92 gene, the entire coatprotein gene, and the 5' terminal region of the p20/21 genes, waspurified from a 1% agarose gel using Gene-Clean Tm and ligated intoEcoRI/KpnI-digested Bluescript. The ligated DNA was used totransform E. coli strain DH5a. Plasmid DNA was purified from thetransformants by alkaline lysis, and screened for CNV coat proteinclones by restriction enzyme mapping. Clone pCNVCP (Fig. 3.1) wasselected for further experimentation.pTBSV101 (Hearne et al., 1990; generously provided by Dr. T.J.Morris, University of Nebraska) is a full-length TBSV cDNA clonefrom which infectious T7 RNA polymerase transcripts can besynthesized. To sub-clone the TBSV coat protein gene, pTBSV101 DNAwas digested with BamHI (nt 2439) and EcoRI (nt 4033). Theresulting 1594 by BamHI -EcoRI fragment encompassed the 3' terminalregion of the TBSV p92 gene, the entire TBSV coat protein gene, andthe 5' terminal region of the TBSV p19/p22 genes. The 1594 byBamHI -EcoRI fragment was gel purified, as above, and ligated intoBamHI/EcoRI-digested Bluescript. Plasmid DNA was purified fromtransformed DH5a, and one clone, pTBSVCP (Fig. 3.1), was selectedafter restriction enzyme mapping.9899pMM200pMM100B HpTBSV101E H WI^I^II pK2/M5pCNVCPligation with EcoRt/Kpnl digested BluescriptH N<11.,HinclI/Ncol digestiongel purificationhEBamHI/Ncol digestiongel purificationB H NEcoRVKpni digestionE H NKI^I^- H N N<^E HI gommwmVBamHI/EcoRI digestionB H NEHincll digestiongel purificationB-1 H'3 ArpTBSVCPligation with BamHI/EcoRI digested BluescriptB H NEreplacement of pCNVCP Hindi/Nicol fragment withHincIVNcol fragment from pTBSVCPE H N<pCTCPEcoRI/Ncol digestiongel purificationE H N<=V Alreplacement of pTBSVCP Hincll/Ncol fragment withHindi/Nicol fragment from pCNVCPB H NEBam1-11/EcoRI digestiongel purificationB H NEpTCCPligation of EcoRI/Ncol fragment of pCTCPwith EcoRI/Ncol digested pK2/M5E H N<ligation of BamHI/EcoRI fragment of pTCCPwith BamHI/EcoRI digested pTBSV101B H NEFig. 3.1. Schematic diagram indicating the sequence of events leading to thecreation of the CNV/TBSV chimeric cDNA clones, pMM100 and pMM200. Regions frompTBSV101 and its subclones are hatched for ease of recognition. Restrictionendonuclease recognition sites are abbreviated as follows: B = BamHI, E = EcoRI.- HincII, K^N - NucJI. Production of the chimeric clonesBoth CNV and TBSV encode a HincII site and a NcoI site in thecorresponding regions of their respective nucleotide and amino acidsequences. The HincII sites occur near the 3' ends of the p92 ORFs(CNV nucleotide 2564; TBSV nucleotide 2577), and the NcoI sitesoccur in the 5' regions of the CNV and TBSV bicistronic ORFs (CNVnucleotide 3830; TBSV nucleotide 3885)(see Fig. 3.2, section111.3.1). The CNV HincII and NcoI sites share the same frameswith the HincII and NcoI sites of TBSV.pCNVCP was digested with HincII (nt 2564) and NcoI (nt 3830)to yield 1266 by HincII -NcoI and 3389 by NcoI -HincII fragments.pTBSVCP was first digested with BamHI (nt 2439) and NcoI (nt 3885)to yield a 1446 by BamHI -NcoI fragment and a 3088 by NcoI -BamHIfragment. After purification from a 1% agarose gel, the 1446 byBamHI -NcoI fragment was subsequently digested with HincII (nt 2577)to produce two fragments of 138 by (BamHI -HincII) and 1308 by(HincII-NcoI). All fragments from the pCNVCP and pTBSVCPdigestions were individually gel-purified using a QIAEX agarose gelextraction kit (QIAGEN) according to the manufacturer'srecommendations. To produce a CNV clone encoding the TBSV coatprotein gene, the TBSV 1308 by HincII -NcoI fragment was ligatedwith the CNV 3389 by NcoI -HincII fragment, to produce clone pCTCP(Fig. 3.1). To produce TBSV clone pTCCP encoding the CNV coatprotein gene (Fig. 3.1), the CNV 1266 by HincII -NcoI fragment wasligated with the TBSV 138 by BamHI -NcoI and 3088 by NcoI -BamHIfragments. Ligated DNA was used to transform DH5a and plasmid DNA100was screened by HincII, NcoI, KpnI, and BglII digestions.To replace the CNV coat protein gene with the TBSV coatprotein gene in the CNV wild-type cDNA clone, pCTCP was digestedwith EcoRI (nt 2156) and NcoI (nt 3830) and the 1716 by fragmentwas ligated into the 6212 by fragment of EcoRI -NcoI digested wild-type CNV plasmid DNA to produce clone pMM100 (Fig. 3.1).Similarly, the CNV coat protein gene was used to replace the TBSVcoat protein gene in pTBSV101 after BamHI (nt 2439) and EcoRI (nt4033) digestion of both pTCCP and pTBSV101. The 1594 by fragmentof pTCCP was ligated into the large fragment of pTBSV101 to produceclone pMM200 (Fig. 3.1). pMM100, in addition to the TBSV coatprotein gene, contains the coding sequences for the terminal 13amino acids of the TBSV p92 gene, the 29 nucleotide long non-translated region between the TBSV p92 and coat protein genes andthe amino terminal region of the TBSV p19/22 (including the 5' non-translated region between the coat protein and p19/22 genes and theputative p19/22 subgenomic promotor). pMM200 contains the codingsequences for the terminal 13 amino acids of the CNV p92 gene andthe 19 nucleotide-long non-translated region between the CNV p92and coat protein genes, the amino-terminal region of the CNV p20/21genes (including the 5' non-translated region between the coatprotein and p20/21 genes and the putative p20/21 subgenomicpromotor), in addition to the CNV coat protein gene. Clones werescreened by HincII digestion followed by Bg/II and KpnI digestionto confirm their chimeric structure.pMM100 and pMM200 were linearized with Smal and in vitro 101transcription, transcript-inoculation of N. clevelandii (replicatedin two independent experiments) and sap-inoculation of plants, wereconducted as described in sections 11.2.5 and Purification of p! 100 and pMM200 virionsThe first steps in the virus purification procedure below werethose of Tremaine et al. (1983). Five grams of pMM100- or pMM200-infected N. clevelandii tissue were homogenized in 2 volumes of 0.1M sodium acetate, pH 5.0, 5 mM 2-mercaptoethanol, and filteredthrough four layers of cheese cloth. Filtrates were left on icefor 10 min, then centrifuged at 10,000 x g for 15 min at 4°C. Thesupernatants were removed, amended with polyethylene glycol to 8%(8,000 MW, Sigma), stirred for 1 hour at 4°C, and then centrifugedas above but for 20 min. In a modification of this purificationmethod, the pellets were resuspended in 10 mM sodium phosphatebuffer, pH 7.0. and adjusted to a density of 1.33 g/cm 3 with CsC1before transfer to 13.5 ml polyallomer Quick-seal Centrifuge Tubes(Beckman). The CsC1 gradients were centrifuged in a Ti 70 rotorat 42,000 rpm for 16 hours at 20°C. After centrifugation theopaque virus-containing bands were withdrawn from the gradientswith a sterile, 20 gauge disposable needle fitted to a 3 mldisposable syringe. The purified virus preparations were dialysedovernight against PBS (0.15 M NaCl, 10 mM sodium phosphate, pH 7.2)and yields were determined spectrophotometrically (a 1 mg/mlsolution of CNV and TBSV has an A260 value of 4.5).102111.2.3 Fungal transmission of purified CNV wild-type, TBSV101,pMM100, and pMM200 virionsTo determine if virions purified from pMM100- and pMM200-infected N. clevelandii could be transmitted by the CNV vector 0.radicale, samples of purified virus (see 111.2.2 above) of CNVwild-type, pTBSV101, pMM100, and pMM200 were sent to Dr. R.N.Campbell, Department of Plant Pathology, University of California,Davis, California, who performed the transmission experimentsbecause fungal culture and transmission techniques were not yetdeveloped at the Vancouver Research Station. The chimeric viruseswere tested for in vitro acquisition and transmission by zoosporesessentially as described by Campbell et al. (1991). Pairs of 5 to7 day-old cucumber seedlings, which were germinated on papertowels, were transplanted into pots containing pasteurized whitesand. Five to seven days after transplantation, the cucumbers wereinoculated with fungus only, fungus + virus, or virus alone.Zoospore suspensions of O. radicale (1.1 x 10 5 zoospores/ml ofisolate SS196; Campbell et al., 1991) were prepared in tap water,and amended with virus (0.1 gg virus/ml zoospore suspension). Tenmillilitres of virus/zoospore inoculum, or of virus/water solutionwere added to each of two replicated pots of cucumber seedlings.Two pots were inoculated with fungus only or with water only. Onemillilitre of virus/water solution was assayed on C. quinoa tocheck the virus titre of the inoculum. The two replicated pots foreach treatment were incubated for 8 days in separate growthchambers at 24 C daytime/18 C nighttime with a 16 hour photoperiod.After incubation the cucumber root systems were washed free of qAnd 103in tap water and then ground in 5 ml of sodium phosphate buffer,pH 7.0 using a mortar and pestle. A portion from each treatmentwas retained by Dr. Campbell; and another portion was sent to theauthor by surface mail. No preservatives were added to the groundroot material prior to shipping. The ground root material wasassayed for successful vector transmission of virus by bio-assayon C. quinoa, a local lesion host for CNV, TBSV and the chimericviruses.111.2.4 Local lesion assay on C. quinoa For the local lesion assay, 50 gl of ground root materialfrom each treatment was amended with non-activated charcoal [usedto adsorb virus inactivators from cucumber (R.N. Campbell, personalcommunication)], and inoculated to each of two leaves of two C.quinoa plants. Local lesions were counted 3 dpi. The bio-assaywas also completed by Dr. Campbell who inoculated each of fourleaves of two C. quinoa plants with corresponding aliquots.111.2.5 ELISAOne hundred microlitres of each of the purified CNV wild-type,pTBSV101, pMM100, and pMM200 virion preparations (10 µg/ml) wasassayed by DAS-ELISA for the presence of either CNV or TBSV coatprotein (Clark and Adams, 1977). The two polyclonal antisera usedwere raised against CNV or TBSV virions (Johnston, 1989; J.H.Tremaine, Agriculture Canada Research Station, Vancouver, personalcommunication). Background A4 05 values were determined by taking104the mean of A405 readings from eight wells treated with PBS insteadof virus. A positive A405 value was conservatively set at ten timesgreater than background (Sutula et al., 1986). ELISA values arepresented as the mean from two replicate plates.105111.3 RESULTS111.3.1 Symptomatology and infectivity of pMM100 and pMM200 in N.clevelandii Synthetic RNA transcribed from each of two CNV/TBSV chimericclones, pMM100 and pMM200 (Fig. 3.2), was infectious and replicatedwell in transcript-inoculated N. clevelandii. These transcripts,and transcripts synthesized from wild-type CNV and pTBSV101 (afull-length TBSV cDNA clone) induced similar symptoms intranscript-inoculated N. clevelandii. Buff-tan local lesions wereproduced 4 to 7 dpi, followed by systemic spread of the viruses by8 to 12 dpi. Systemic infection culminated in a lethal necrosisin all transcript- or sap-inoculated N. clevelandii.Purified CNV wild-type, pTBSV101, pMM100 and pMM200 virionswere inoculated to each of two leaves of the following plants todetermine if the chimeric viruses induced symptoms distinct fromeach other or from wild-type CNV and/or TBSV: Chenopodium amaranticolor, cowpea, Cucumis sativus, Daturia stramonium,Gomphrena qlobosa, Nicotiana benthamiana, N. clevelandii, N.qlutinosa,and N. tabacum cv. White Burley. No distinctivesymptoms were produced on any of these hosts such thatsymptomatology could be used to differentiate any of the fourinocula used. Local lesions were produced on all hosts andsystemic infections were induced in N. clevelandii and N.benthamiana. Additionally, the specific infectivity of wild-typeCNV, pTBSV101, pMM100, and pMM200 virions was determined byinoculating two leaves of each of two C. quinoa plants with ten-106pMM100p33 p92 coat protein p20/21NIiA pMM200p33^p92^coat protein^p19/22Fig. 3.2. Schematic diagram illustrating the structure of theCNV/TBSV chimeric clones, pMM100 and pMM200. pMM100 is a full-length CNV cDNA clone in which 3'-terminus of the p92 gene, thecoat protein gene, and the 5'-terminal region of the p20/21 geneswere replaced with the corresponding sequences from TBSV. pMM200is a full-length TBSV cDNA clone in which the 3'-terminus of thep92 gene, the coat protein gene, and the 5' terminal regions of thep19/22 genes were replaced with the corresponding sequences fromCNV. TBSV sequences are hatched for ease in identification. H =N = NcoI.107fold dilutions of 1 mg/ml purified virus preparations. At a virusconcentration of 10' mg/ml an average of 18, 12, 26, and 24 locallesions/leaf were observed for wild-type CNV, pTBSV101, pMM100 andpMM200, respectively. These experiments demonstrated that thechimeric RNAs encapsidated within pMM100 and pMM200 virions wereinfectious and had specific infectivities similar to wild-type CNVand pTBSV101.111.3.2 Purification of pMM100 and pMM200 virionsVirions were readily purified from systemically-infectedleaves of pMM100 and pMM200 sap-inoculated N. clevelandii.Electron micrographs of negatively stained preparations of pMM100and pMM200 virions were compared with those purified from CNV wild-type- and pTBSV101- infected N. clevelandii (Fig. 3.3). There wereno apparent differences in the structure of the virions producedin wild-type CNV-, pTBSV101-, pMM100-, and pMM200- infected N.clevelandii.111.3.3 ELISATo determine if virions produced in pMM100- or pMM200-transcript-inoculated N. clevelandii were truly chimeric, purifiedvirions were assayed for the presence of CNV or TBSV coat proteinby DAS-ELISA (an A405 value 10 times greater than background wasconsidered positive). It has previously been determined that CNVand TBSV-Ch are serologically unrelated (Tremaine, 1970). The108Fig. 3.3 Electron micrographs of negatively stained virionspurified from N. clevelandii infected with A. CNV wild-type; B.pTBSV101; C. pMM100; D. pMM200. Bar = 100 nm. Electronmicrographs courtesy of Mr. F. Skelton, Vancouver Research Station.109Table 3.2 DAS-ELISA A405 1 values for purified CNV, pTBSV101, pMM100,and pMM200 virions using polyclonal antisera raised against CNV orTBSV virions.Treatment 2 CNV antiserum3 TBSV antiserum4CNV wild-type virions 0.529 0.003pTBSV101 virions 0.001 1.783pMM100 virions 0.002 1.670pMM200 virions 0.222 0.0011 A405 values were considered positive when ten times greater thanbackground.2 Purified virions were used at a concentration of 10 µg/ml.3 Background value for CNV antiserum was 0.0054 Background value for TBSV antiserum was 0.001110results presented in Table 3.2 show that CNV wild-type and pMM200virions reacted with CNV antiserum, but not with TBSV antiserum.Conversely, pTBSV101 and pMM100 virions reacted with TBSVantiserum, but not with CNV antiserum. These data, in conjunctionwith the electron microscopy results and the results described insection 111.3.1 demonstrate that the chimeric viruses resultingfrom the exchange of the CNV and TBSV coat protein genes areinfectious and are capable of producing virions with a morphologycharacteristic of the corresponding viruses.111.3.4 Fungus transmissibility of pMM100 and pMM200 virionsTo determine if the chimeric virions purified from pMM100- andpMM200- infected N. clevelandii could be transmitted by 0.radicale, pMM100, pMM200, CNV wild-type, and pTBSV101 virions,alone or in combination with 0. radicale zoospores, were used toinoculate the roots of cucumber seedlings (the transmissionexperiments were completed by Dr. R.N. Campbell, University ofCalifornia, Davis, California). After incubation for several daysthe roots of the inoculated cucumber plants were harvested andground in sodium phosphate buffer, pH 7.0. To determine whichviruses were transmitted to cucumber by 0. radicale, ground rootmaterial from each treatment was used in a local lesion assay onC. quinoa. The mean number of local lesions produced on eachinoculated C. ciuinoa by each virus treatment is recorded in Table3.3. CNV wild-type and pMM200 virions (chimeric TBSV/CNV RNAencapsidated in CNV coat protein) but not TBSV101 or pMM100 virions111(chimeric CNV/TBSV RNA in TBSV capsids) were successfullytransmitted to cucumber by the zoospores of 0. radicale. Thisindicates that the coat protein of CNV is specifically required forits transmission by 0. radicale.A greater number of local lesions was recorded for the CNVwild-type and pMM200 treatments in bio-assay 1 than bio-assay 2(Table 3.3). This was probably a consequence of 1) differences inbio-assay conditions (e.g. age of the C. cruinoa used in the assay,environmental conditions in the greenhouses); and 2) inactivationof some of the virus during transport of the ground leaf materialto the author. The root material was transported in sodiumphosphate buffer, pH 7.0. with no added preservative, and at roomtemperature.112Table 3.3. Bio-assay to determine the fungus transmissibility of the CNV/TBSV c imericviruses, pMM100 and pMM200.Treatment' Mean Number of Local Lesions/Plant 2Bio-assay 1 3 Bio-assay 2 4Plant 1^'Plant 2 Plant 1 Plant 2CNV wild-type + 0. radicale 83 19 39.5 3.5pTBSV101 + 0. radicale 0 0 0 0pMM100 + 0. radicale 0 0 0 0pMM200 + 0. radicale 92 42 8.5 70. radicale, no virus 0 0 0 0CNV wild-type, no fungus 0 0 0 0pTBSV101, no fungus 0 0 0 0pMM100, no fungus 0 0 0 0pMM200, no fungus 0 0 0 0' Treatments used to inoculate the root systems of cucumber seedlings. After incub tion,the root systems were harvested and ground in sodium phosphate buffer. The groun rootmaterial was then used to inoculate C. quinoa.2 The number of local lesions on C. quinoa were counted 3 dpi.3 Bio-assay 1 was completed by Dr. R.N. Campbell, U.C. Davis, California and the resul s arepresented with his permission. The number of local lesions presented is the mean nu er oflocal lesions on four leaves of one C. quinoa.4 Bio-assay 2 was completed by the author. The number of local lesions presented s themean number of local lesions on two leaves of one C. quinoa.111.4 DISCUSSIONIn vitro transcribed RNA produced from the CNV/TBSV chimericclones, pMM100 and pMM200, was highly infectious and producedbiologically active virions in N. clevelandii. No differences inthe phenotypic response of N. clevelandii to either clone wereobserved, and the symptoms produced were essentially the same asfor CNV wild-type and pTBSV101 transcripts which produceindistinguishable symptoms on N. clevelandii. Systemic infectionsconsistently occurred and were always lethal. Virions from pMM100and pMM200 sap-inoculated N. clevelandii were purified andcharacterised. There were no apparent differences in thestructural appearances of CNV wild-type, pTBSV101, pMM100 andpMM200 virions. When tested by DAS-ELISA, pMM100 virions reactedwith TBSV, but not CNV, polyclonal antiserum, and pMM200 virionsreacted with CNV, but not TBSV, polyclonal antiserum. The electronmicrographs and ELISA results indicate that the CNV and TBSVreciprocal coat protein gene exchanges were successful; coatprotein production and RNA encapsidation were apparently notaffected by the chimeric nature of the pMM100 and pMM200 genomes.Transmission of CNV by 0. radicale was confirmed, as was theinability of 0. radicale to transmit TBSV-Ch.CNV wild-type and pMM200 virions were successfully transmittedto C. sativus by 0. radicale, while pTBSV101 and pMM100 virionswere not. pMM200 contains, in addition to the CNV coat proteingene in a full-length TBSV background, the 3' terminus of the CNVp92 gene, the non-translated region between p92 and the CNV coat114protein ORF, and the 5' terminal region of the CNV p21 ORF.Conversely, pMM100 contains, in addition to the TBSV coat proteingene in a full-length CNV background, the 3' terminus of the TBSVp92 gene, the non-translated region between p92 and the TBSV coatprotein ORF, and the 5' terminal region of the p22 ORF. ThatpMM200 virions were transmitted to cucumber by 0. radicale, whilepMM100 virions were not, suggests that the CNV coat protein is themajor viral determinant that governs recognition, acquisition, andtransmission of CNV by O. radicale. However, because geneticelements other than the coat protein of the donor virus werepresent in the chimeric constructs, a function for non-structuralproteins in the transmission of CNV cannot be completely ruled out.Transmission of plant viruses by their vectors involves athree-way interaction between the virus, the vector, and the hostplant. Additionally, environmental or other influences on any ofthese three components may also affect transmission. Transmissionof a virus, and subsequent host plant infection, depends on therecognition of the vector by the virus, acquisition of the virusby the vector, release of the virus in the host plant, andsuccessful replication of the virus in its host. For example,transmission of CNV and tobacco necrosis virus (TNV) by theirrespective vectors, O. radicale and O. brassicae, is determined inpart by virus/vector recognition. Zoospores of O. radicale adsorbed particles of CNV but not tobacco necrosis virus (TNV),while zoospores of O. brassicae, adsorbed TNV particles, but notCNV particles (Temmink et al., 1970). That transmission of115pTBSV101 and pMM100 virions by 0. radicale zoospores was notsuccessful may have resulted from: 1) a lack of recognition betweenthe viruses and the vector; 2) failure of the viruses to beadsorbed by 0. radicale zoospores; or 3) failure of release of theviruses in cucumber, had they been adsorbed by the zoospores. Itis also possible that acquisition and/or release of pTBSV101 orpMM100 virions occurred but was at a level too low to promote adetectable level of virus transmission. In vitro transmissionof plant viruses by their fungal vectors is characterized by a veryspecific association between the virus and its vector and isbelieved to be a result of the interaction of the viral capsid andthe plasmalemma of the vector zoospore (Adams, 1991). Thisselective association was demonstrated by Campbell et al. (1991)who found four cultures of 0. radicale to vary in their ability totransmit a variety of tombus- and carmo- viruses. All fourcultures transmitted MNSV and CNV, but only three transmittedcucumber leaf spot virus (CLSV) and the cucumber fruit streakstrain of CLSV. None of the cultures transmitted squash necrosisvirus, cucumber soil-borne carmovirus, or petunia asteroid mosaictombusvirus [which is serologically identical to TBSV-Ch (Koenigand Kunze, 1982)]. The results of Campbell et al. (1991) indicatethat some form of recognition phenomenon is involved in thetransmission of these viruses by 0. radicale zoospores, which ischaracteristic of in vitro acquisition of such viruses and theirsubsequent transmission.Fungal transmission of viruses shares many characteristics116with virus transmission by nematodes. The nematode-transmittedviruses (nepoviruses and tobraviruses) are thought to beselectively and specifically adsorbed at the retention sites on thenematode vector, and, as with tombusvirus transmission (see above),the specificity of nematode transmission may also be due todifference between populations of vector species (Taylor and Brown,1981). Nematode transmission of raspberry ringspot (Harrison etal., 1974; Hanada and Harrison, 1977) and tomato black ringnepoviruses (Randles et al., 1977) has been mapped to RNA 2 ofthese viruses, which contains the coat protein gene. Taylor andBrown (1981) speculate that transmission of these viruses by theirnematode vectors is dependent on the protein surface of the virusparticles. Robertson and Henry (1986) reported that particles ofarabis mosaic and strawberry latent ringspot nepoviruses wereobserved attached to carbohydrate zones which discontinuously linedthe odontophore and oesophagus of their vector Xiphenema diversicaudatum. Taylor (1990) speculates that virus retention inXiphenema and trichorid nematode vectors may involve theinteraction between carbohydrate moieties on the wall of thenematode's food canal and complementary lectin-like molecules onthe virus capsid. This may be analogous to the association of theCNV coat protein with the plasmalemma of 0. radicale zoospores (seebelow).The region of the CNV coat protein that specificallyrecognizes and associates with the plasmalemma of 0. radicale zoospores is not known, but the P domain is a probable candidate117for a number of reasons.^First, the viruses within thetombusvirus, carmovirus, and dianthovirus groups are soilborne, andthose that have known vectors (CNV, MNSV, RCNMV) are alltransmitted by 0. radicale (that not all tombus-, carmo- anddianthoviruses are transmitted by 0. radicale does not precludeother fungi from acting as vectors of these viruses). The Pdomain, which is unique to these groups of viruses, could play somerole as a determinant in their fungal transmission. Second, a dotmatrix comparison of the amino acid sequences of severaltombusvirus and carmovirus P domains revealed significant sequencesimilarity only between CNV and MNSV (Riviere et al., 1989), bothof which are transmitted naturally to cucurbits by 0. radicale.Third, P domain dimers project upwards from the capsid and wouldappear to be the first part of the capsid to be physicallypresented to the plasmalemma of the zoospore. Gibson and Argos(1990) proposed that the tombusvirus P domain was acquired by anancestral tombusvirus that captured a cellular gene specifying thisdomain. The topologies of the Class I (S domains) and Class II (Pdomains) "jellyrolls" (section 11.1.2) are very similar. Severalproteins that have Class I "jellyroll" topologies are known tofunction in ligand-binding and Gibson and Argos (1990) propose thatthe P domain of TBSV may have a similar function in binding to acellular receptor. The P domain of CNV could function as a bindingsite for a receptor on the zoospore plasmalemma, in addition to,or instead of, the function proposed by Gibson and Argos (1990) inintracellular recognition.118Attempts to specifically investigate the role of the CNV Pdomain in fungal transmission were unsuccessful. Initial CNVconstructs were designed such that XhoI restriction endonucleaserecognition sites were introduced to the borders of the P domain.Deletion of the P domain and subsequent inoculation of N.clevelandii with synthetic RNA from the P-domainless clone, PD(-),is described in Chapter II. PD(-) failed to produce virions ininfected tissue, and so fungal transmission experiments with a P-domainless virus were not possible.XhoI restriction endonuclease recognition sites were thenintroduced to the borders of the P domain of the coat protein geneof MNSV, so as to facilitate replacement of the CNV P domain withthat of MNSV. A chimeric cDNA clone, pM2, was produced in whichthe P domain sequence of CNV was replaced with the comparablesequence from MNSV (Fig. 3.4). pM2 transcripts were infectious inN. clevelandii but virions were not observed by transmissionelectron microscopy nor could coat protein be detected by DAS-ELISAusing CNV and MNSV polyclonal antisera. The fact that virusparticles were not produced in pM2-infected plants may have beendue to a flaw in the design of the chimeric construct. The 5' XhoIsite in XpK2/M5 was introduced at CNV nt 3417, immediately afterthe CNV hinge sequence, whereas the MNSV 5' XhoI site wasintroduced at MNSV nt 3619, six nucleotides downstream from thelast amino acid of the MNSV hinge. While the MNSV P domainfragment used to replace that from CNV was similar in size (CNV 316nt, MNSV 313 nt), the replacement of the CNV XhoI-XhoI P domain119sequence with that from MNSV resulted in a positional loss of threeamino acids that would normally follow the CNV hinge sequence (Fig3.4, B). The loss of these three amino acids may have compromisedthe intricate folding of the chimeric coat protein P domain so thatparticle formation was not possible. When it became apparent thatCNV/MNSV chimeric virus particles were not produced, furtherexperimentation with the CNV/MNSV chimeric clone, pM2, was notpursued.120R a SXhol nt3417Hill XpK2/M5Xhol nt hingeXhol, CNV nt 3417 Xhol, CNV nt 3733MNSV nt 3620^MNSV nt 3953—rs4^pM2/I R SBXpK2/M5 5' AQPTSPLLESLFRESAS 3'MNSV 5' PQPTAGMVCMLERLVSL 3'pM2 5' PQPTSPLLE---RLVSL 3'Fig. 3.4 Diagram of a CNV/MNSV chimeric clone. A. The coatprotein P domain coding sequence of the full-length CNV cDNA cloneXpK2/M5 was replaced with that from MNSV to produce the chimericCNV/MNSV cDNA clone pM2 (only the coat protein region is shown).Two XhoI sites were introduced into the MNSV P domain sequenceusing the polymerase chain reaction primed with mutagenicoligonucleotides. The MNSV P domain sequence in pM2 is hatched foreasier recognition. B. The amino acid sequences of XpK2/M5, MNSV(after mutagenesis), and pM2 in the region following the hinge areshown. The five amino acids of the hinge are underlined. Thedashed lines indicate the loss of three amino acids resulting fromligation of the XhoI-XhoI MNSV P domain cassette and XhoI digestedXpK2/M5. The two amino acids in the MNSV sequence in lonlci type arP those changed by PCR mutagenesis (from FD to LE).121CHAPTER IV. FUTURE DIRECTIONSCucumber necrosis virus has proven to be an interesting andvery useful model for the study of viral gene function. Mutationalanalysis of the CNV coat protein gene has provided insight into therole of the coat protein in the life cycle of CNV. The CNV coatprotein is not essential for cell-to-cell and systemic movement ofthe virus, induction of the necrotic response, or for mechanicaltransmission. Encapsidation of CNV RNA was found, however, toenhance both its stability in sap and the efficiency with which itis transmitted mechanically. It remains to be determined, however,whether the CNV coat protein has subtle effects on the regulation(timing or level) of expression of CNV genes. Additionally, theCNV coat protein may be the only viral protein required tofacilitate CNV transmission by zoospores of 0. radicale.The coat proteinless mutant CP(-) could prove to be useful instudying intra-plant movement of CNV. As the CNV coat protein isnot required for viral systemic movement the question is raised -how does this virus move in vivo and in what form? Evidence waspresented that RNA of another mutant, PD(-), was protected in somemanner from host ribonucleases, probably by a protein. Monoclonalantibodies raised against the non-structural proteins of CNV couldbe used in affinity chromatography and immunogold labelling todetermine which viral protein(s) associate with CP(-) RNA topotentiate its movement, and what form such a ribonucleoproteincomplex takes as it moves from cell-to-cell and systemically.Further reciprocal coat protein gene exchanges between CNV and122TBSV at the domain and sub-domain level could be used to determineif a specific region of the CNV coat protein is actuallyresponsible for association of CNV with 0. radicale zoospores andcould also provide a means for identifying viral receptors on thezoospore plasmalemma. To date no cellular receptors for plantviruses have been identified and the CNV/O. radicale system mayprovide insight into both the molecular basis for fungaltransmission of plant viruses and also the more fundamentalquestion of how intracellular recognition of plant viruses isachieved by their host plants.123BIBLIOGRAPHYAdams, M.J. 1988. Evidence for virus transmission byplasmodiophorid vectors. in "Developments in Applied Biology2: Viruses with Fungal Vectors". J.I. Cooper and M.J.C. Asher(eds.). Association of Applied Biologists, Wellesbourne, U.K.p.203-212.Adams, M.J. 1991. Transmission of plant viruses by fungi. Annalsof Applied Biology 118, 479-492.Adams, M.J., A.G. Swaby and P. Jones. 1988. Confirmation of thetransmission of barley yellow mosaic virus (BaYMV) by thefungus Polvmvxa qraminis. Annals of Applied Biology 112, 133-141.Ahlquist, P., V. Luckow, and P. Kaesberg. 1981. Complete nucleotidesequence of brome mosaic virus RNA 3. Journal of MolecularBiology 153, 23-38.Allison, R.F., C. Thompson, and P. Ahlquist. 1990. Regenerationof a functional RNA virus genome by recombination betweendeletion mutants, and requirements for cowpea chlorotic mottlevirus 3a and coat protein genes for systemic infection.Proceedings of the National Academy of Sciences 87, 1820-1824.Appiano, A., G. D'Agostino, M. Bassi, N. Barbieri, G. Viale and P.Dell'Orto. 1986. Origin and function of tomato bushy stuntvirus-induced inclusion bodies: II. Quantitative electronmicroscope autoradiography and immunogold cytochemistry.Journal of Ultrastructure and Molecular Structure Research 97,31-38.Armour, S.L., U. Melcher, T.P. Pirone, D.J. Lyttle, and R.C.Essenberg. 1983. Helper component for aphid transmissionencoded by region II of cauliflower mosaic virus DNA. Virology129, 25-30.Atabekov, J.G. and Y.L. Dorokhov. 1984. Plant virus-specifictransport function and resistance of plants to viruses.Advances in Virus Research 29, 313-364.Atabekov, J.G. and M.E. Taliansky. 1990. Expression of a plantvirus-coded transport function by different viral genomes.Advances in Virus Research 38, 201-248.Bailey, J.M. and N. Davidson. 1976. Methylmercury as a reversibledenaturing agent for agarose gel electrophoresis. Annals ofBiochemestry 70, 75-85.Barr, D.J.S. 1968. A new species of Olpidium parasitic on cucumberroots. Canadian Journal of Botany 46, 1087-1091. 124Barr, D.J.S. 1988. Zoosporic plant parasites as fungal vectors ofviruses: taxonomy and life cycles of species involved. in"Developments in Applied Biology 2: Viruses with FungalVectors". J.I. Cooper and M.J.C. Asher (eds.). Association ofApplied Biologists, Wellesbourne, U.K. p 123-138.Bol, J.F., L. van Vloten-Doting, and E.M.J. Jaspars. 1971. Afunctional equivalence of top component a RNA and coat proteinin the initiation of infection by alfalfa mosaic virus.Virology 46, 73-85.Bowen, R. and R.T. Plumb. 1979. The occurrence and effects of redclover necrotic mosaic virus in red clover (Trifoliumpratense). Annals of Applied Biology 91, 227-236.Boyko, V.P. and A.V. Karasev. 1992. Tombusvirus genome may encodethe sixth small protein near its 3' terminus. Virus Genes 6,143-148.Brault, V. and W.A. Miller. 1992. Translational frameshiftingmediated by a viral sequence in plant cells. Proceedings ofthe National Academy of Sciences 89, 2262-2266.Brunt, A.A. 1988. Labile rod shaped viruses transmitted byplasmodiophorid fungi - Furoviruses. in "Developments inApplied Biology 2: Viruses with Fungal Vectors". J.I. Cooperand M.J.C. Asher (eds.). Association of Applied Biologists,Wellesbourne, U.K. p. 3-18.Campbell, R.N. 1968. Transmission of tomato bushy stunt virus isunsuccessful with Olpidium. Plant Disease Reporter 52, 379-380.Campbell, R.N. 1985. Longevity of Olpidium brassicae in air-drysoil and the persistence of the lettuce big vein agent.Canadian Journal of Botany 63, 2268-2289.Campbell, R.N. and P.R. Fry. 1966. The nature of the associationsbetween Olpidium brassicae and lettuce big-vein and tobacconecrosis viruses. Virology 29, 222-233.Campbell, R.N. and R.G. Grogan. 1964. Acquisition and transmissionof lettuce big vein virus by Olpidium brassicae.Phytopathology 54, 681-690.Campbell, R.N., H. Lecoq, C. Wipf-Scheibel, and S.T. Sim. 1991.Transmission of cucumber leaf spot virus by Olpidium radicale.Journal of General Virology 72, 3115-3119.Campbell, R.N. 1993. Persistence is not applicable to fungalvector-virus relationship. Phytopathology 83, 363-364.125Carrington, J.C., T.J. Morris, P.G. Stockley, and S.C. Harrison.1987. Structure and assembly of turnip crinkle virus. IV.Analysis of the coat protein gene and implications of thesubunit primary structure. Journal of Molecular Biology 194,265-276.Chen, B. and R.I.B. Francki. 1990. Cucumovirus transmission by theaphid Myzus persicae is determined solely by the propertiesof its coat protein. Journal of General Virology 71, 939-944.Citovsky, V. and P. Zambryski. 1991. How do plant virus nucleicacids move through intercellular connections? BioEssays 13,373-379.Citovsky, V., D. Knorr, and P. Zambryski. 1991. Gene I, a potentialcell-to-cell movement locus of cauliflower mosaic virus,encodes an RNA-binding protein. Proceedings of the NationalAcademy of Sciences, 88, 2476-2480.Citovsky, V., D. Knorr, G. Schuster, and P. Zambryski. 1990. TheP30 movement protein of tobacco mosaic virus is a single-strand acid binding protein. Cell 60, 637-647.Citovsky, V., M.L. Wong, A.L. Shaw, B.V. Venkataram Prasad and P.Zambryski. 1992. Visualization and characterization of tobaccomosaic virus movement protein binding to single-strandednucleic acids. Plant Cell 4, 397-411.Clark, M.F. and A.N. Adams. 1977. Characteristics of the microplatemethod of enzyme-linked immunosorbent assay for the detectionof plant viruses. Journal of General Virology 34, 475-483.Dalmay, T., L. Rubino, J. Burgyan, and M. Russo. 1992.Replication and movement of a coat protein mutant ofcymbidium ringspot tombusvirus. Molecular Plant-MicrobeInteractions 5, 379-383.Dawson, W.O., P. Bubrick, and G.L. Grantham. 1988. Modificationsof the tobacco mosaic virus coat protein gene affectingreplication, movement, and symptomatology. Phytopathology 78,783-789.Dias, H.F. 1970a. Transmission of cucumber necrosis virus byOlpidium cucurbitacearum. Virology 40, 828-839.Dias, H.F. 1970b. The relationship between cucumber necrosis virusand its vector Olpidium cucurbitacearum. Virology 42, 204-211.Dinesh-Kumar, S.P. and W.A. Miller. 1992. Control of translationof three genes from a single subgenomic RNA of barley yellowdwarf virus (BYDV). Phytopathology 82, 1071. 126Dolja, V.V. and E.V. Koonin. 1991. Phylogeny of capsid proteins ofsmall icosahedral RNA plant viruses. Journal of GeneralVirology 72, 1481-1486.Dougherty, W.G., J.C. Carrington, S.M. Cary, and T.D. Parks. 1988.Biochemical and mutational analysis of a plant viruspolyprotein cleavage site. EMBO Journal 7, 1281-1287.Finnen, R.L. and D.M. Rochon. 1993. Sequence and structure ofdefective interfering RNAs associated with cucumber necrosisvirus infections. Journal of General Virology (in press).French, R. and P. Alhquist. 1988. Characterization and engineeringof sequences controlling in vivo synthesis of brome mosaicvirus subgenomic RNA. J. Virology 62, 2411-2420.Gallitelli, D. and R. Hull. 1985. Characterisation of satelliteRNAs associated with the tomato bushy stunt virus and fiveother definitive tombusviruses. Journal of General Virology66, 1533-1543.Gardiner, W.E., G. Sunter, L. Brand, J.S. Elmer, S.G. Rogers, andD.M. Bisaro. 1988. Genetic analysis of tomato golden mosaicvirus: The coat protein is not required for systemic spreador symptom development. EMBO Journal 7, 899-904.Gibson, T.J. and P. Argos. 1990. Protruding domain of tomato bushystunt virus coat protein is a hitherto unrecognized class ofjellyroll conformation. Journal of Molecular Biology 212, 7-9.Goldbach, R. 1987. Genome similarities between plant and animal RNAviruses. Microbiological Sciences 4, 197-202.Gorbalenya, A.E. and E.V. Koonin. 1989. Viral proteins containingthe purine NTP-binding sequence pattern. Nucleic AcidsResearch 17, 8413-8440.Gorbalenya, A.E., E.V. Koonin, A.P. Donchenko and V.M. Blinov.1988. A conserved NTP-motif in putative helicases. Nature 333,22.Greico, P. and D. Gallitelli. 1990. Nucleotide sequence of the 3'-terminal region of artichoke mottle crinkle tombusvirus RNA.Nucleic Acids Research 18, 1300.Greico, F., J. Burgyan and M. Russo. 1989. The nucleotide sequenceof cymbidium ringspot virus RNA. Nucleic Acids Research 17,6383.127Habili, N. and R.H. Symons. 1989. Evolutionary relationship betweenluteoviruses and other RNA plant viruses based on sequencemotifs in their putative RNA polymerases and nucleic acidhelicases. Nucleic Acids Research 17, 9543-9555.Hacker, D.L., I.T.D. Petty, N. Wei, and T.J. Morris. 1992. Turnipcrinkle virus genes required for RNA replication and virusmovement. Virology 186, 1-8.Hamilton, R.I. 1980. Defences triggered by previous invaders:viruses. in "Plant Diseases, an Advanced Treatise". J.G.Horsfall and E.B. Cowling (eds.). Academic Press, New York.p. 279-303.Hanada, K. and B.D. Harrison. 1977. Effects of virus genotype andtemperature on seed transmission of nepoviruses. Annals ofApplied Biology 85, 79-92.Harrison, B.D. and A.F. Murant. 1977. Nematode transmissibility ofpseudo-recombinant isolates of tomato black ring virus. Annalsof Applied Biology 86, 209-212.Harrison, B.D., A.F. Murant, M.A. Mayo, and I.M. Roberts. 1974.Distribution of determinants for symptom production, hostrange and nematode transmissibility between two RNA componentsof raspberry ringspot virus. Journal of General Virology 22,233-247.Harrison, S.C. 1984. Multiple modes of subunit association in thestructures of simple spherical viruses. Trends in BiochemicalSciences, 345-351.Harrison, S.C. 1989. Common features in the design of small RNAviruses. in "Concepts in Viral Pathogenesis III". M.F.A.Oldstone and S.A. Notkins (eds.). Springer-Verlag, New York.p. 3-19.Harrison, S.C., A.J. Olson, C.E. Schutt, F.K. Winkler, and G.Bricogne. 1978. Tomato bushy stunt virus at 2.9 A resolution.Nature 276, 368-373.Hearne, P.Q., D.A. Knorr, B.I. Hillman, and T.J. Morris. 1990. Thecomplete genome structure and synthesis of infectious RNA fromclones of tomato bushy stunt virus. Virology 177, 141-151.Heaton, L.A., T.C. Lee, N. Wei and T.J. Morris. 1991. Pointmutations in the turnip crinkle virus capsid protein affectthe symptoms expressed by Nicotiana clevelandii. Virology 183,143-150.128Hemenway, C., R.X. Fang, W.K. Kaniewski, N.H. Chua, N.E. Turner.1988. Analysis of the mechanism of protection in transgenicplants expressing the potato virus X coat protein or itsantisense RNA. EMBO Journal 7, 1273-1280.Hillman, B.I., J.C. Carrington and T.J. Morris. 1987. A defectiveinterfering RNA that contains a mosaic of a plant virusgenome. Cell 51, 427-433.Hiruki, C. 1965. Transmission of tobacco stunt virus by Olpdiumbrassciae. Virology 25, 541-549.Hiruki, C. 1987. Recovery and identification of tobacco stunt virusfrom air-dried resting spores of Olpidium brassicae. PlantPathology 36, 224-228.Hodgman, T.C. 1988. A new superfamily of replicative proteins.Nature 333, 22-23.Hodgman, T.C. 1986. An amino acid sequence motif linking viral DNApolymerases and plant virus proteins involved in RNAreplication. Nucleic Acids Research 14, 6769.Hogle, J.M., A. Maeda, and S.C. Harrison. 1986. Structure andassembly of turnip crinkle virus. I. X-ray crystallographicstructure analysis at 3.2 A resolution. Journal of MolecularBiology 191, 625-638.Hopper, P., S.C. Harrison, and R.T. Sauer. 1984. Structure oftomato bushy stunt virus V. Coat protein sequencedetermination and its structural implications. Journal ofMolecular Biology 177, 701-713.Johnston, J.C. 1989. In vitro translation of cucumber necrosisvirus RNA. M.Sc. Thesis. University of British Columbia. 82pp.Johnston, J.C. and D.M. Rochon. 1990. Translation of cucumbernecrosis virus RNA in vitro. Journal of General Virology 71,2233-2241.Jones, R.A.C. and B.D. Harrison. 1969. The behaviour of potato mop-top virus in soil, and evidence for its transmission bySponcrospora subterranea (Wallr.)Lagerh. Annals of AppliedBiology 63, 1-17.Jones, R.W., A.O. Jackson, and T.J. Morris. 1990. Defective-interfering RNAs and elevated temperatures inhibit replicationof tomato bushy stunt virus in inoculated protoplasts.Virology 176, 539-545.129Jupin, I., L. Quillet, U. Niesbach-Klosgen, S. Bouzoubaa, K.Richards, H. Guilley, and G. Jonard. 1990. Infectioussynthetic transcripts of beet necrotic yellow vein virus RNAsand their use in investigating structure-functionrelationships. in "Viral Genes and Plant Pathogenesis". T.P.Pirone and J.G. Shaw (eds.). Springer-Verlag, New York. p 187-204.Kamer, G. and P. Argos. 1984. Primary structural comparison of RNA-dependent polymerases from plant, animal, and bacterialviruses. Nucleic Acids Research 12, 7269-7282.Karling, J.S. 1968. The Plasmodiophorales. Hafner Publishing Co.London. pp.256.Kim, K.S. and K.W. Lee. 1992. Geminivirus-induced macrotubules andtheir suggested role in cell-to-cell movement. Phytopathology82, 664-669.Kirkegaard, K. and D. Baltimore. 1986. The mechanism of RNArecombination in poliovirus. Cell 47, 433-443.Kozak, M. 1978. How do eukaryotic ribosomes select initiationregions in messenger RNA? Cell 15, 1109-1123.Kozak, M. 1984. Selection of initiation sites by eukaryoticribosomes: effect of inserting AUG triplets upstream from thecoding sequence for preproinsulin. Nucleic Acids Research 12,3873-3893.Kozak, M. 1986a. Point mutations define a sequence flanking the AUGinitiator codon that modulates translation by eukaryoticribosomes. Cell 44, 283-292.Kozak, M. 1986b. Regulation of protein synthesis in virus-infectedanimal cells. Advances in Virus Research 31, 229-292.Kozak, M. 1989. The scanning model for translation: an update.Journal of Cell Biology 108, 229-241.Krausslich, H-G and E. Wimmer. 1988. Viral proteinases. AnnualReview of Biochemistry 57, 701-54.Kunkel, T.A., J.D. Roberts, and R.A. Zakour. 1987. Rapid andefficient site-specific mutagenesis without phenotypeselection. Methods in Enzymology. 154, 367-382.Laakso, M.M. and L.A. Heaton. 1992. Turnip crinkle virus coatprotein is translated in vitro from swollen virions.Phytopathology 82, 1177.130Laemmli, U.K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227, 680-685.Lai, M.M.C. 1992. RNA recombination in animal and plant viruses.Microbiological Reviews 56, 61-79.Lange, L. and V. Inunza. 1977. Root inhabiting Olpidium species:the 0. radicale complex. Transactions of the BritishMycological Society 69, 37-384.Lazarowitz, S.G., A.J. Pinder, V.D. Damsteegt, and S.G. Rogers.1989. Maize streak virus genes essential for systemic spreadand symptom development. EMBO Journal 8, 1023-1032.Li, X.H., L.A. Heaton, T.J. Morris, and A.E. Simon. 1989. Turnipcrinkle virus defective interfering RNAs intensify viralsymptoms and are generated de novo. Proceedings of theNational Academy of Sciences 86, 9173-9777.Linstead, P.J., G.J. Hills, K.A. Plaskitt, I.G. Wilson, C.L.Harker,and A.J. Maule. 1988. The subcellular location of the gene1 product of cauliflower mosaic virus is consistent with afunction associated with virus spread. Journal of GeneralVirology 69, 1809-1818.Loesch-Fries, L.S., D. Merlo, T. Zinnen, L. Burhop, K. Hill, K.Krahn, N. Jarvis, S. Nelson, and E. Halk. 1987. Expression ofalfalfa mosaic virus RNA4 in transgenic plants confersresistance. EMBO Journal 6, 1845-1852.Lutke, H.A., K.C. Chow, F.S. Mickel, K.A. Moss, H.F. Kern, and G.A.Scheele. 1987. Selection of AUG initiation codons differs inplants and animals. EMBO Journal 6, 43-48.Martelli, G.P., D. Gallitelli and M. Russo. 1988. Tombusviruses.in "The Plant Viruses, Volume 3". R. Koenig (ed.). PlenumPress, New York. p 13-72.Matthews, R.E.F. 1991. Plant Virology, 3rd edition. Academic Press,London. 835 pp.Maule, A.J. 1991. Virus movement in infected plants. CriticalReviews in Plant Sciences 9, 457-473.Mayo, M.A., H. Barker, and B.D. Harrison. 1982. Specificity andproperties of the genome-linked proteins of nepoviruses.Journal of General Virology 59, 149-162.McKeen, C.D. 1959. Cucumber necrosis virus. Canadian Journal ofBotany 37, 913-925.131Miller, W.A., P.M. Waterhouse, W.L. Gerlach. 1988. Sequence andorganization of barley yellow dwarf virus genomic RNA. NucleicAcids Research 16, 6097-6111.Nassuth, A. and J.F. Bol. 1983. Altered balance of the synthesisof plus- and minus-strand RNAs induced by RNA 1 and 2 ofalfalfa mosaic virus in the absence of RNA3. Virology 124, 75-85Neeleman, L., A.C. van der Kuyl and J.F. Bol. 1991. Role of alfalfamosaic virus coat protein gene in symptom formation. Virology181, 687-693.Nishiguchi, M., S. Kikuchi, Y. Kiho, T. Ohno, T. Meshi, and Y.Okada. 1985. Molecular basis of plant viral virulence; thecomplete nucleotide sequence of an attenuated strain oftobacco mosaic virus. Nucleic Acids Research 13, 5585-5590.Pelletier, J. and N. Sonenberg. 1988. Internal initiation oftranslation of eukaryotic mRNA directed by a sequence derivedfrom poliovirus RNA. Nature 334, 320-325.Petty, I.T.D., and A.O. Jackson. 1990. Mutational analysis ofbarley stripe mosaic virus RNA Z. Virology 170, 712-718.Pirone, T.P. 1981. Efficiency and selectivity of the helper-component-mediated aphid transmission of purified potyviruses .Phytopathology 71, 922-924.Powell Abel, P., R.S. Nelson, B. De, N. Hoffman, S.G. Rogers, R.T.Fraley, and R.N. Beachy. 1986. Delay of disease developmentin transgenic plants that express the tobacco mosaic viruscoat protein gene. Science 232, 738-743.Randles, J.W., B.D. Harrison, A.F. Murant, and M.A. Mayo. 1977.Packaging and biological activity of the two essential RNAspecies of tomato black ring virus. Journal of GeneralVirology 36, 187-193.Rao, A.S. and M.K. Brakke. 1969. Relation of soil-borne wheatmosaic virus and its fungal vector Polymyxa graminis.Phytopathology 59, 581-587.Reed, K.C. and D.A. Mann. 1985. Rapid transfer of DNA from agarosegels to nylon membranes. Nucleic Acids Research 13, 7207-7221.Resende, R. de 0, P. de Hann, E. van Rossen, A.C. de Avila, R.Goldbach, and D. Peters. 1992. Defective interfering L RNAsegments of tomato spotted wilt virus retain both virus genometermini and have extensive internal deletions. Journal ofGeneral Virology 73, 2509-2516.132Rigby, P., W.J. Dieckmann, M. Rhodes, and P. Berg. 1977. Labellingdeoxyribonucleic acid to high specific activity in vitro bynick translation with DNA polymerase I. Journal of MolecularBiology 113, 237-251.Riviere, C.J. and D.M. Rochon. 1990. Nucleotide sequence andgenomic organization of melon necrotic spot virus. Journal ofGeneral Virology 69, 395-400.Riviere, C.J., J. Pot, J.H. Tremaine, and D.M. Rochon. 1989. Thecoat protein of melon necrotic spot carmovirus is more similarto those of tombusviruses than those of carmoviruses. Journalof General Virology 70, 3033-3042.Robertson, W.M. and C.E. Henry. 1986. An association ofcarbohydrates with particles of arabis mosaic retained withinXiphenema diversicaudatum. Annals of Applied Biology 109,299-305.Rochon, D.M. 1991. Rapid de novo generation of defectiveinterfering RNA by cucumber necrosis virus mutants that do notexpress the 20-kDa nonstructural protein. Proceedings of theNational Academy of Sciences 88, 11153-11157.Rochon, D.M. and J.C. Johnston. 1991. Infectious transcripts fromcloned cucumber necrosis virus cDNA: evidence for abifunctional subgenomic mRNA. Virology 181, 656-665.Rochon, D. and J.H. Tremaine. 1988. Cucumber necrosis virus is amember of the Tombusvirus group. Journal of General Virology69, 395-400.Rochon, D.M. and J.H. Tremaine. 1989. Complete nucleotide sequenceof the cucumber necrotis virus genome. Virology 169, 251-259.Rochon, D.M., J.C. Johnston, and C.J. Riviere. 1991. Molecularanalysis of the cucumber necrosis virus genome. CanadianJournal of Plant Pathology 13, 142-154.Romero, J., Q. Huang, J. Pogany, and J.J. Bujarski. 1993.Characterization of defective interfering RNA components thatincrease symptom severity of broad bean mottle virusinfections. Virology, 194, 576-584.Rossman, M.G. and J.E. Johnson. 1989. Icosahedral RNA virusstructure. Annual Review of Biochemistry. 58, 533-573.Roux, L., A.E. Simon, and J.H. Holland. 1991. Effects of defectiveinterfering viruses on virus replication and pathogenesis invitro and in vivo. Advances in Virus Research 40, 181-211.133Rubino, L., J. Burgyan, F. Greico, and M. Russo. 1990. Sequenceanalysis of cymbidium ringspot virus satellite and defectiveinterfering RNAs. Journal of General Virology 71, 1655-1660.Sacher, R., and P. Ahlquist. 1989. Effects of deletions in the N-terminal basic arm of brome mosaic virus coat protein on RNApackaging and systemic infection. Journal of Virology 63,4545-4552.Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. MolecularCloning, a Laboratory Manual (2nd ed.). Cold Spring HarbourPress, Cold Spring Harbour.Schlosser, E. 1988. Epidemiology and management of Polymyxa betae and beet necrotic yellow vein virus. in "Developments inApplied Biology 2: Viruses with Fungal Vectors". J.I. Cooperand M.J.C. Asher (eds.). Association of Applied Biologists,Wellesbourne, U.K. p.281-292.Scholthof, H.B., T.J. Morris and A.O. Jackson. 1993. The capsidprotein gene of tomato bushy stunt virus is dispensable forsystemic movement and can be replaced for localized expressionof foreign genes. Molecular Plant Microbe Interactions 6, 309-322Shintaku, M. and P. Palukaitis. 1990. Genetic mapping of cucumbermosaic virus. in "Viral Genes and Plant Pathogenesis"T.P.Pirone and J.G. Shaw (eds.). Springer-Verlag, New York.p. 156-164.Slykhuis, J.T. and D.J.S. Barr. 1978. Confirmation of Polvmyxa graminis as a vector of wheat spindle streak mosaic virus.Phytopathology 68, 639-643.Stanley, J. and R. Townsend. 1985. Characterization of DNA formsassociated with cassava latent virus infection. Nucleic AcidsResearch 13, 2189-2206.Stanley, J. and R. Townsend. 1986. Infectious mutants of cassavalatent virus generated in vivo from intact recombinant DNAclones containing single copies of the genome. Nucleic AcidsResearch 14, 5981-5998.Stobbs, L.W., G.W. Cross and M.S. Manocha. 1982. Specificity andmethods of transmission of cucumber necrosis virus by Olpidiumradicale zoospores. Canadian Journal of Plant Pathology 4,134-142.Sutula, C.L., J.M. Gillett, S.M. Morrissey, and D.C. Ramsdell.1986. Interpreting ELISA data and establishing the positive-negative threshold. Plant Disease 70, 722-726.134Suzuki, M., S. Kuwata, J. Kataoka, C. Masuta, N. Nitta and Y.Takanami. 1991. Functional analysis of deletion mutants ofcucumber mosaic virus RNA3 using an in vitro transcriptionsystem. Virology 183, 106-113.Taylor, C.E. 1990. Nematode interactions with other pathogens.Annals of Applied Biology 116, 405-416.Taylor, C.E. and D.J.F. Brown. 1981. Nematode-virus interactions.in "Plant Parasitic Nematodes, Volume III". B.M. Zuckerman andR.A. Rohde (eds.). Academic Press, London. p. 281-301.Teakle, D.S. 1983. Zoosporic fungi and viruses: double trouble. in"Zoosporic Plant Pathogens". S.T. Buczaki (ed.) AcademicPress, London. p 233-248.Teakle, D.S. and A.H. Gold. 1963. Further studies of Olpidium asa vector of tobacco necrosis virus. Virology 19, 310-315.Temmink, J.H.M., R.N. Campbell, and P.R. Smith. 1970. Specificityand site of in vitro acquisition of tobacco necrosis virus byzoospores of Olpidium brassicae. Journal of General Virology9, 201-213.Thouvenel, J.C. and C. Fauquet. 1980. Polymyxa graminis on newSorghum species in Africa. Plant Disease 64, 957-958.Tremaine, J.H. 1972. Purification and properties of cucumbernecrosis virus and a small top component. Virology 48, 582-590.Tremaine, J.H., W.P. Ronald, and E.M. McGauley. 1983. Effects ofsodium dextran sulfate on some isometric plant viruses.Phytopathology 73, 1241-1246.van der Kuyl, A.C., L. Neeleman, and J.F. Bol. 1991. Deletionanalysis of cis- and trans-acting elements involved inreplication of alfalfa mosaic virus RNA 3 in vivo. Virology183, 687-694.van Lent, J. Wellink, and R. Goldbach. 1990. Evidence for theinvolvement of the 58K and 48K proteins in the intercellularmovement of cowpea mosaic virus. Journal of General Virology71, 219-223.van Lent, J., M. Storms, F. van der Meer, J. Wellink and R.Goldbach. 1991. Tubular structures involved in movement ofcowpea mosaic virus are also formed in infected cowpeaprotoplasts. Journal of General Virology 72, 2615-2623.van Loon, L.C. 1987. Disease induction by plant viruses. Advancesin Virus Research 33, 205-255.135Verver, J., 0. Le Gall, A. van Kammen and J. Wellink. 1991. Thesequence between nucleotides 161 and 512 of cowpea mosaicvirus M RNA is able to support internal initiation oftranslation in vitro. Journal of General Virology 72, 2339-2345.Viera, J. and J. Messing. 1987. Production of single-strandedplasmid DNA. Methods in Enzymology 153, 3-11.Webster, J. 1980. Introduction to Fungi (2nd ed). CambridgeUniversity Press, London. p.45-53.Wellink, J. and A.B. van Kammen, A.B. 1989. Cell-to-cell transportof cowpea mosaic virus requires both the 58K/48K proteins andthe capsid proteins. Journal of General Virology 70, 2279-2286.White, K.A., J.B. Bancroft, and G.A. Mackie. 1991. Defective RNAsof clover yellow mosaic virus encode nonstructural/coatprotein fusion products. Virology 183, 479-486.Wieczorek, A. and H. Sanfacon. 1993. Characterization and sub-cellular localization of tomato ringspot nepovirus putativemovement protein. Virology (in press).Xiong, Z. and S.A. Lommel. 1989. The complete nucleotide sequenceand genome organization of red clover necrotic mosaic virusRNA-1. Virology 171, 543-554.Xiong, Z., K.H. Kim, and S.A. Lommel. 1991. The role of the redclover necrotic mosaic virus capsid and cell-to-cell movementproteins in viral infection. Phytopathology 81, 1183.Xiong, Z., K.H. Kim, D. Giesman-Cookmeyer, and S.A. Lommel. 1993.The roles of the red clover necrotic mosaic virus capsid andcell-to-cell movement proteins in systemic infection. Virology192, 27-32.136


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items