TRANSCRIPTIONAL AND TRANSLATIONAL REGULATION OF A SUBGENOMIC MRNA OF CUCUMBER NECROSIS VIRUS
by
JULIE CATHERINE JOHNSTON
B.Sc, University of British Columbia, 1988 M.Sc, University of British Columbia, 1990
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
in
THE FACULTY OF GRADUATE STUDIES
(Department of Microbiology and Immunology)
We accept this thesis as conforming to the required standard
THE UNIVERSITY OF BRITISH COLUMBIA
December 1995
© Julie Catherine Johnston, 1995 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.
Department
The University of British Columbia Vancouver, Canada
Date X^"<£ . 2-Q , l^^tC
DE-6 (2/88) Abstract
Cucumber necrosis virus (CNV) is a spherical virus which encapsidates a small messenger sense RNA genome. During infection, CNV generates a 0.9 kb subgenomic mRNA which directs the synthesis of two distinct proteins, p20 and p21, from different nested open reading frames (ORFs). Sequences comprising the core promoter for the synthesis of the CNV 0.9 kb subgenomic mRNA were determined using deletion analysis and site-directed mutagenesis. The results indicated that the CNV 0.9 kb subgenomic mRNA core promoter lies within a region located 20 nucleotides upstream and 6 nucleotides downstream of the transcription initiation site and that nucleotides immediately surrounding the initiation site also regulate promoter activity.
Comparison of sequences within the core promoter region with the corresponding region in other tombusviruses revealed that the tombusvirus promoter shares a region of near complete identity in 14 of the 26 core promoter nucleotides. Similarities to other well studied plant and animal virus promoters or to other putative CNV promoters were not apparent. Expression of both CNV p20 and p21 from the 0.9 kb subgenomic mRNA represents one of the rare cases of production of two proteins from the same coding region of a single mRNA. In vitro translation of synthetic transcripts corresponding to the 0.9 kb subgenomic mRNA but containing point substitutions in the AUG codons for either p20 or p21 indicated that these proteins are indeed separately initiated from different nested ORFs. The regulation of the synthesis of these proteins was investigated through examining the effects of codon context and leader length on the efficiency of translation.
Nucleotide substitutions introduced into the -3 and +4 positions of the p21 AUG codon verified that purines in these positions are favored and demonstrated the similar contribution of the -3 and
+4 positions to the efficiency of initiation codon selection in plants. Further analyses also indicated that the codon context of the upstream p21 AUG codon affects expression from the downstream p20 AUG codon and that an increase in the length of the subgenomic mRNA leader decreases expression from the downstream site. These observations are in accordance with the
"Kozak rules" for accession of internal AUG codons by leaky ribosomal scanning and provide the first example of an effect of leader length on the efficiency of translation initiation in a plant
(viral) mRNA. Table of Contents
Abstract ii
Table of Contents iii
List of Tables viii
List of Figures ix
List of Abbreviations xi
Acknowledgments xiv
Chapter 1 Introduction 1
1.1 Positive strand RNA viruses 3
1.1.1 Genome structure and organization 4
1.1.2 Viral Proteins 5
1.1.3 Replication of genomic RNA 10
1.1.4 Generation of subgenomic mRNAs 14
1.1.5 Production of viral proteins 17
1.2 The Tombusvirus Group 21
1.2.1 Cucumber necrosis virus 23
1.3 Thesis Objectives , 28
Chapter 2 Materials and Methods 30
2.1 Plasmid construction 30
2.1.1 Construction of plasmids used to map the 0.9 kb subgenomic mRNA
promoter 28
2.1.2 Construct containing mutations flanking the 0.9 kb subgenomic
mRNA start site 32
2.1.3 Construction of plasmids for transient expression in protoplasts 33
2.1.4 Construction of plasmids to generate subgenomic-length templates
for in vitro translation 35 iv
2.1.5 CaMV 35S promoter based constructs to map the promoter for the
0.9 kb subgenomic mRNA 40
2.2 In vitro transcription 43
2.3 Transcript inoculation 44
2.4 Protoplast isolation and transfection 45
2.5 RNA extraction 46
2.6 Northern blot analysis 47
2.7 In vitro translation and SDS-PAGE 47
2.8 Determination of relative GUS activity 48
Chapter 3 Results 49
3.1 Analysis of CNV 0.9 kb subgenomic mRNA production 49
3.1.1 Kinetics of CNV subgenomic RNA production in protoplasts 49
3.2 Deletion analysis of the CNV 0.9 kb subgenomic mRNA promoter 51
3.2.1 Large scale deletion analysis of sequences 5' of the CNV 0.9 kb
subgenomic mRNA start site '. 51
3.2.2 Deletion analysis of the 5' border of the 0.9 kb subgenomic mRNA
promoter 55
3.2.3 Large scale deletion analysis of sequences 3' of the CNV 0.9 kb
subgenomic mRNA start site 57
3.2.4 Deletion analysis of the 3' border of the 0.9 kb subgenomic mRNA
promoter '. 60
3.3 Mutational analysis of the core promoter for the 0.9 kb subgenomic mRNA 60
3.3.1 Effect of mutations in the 0.9 kb subgenomic core promoter on RNA
accumulation in protoplasts 62
3.3.2 Effect of mutations in the core promoter on 0.9 kb subgenomic
mRNA production in plants 62 3.3.3 Isolation of 0.9 kb subgenomic mRNA promoter revertants from
plants 67
3.4 Characterization of a CNV 0.35 kb subgenomic RNA species 67
3.4.1 In vitro translation of wild type and mutant 0.35 kb subgenomic RNA
transcripts 69
3.4.2 Effect of mutations in the pX ORF on infectivity of CNV transcripts .... 70
3.5 Production of p20 and p21 from wild type and mutant 0.9 kb subgenomic RNA
transcripts 72
3.5.1 In vitro production of p20 and p21 from CNV AUG codon mutants 74
3.5.2 Effect of mutations in the start codons of p20 and p21 on infectivity 76
3.5.3 Accumulation of CNV p21 and p20 AUG codon mutants in
cucumber protoplasts 76
3.6 Investigations into the restoration of systemic movement by coat protein
deletion derivatives 78
3.6.1 Production of p41, p20 and p21 from coat protein deletion mutants 79
3.7 Analysis of translational regulation in the production of p20 and p21 81
3.7.1 Effect of mutations surrounding the AUG codon for p21 82
3.7.2 Effect of mutations surrounding the p21 AUG codon on initiation
from the downstream p20 initiation codon 86
3.7.3 Effect of codon context on relative production of p20 and p21
in vitro 89
3.7.4 Effect of leader length of the 0.9 kb subgenomic mRNA on
production of p20 and p21 89
3.8 Trans-complementation assay 92 ;er 4 Discussion 96
4.1 Delineation of the promoter for 0.9 kb subgenomic mRNA synthesis 96
4.1.1 The 0.9 kb subgenomic mRNA core promoter is located between
nucleotides -20 and +6 relative to the subgenomic start site 96
4.1.2 The 0.9 kb subgenomic mRNA promoter shares little homology with
ICR2-like sequences or other CNV putative cis -acting sequences 99
4.1.3 The 0.9 kb subgenomic mRNA promoter shares considerable
sequence similarity with the putative promoter region in other
tombusviruses :. 100
4.1.4 Nucleotides immediately surrounding the 0.9 kb subgenomic mRNA
start site regulate promoter activity 100
4.2 Characterization of the 0.35 kb subgenomic RNA 101
4.2.1 A third subgenomic RNA of 0.35 kb is generated during CNV
infection 101
4.2.2 0.35 kb subgenomic transcripts direct the synthesis of pX in vitro 102
4.2.3 Mutations in the pX ORF alter infectivity of CNV genomic
transcripts ..; : 102
4.3 Functional analysis of CNV proteins 103
4.3.1 CNV p21 is associated with viral cell-to-cell movement 103
4.3.2 CNV p20, p21 and p41 are dispensible for RNA accumulation in
protoplasts 104
4.3.3 CNV mutants lacking the coat protein coding region have the
potential to overexpress the p21 movement protein 105
4.4 Translation control of CNV p20 and p21 production 107
4.4.1 The 0.9 kb subgenomic mRNA is bifunctional 107
4.4.2 Efficient initiation codon selection requires purines in either the -3 or
+4 position 108 4.4.3 Accession of CNV p20 ORF is consistent with leaky ribosomal
scanning 110
4.4.4 Leader length of the 0.9 kb subgenomic mRNA contributes to
translation of p20 via leaky ribosomal scanning Ill
4.5 Concluding Remarks 114
References 116
Chapter 5 Appendix 135
5.1 The (3-glucuronidase (GUS) enzyme system 135
5.1.1 p-Nitrophenyl P-D-glucuronide (pNPG) substrate 135
5.1.2 Quantitative analysis of GUS activity 136
5.1.3 Determination of relative GUS activity from transfected protoplasts 136
5.1.4 The pAGUS-1 expression vector...: 137 viii List of Tables
Table 5.1 Spectrophotometric measurements of p-nitrophenol absorbance in protoplast
samples transfected with pCGUS constructs 138
Table 5.2 GUS activity computed from kinetic spectrophotometric measurement of p-
nitrophenol absorbances in Table 6.1 140
Table 5.3 GUS activity computed from three independent experiments 140
Table 5.4 Spectrophotometric measurement of p-nitrophenol absorbance in protoplast
samples transfected with pBGUS constructs 142
Table 5.5 GUS activity computed from kinetic spectrophotometric measurement of p-
nitrophenol absorbance in Table 6.4 143
Table 5.6 GUS activity computed from two independent experiments 143 List of Figures
Figure 1.1 Model for the generation of subgenomic mRNA by internal initiation of
transcription 16
Figure 1.2 Schematic representation of the organization and expression of the CNV
genome 25
Figure 3.1 Kinetics of the accumulation of CNV subgenomic RNAs in protoplasts 50
Figure 3.2 Description of deletion mutants used to analyze the 5' border of the CNV 0.9 kb
subgenomic mRNA 52
Figure 3.3 Accumulation of PD(-) and CP(-) 0.9 kb subgenomic mRNAs in cucumber
protoplasts 54
Figure 3.4 Deletion analysis of the 5' border of the CNV 0.9 kb subgenomic mRNA 56
Figure 3.5 Description of deletion mutants used to analyze the 3' border of the CNV 0.9 kb
subgenomic mRNA 58
Figure 3.6 Large scale deletion analysis of the sequences 3' of the CNV 0.9 kb subgenomic
mRNA start site 59
Figure 3.7 Deletion analysis of the 3' border of the CNV 0.9 kb subgenomic mRNA 61
Figure 3.8 Nucleotide sequence of the region surrounding the 0.9 kb subgenomic start site
in CNV WT RNA and original M5Bam mutant and revertant RNAs 63
Figure 3.9 Accumulation of WT and M5Bam 0.9 kb subgenomic mRNAs in cucumber
protoplasts 64
Figure 3.10 Comparisons of infections produced by CNV WT and M5Bam transcript RNA
and M5Bam passaged RNA 65
Figure 3.11 Effects of mutations surrounding the 0.9 kb subgenomic mRNA transcription
start site on subgenomic RNA levels in protoplasts and plants 66
Figure 3.12 Nucleotide sequence surrounding the putative translation initiation site of pX 68
Figure 3.13 In vitro translation of synthetic pX subgenomic-length transcripts 71
Figure 3.14 Nucleotide sequences surrounding the translation initiation sites for CNV p20
andp21 73 X
Figure 3.15 In vitro translation of natural and synthetic CNV subgenomic mRNAs
containing the p20 and p21 ORFs 75
Figure 3.16 Northern blot demonstrating replication of WT, M5215 and M5201 mutant
RNA in cucumber protoplasts 77
Figure 3.17 Characterization of WT, PD(-) and CP(-) subgenomic RNAs and their in vitro
translation products .'. 80
Figure 3.18 Diagrammatic representation of pCGUS constructs used to analyze nucleotides
which regulate p21 translation initiation 83
Figure 3.19 GUS activity directed by pCGUS construct series in protoplasts 84
Figure 3.20 Diagrammatic representation of pBGUS constructs used to analyze p20
expression '. 87
Figure 3.21 GUS activity directed by pBGUS construct series in protoplasts 88
Figure 3.22 In vitro translation of 0.9 kb subgenomic RNA transcripts containing mutations
downstream of the initiation codon for p20 90
Figure 3.23 In vitro translation of wild-type 0.9 kb subgenomic mRNA transcripts and
extended leader ANM2 subgenomic length mRNA transcripts 91
Figure 3.24 Diagrammatic representation of constructs generated for the purpose of
mapping the CNV 0.9 kb subgenomic mRNA promoter 93
Figure 3.25 Accumulation of CNV RNA from T7- and CaMV 35S promoter-based
constructs in protoplasts 95
Figure 4.1 Sequences surrounding the CNV 0.9 kb subgenomic promoter and comparison
with other putative promoters 97
Figure 4.2 Predicted secondary structure of the 5' untranslated leader and initial coding
region of CNV subgenomic length transcripts 112
Figure 5.1 Time course of GUS activity as determined by p-nitrophenol absorbance 139
Figure 5.2 Relative GUS activity directed by pCGUS construct series in three
experiments 141
Figure 5.3 Relative GUS activity directed by pBGUS constructs in two experiments 144 List of Abbreviations
A adenine
A angstrom a arm
A1MV alfalfa mosaic virus
AMCV artichoke mottle crinkle virus
ATP adenosine triphosphate
BMV brome mosaic virus
BRL Bethesda Research Laboratories
BYDV-PAV barley yellow dwarf virus serotype PAV
C cytosine
°C Celsius (degrees) ca. circa; approximately
CaMV cauliflower mosaic virus cDNA complementary DNA
CIP calf intestinal phosphatase
CMI cucumber media I
CNV cucumber necrosis virus
CP coat protein
CTP cytidine triphosphate
CymRSV cymbidium ringspot virus
DI RNA defective interfering RNA
DNA deoxyribonucleic acid dsRNA double stranded RNA
DTT dithiothreitol
EDTA ethylenediaminetetraacetic acid
EF elongation factor
E. coli Escherichia coli G guanine
GDD glycine-aspartate-aspartate
GTP guanosine triphosphate
GUS glucuronidase h hinge hr hour
ICR internal control region kb kilobase kDa kilodalton
LB Luria-Bertani (medium)
M molar mM millimolar
MES 2[N-morpholino]ethanesulphonic acid met methionine min minute mRNA messenger RNA m7G methyl 7 guanine
NAPS Nucleic Acid - Protein Service unit nm nanometer
NOS nopaline synthetase oligo oligonucleotide
ORF open reading frame
P protein
(P),PD protruding domain
PCR polymerase chain reaction
PEG polyethylene glycol pNPG para-nitrophenol glucuronide pol polymerase poly(A) polyadenylate RdRp RNA-dependent RNA polymerase
RNA ribonucleic acid rpm revolutions per minute
RT-PCR reverse transcriptase PCR
S sedimentation coefficient
(S) shell domain
SDS sodium dodecyl sulfate
SDS-PAGE SDS-polyacrylamide gel electrophoresis ssDNA single stranded DNA
T thymine
(T) triangulation number
TBSV tomato bushy stunt virus
TCV turnip crinkle virus
TMV tobacco mosaic virus
TNE Tris-HCl, sodium chloride, EDTA buffer
Tris Tris (hydroxymethyl) aminomethane tRNA transfer RNA
TTP thymidine triphosphate
TYMV turnip yellow mosaic virus
U uracil
*F pseudouridine ug microgram ul microlitre
USB United States Biochemical
UTP uridine triphosphate
VpG virion protein, genome-linked
WT wild type
(+) positive or messenger (sense)
(-) negative (sense) xiv Acknowledgments
I would first like to thank my supervisor, Dr. D'Ann Rochon, for sharing not only her knowledge of science, but also her love of it, and for making the lab a special place where ideas
and inspiration freely mix with encouragement and sound advice. Her faith in and support of her
students still amazes me everyday and without these things I know I would never have come this far. I would also like to sincerely thank my committee members, Drs. Tony Warren, Frank
Tufaro, Carl Douglas and Dave Theilmann for their interest and suggestions as well as their accessibility and support. Special thanks also to Dr. Dave Theilmann and Dr. Helene Sanfa§on for their many helpful discussions and very generous advice as well as for their critical reviews of manuscripts.
I feel very fortunate in having been able to conduct this research through the Department of
Microbiology and Immunology at the Vancouver Research Station of Agriculture Canada and there are many people to whom I owe a great debt of thanks. I am grateful to my past and present labmates who contributed greatly to this thesis through their knowledge, advice and friendship. In the order of their appearance, thanks to Carol Riviere, Mike Rott, Morven
McLean, Admir Purac, Lawrence Lee, Angus Gilchrist, Renee Finnen, Tim Sit, Marjorie
Robbins, Ron Reade and Howard Damude whose encouragement, empathy, wisdom and wit
definitely helped me through the rough spots. Also to the many additional students, post-docs, technicians and visiting scientists at Agriculture Canada, especially, Lucia Fuentes, Murray
Bulger, Andrew Wieczorek and Claire Huguenot...thanks for making life in and out of the lab challenging, interesting and fun!
A very special thanks also to those who contributed to this thesis by providing technical
assistance, protocols, or materials. Thanks to Drs. J. Skuzeski and R. Gesteland (at the
University of Utah School of Medicine, Salt Lake City) for providing the pAGUS-1 vector.
Thanks also to Tim Sit for the pSC/0.9sg and pSC/PD(-) sg clones as well as his 'new and
improved' protocols, expert advice, and never to be forgotten sense of humor! Thanks to Angus
Gilchrist for his help in screening clones through sequence analysis and to Howard Damude for
assisting in the characterization of mutants using RT-PCR. Special thanks to Lucia Fuentes for first introducing me to the perils of protoplasts but more importantly, for always being ready with a coffee mug and a shoulder in times of need! And thanks also to John Hall and Andrew
Parker for their remarkable patience and fortitude in attempting to explain statistical analysis to a very difficult subject. I am also especially grateful to Andrew Wieczorek for his incredibly generous assistance and excellent advice in so many aspects of my work ranging from monoclonal antibody production and serological analyses to the production of protoplasts and
GUS assays. Even more than that, I appreciate being taken by the hand and introduced again to the world underwater (where the consequences of not following his advice became only slightly more serious).
Finally, thanks to my friends, Carmen for adopting my dog as her own during many late nights in the lab and Jono for the many stimulating discussions regarding the application of general relativity and quantum field theory to this project. My deepest thanks to Howard for becoming my focus and to my parents whose love, patience and support has carried me through this time and who I'm sure will be more relieved than I when this thesis is finally done! Chapter 1 Introduction
The concept of a virus was first introduced in the early 1900s by Beijerinck in describing a new form of infective agent that had the ability to pass through a bacteria-proof filter and could not be detected or cultivated (see Matthews, 1991). The object of these early observations of
Ivanovski (1892) and later Beijerinck (1898) was the causal agent of a disease of tobacco, now known to be tobacco mosaic virus (TMV). Since that time, the study of plant viruses has provided considerable information about the nature of viruses and the functions of their components. TMV was the first virus to be isolated in paracrystalline form, earning Stanley
(1935) the Nobel Prize in Chemistry as well as fueling debate over whether viruses constituted biological entities or inanimate chemicals (see Hughes, 1977). Further purification of TMV by
Bawden and Pirie (1937) demonstrated the virus to be a nucleoprotein complex, which was also announced in the same year by Schlesinger working on bacteriophage (see Hughes, 1977). The observation that viruses consist of only protein and nucleic acid came even before the nature of genetic material was known and attention was initially focused on the protein element as being the infectious component (see Matthews, 1991). However, following the classic experiments of Hershey and Chase (1952), which demonstrated the independent functions of bacteriophage protein and nucleic acid, Gierer and Schramm (1956), Fraenkel-Conrat and Williams (1955) and Fraenkel-Conrat (1956) determined TMV RNA to be the infectious component and the protein coat to serve a protective role (see Matthews, 1991).
With the hereditary role of nucleic acid established, viruses, with their stability, small genome size, and potential for manipulation, became 'windows' through which events occurring inside the cell could be viewed. Being the genetic material, viral nucleic acid was to help solve many of the mysteries of cellular replication, transcription and protein synthesis. In particular, positive strand viral RNA, with its potential to function directly as mRNA, was to serve as a model for probing basic processes underlying the regulation of gene expression. Among the contributions made by plant viruses are those which assisted in establishing the monocistronic nature of eukaryotic cellular mRNA (Shih and Kaesberg, 1973), aided in the study of macromolecular assembly of proteins (Harrison, 1983) as well as autocatalytic (ribozyme) cleavage of RNA (Prody etal, 1986; reviewed in Long and Uhlenbeck, 1993), and provided insight into promoter function and polyadenylation signals in plant cells (reviewed in Benfey and Chua, 1990; Sanfagon etal., 1991). The genomes of plant viruses were also among the first of RNA viruses to which the techniques of reverse genetics were applied (Ahlquist et al.,
1984b) and, together with the development of powerful methods to manipulate DNA in vitro, provided useful tools for the study of cellular processes. Thus with their small genome size and ability to replicate to high levels in plant cells, combined with the use of cell-free translation, plant protoplast systems and easily assayed reporter genes, plant RNA viruses have become convenient model systems for understanding the organization and expression of genetic information.
The following chapter is meant to provide some background on the molecular biology of positive strand RNA plant viruses, in particular, their genome organizations and replication strategies. Most relevant to this thesis are the sections concerning the generation of subgenomic mRNAs and the production of viral proteins which follow the more general topics, above, as well as sections describing the biology and molecular biology of cucumber necrosis virus (CNV). While the following sections will focus on RNA plant viruses, where appropriate, examples from RNA bacteriophage as well as RNA animal virus systems will be provided to acknowledge their important contributions towards understanding molecular aspects of plant virology as well as to place the study of RNA plant viruses into the perspective of virology as a whole. It is interesting to note that after being largely responsible for ushering in the era of modern virology, research on plant RNA viruses lagged behind that of bacterial and vertebrate viruses due in part to a lack of plant cell culture systems as well as tools for the study and manipulation of RNA genomes. With the development of new technologies, as discussed above, RNA plant viruses are making important contributions in such areas as viral
RNA evolution, recombination and replication which form the basis of molecular virology. 1.1 Positive strand RNA viruses
The genomes of viruses may be composed of RNA or DNA (and may be single or double stranded); however, by far the majority of eukaryotic viruses contain RNA genomes which are single stranded and of messenger sense (+) polarity (Francki et al, 1991). One feature unique to RNA viruses is their capacity for rapid change which is a consequence of both a high mutation rate due to the absence of a proofreading function associated with the RNA-dependent
RNA polymerase (RdRp; Holland et al., 1982) as well as the ability of the RdRp to dissociate and reassociate with the template RNA resulting in recombinant molecules (Lai, 1992).
Despite their potential for rapid evolution, certain sequence motifs have been found to be conserved over a broad range of divergent virus groups (Haseloff et al., 1984; Goldbach et al,
1991; Koonin and Dolja, 1993). The most universal of these motifs are those contained within enzymes which mediate genome replication and expression, with the RdRp domains being the best conserved (Zimmern, 1988; Koonin and Gorbalenya, 1989; Koonin and Dolja, 1993).
Based on the degree of sequence conservation between RdRp motifs, relationships have been found between all (+) strand RNA viruses sequenced to date, including those of plants, animals and bacteria. These relationships have constituted their classification into three large supergroups, each of which contain several well defined lineages (see Koonin and Dolja, 1993 for most recent review). Supergroup 1, also referred to as the picornavirus-like supergroup, consists of the picorna-like lineage (which includes picornaviruses, comoviruses, nepoviruses and calciviruses), the poty-like lineage (including potyviruses and bymoviruses), the lineage made up of nodaviruses, sobemoviruses and luteoviruses, the dsRNA lineage and the arteri-like lineage (coronaviruses, toroviruses and arteriviruses). Supergroup 2 or the flavivirus-like supergroup includes the RNA phage lineage, the flavivirus and pestivirus lineages and the lineage consisting of plant viruses with small genomes (BYDV-PAV luteovirus, dianthoviruses, necroviruses, carmoviruses and tombusviruses). Supergroup 3, also called the alphavirus-like supergroup, is composed of the tymo-like lineage (tymoviruses, carlaviruses, potexviruses, capilloviruses), the rubi-like lineage (rubella virus and alphaviruses) and the tobamo-like lineage (tobamoviruses, tricornaviruses, hordeiviruses, tobraviruses, and closteroviruses). Each of these supergroups contain viruses which differ widely in genome size, organization and translation strategy suggesting the parallel evolution of important features necessary for genome replication and expression (Koonin and Dolja, 1993)
1.1.1 Genome structure and organization
The size of non-defective (+) strand RNA viral genomes as well as the organization of the genes which they encode vary considerably. While the genomes of (+) strand RNA viruses range from under 3.5 kb for several RNA phage to over 30 kb in the case of coronaviruses, most are in the range of ca. 5 to 10 kb (see Koonin and Dolja, 1993). The information required to produce a complete infection by (+) strand RNA viruses may be encoded by a single RNA molecule (i.e. monopartite) or by a segmented genome comprising more than one RNA component (i.e. multipartite). Various structures are found at the termini of viral genomes; 5' terminal structures include a m7G5'pppXpYp3' cap (note the X and Y nucleotides are not methylated as they are in cellular and animal virus mRNAs; see Matthews, 1991), a di- or triphosphate, or a small covalently linked protein (VpG), and 3' structures include a polyadenylate sequence, a hydroxyl group, or a tRNA-like structure. The overall organization of the genome is reflected in the strategy by which the genes encoded by a particular virus are expressed (for reviews see Kaesberg, 1987; Mayo, 1987; Morch and Haenni, 1987). With the exception of retroviruses (which will not be considered here), all (+) strand RNA viruses require the synthesis of nonstructural proteins in order to initiate infection. Therefore, the first genes to be encoded are usually those for enzyme(s) which are essential for replication unless an alternative translation strategy such as proteolytic processing enables their production from other regions of the genome (see section 1.1.5). In addition to proteins which mediate replication, the genomes of many (+) strand RNA plant viruses also contain coding regions for one or more capsid proteins and may encode discrete proteins involved in viral transport and vector transmission. 1.1.2 Viral Proteins
Nonstructural (Replication-associated) Proteins
All non-defective (+) strand RNA viruses encode a component or components of the RNA replicase (Koonin and Dolja, 1993). This enzyme complex contains the viral encoded RdRp activity as well as other activities which may be encoded by separate viral proteins, different domains of the same protein or by host factors (for recent reviews see David et al., 1992;
Duggal et al, 1994; Pogue et al, 1994). The functions of the nonstructural proteins are usually inferred from the presence of sequence motifs similar to those present in biochemically characterized enzymes. The RdRp domains of all (+) strand RNA viruses share at least three core sequence motifs with the signature motif being a glycine-aspartate-aspartate (GDD) tripeptide within a conserved sequence context (Koonin and Dolja, 1993). These core motifs have been demonstrated to be essential for RdRp activity in the replicase of the picornavirus, encephalomyocarditis virus, and are suggested to be involved in the binding of nucleotide triphosphates (Sankar and Porter, 1992; Koonin and Dolja, 1993).
In addition to RdRp activity, the replicases of some (+) strand RNA viruses, generally those with a genome of over 6 kb, also contain RNA helicase activity (those with genomes under 6 kb may recruit cellular factors for this activity; Koonin and Dolja, 1993). RNA helicase activity, associated with duplex unwinding during RNA transcription and replication, has been demonstrated for the cylindrical inclusion protein of plum pox poty virus (Lain etal,
1990; 1991) and can be inferred by the presence of conserved sequence motifs (Koonin and
Dolja, 1993). For the replication of (+) strand RNA viruses with capped 5' termini, the replicase is also proposed to contain methyltransferase and guanylyltransferase functions required for capping activity (Koonin and Dolja, 1993; Strauss and Strauss, 1994).
Methyltransferase activity, responsible for methylation of the 5' guanosine of the cap structure, has been demonstrated in Sindbis virus (Durbin and Stollar, 1985; Mi et al, 1989; Mi and
Stollar, 1991) and tentatively identified in a number of related virus groups (Koonin and Dolja, 1993). The domains of viral methyltransferases contain conserved motifs which may or may not be related to those found in cellular methyltransferases (Koonin and Dolja, 1993).
RNA viruses which express their genomes through the production of polyprotein precursors also encode proteases necessary for the liberation of individual viral proteins required for replication or assembly. The two main classes of proteases encoded by (+) strand RNA viruses are the chymotrypsin-related cysteine and serine proteases and the papain-like cysteine proteases, the activities of which have been demonstrated in a number of viruses (see Koonin and Dolja, 1993). Some of these proteases are necessary for the processing of the majority of the proteins encoded by the virus (e.g. picornaviruses and potyviruses) while others are required to perform only one cleavage (e.g. the capsid protein of alphaviruses which functions as an autoprotease to liberate its carboxy-terminus; Hahn et al., 1985). RNA viruses which require proteolytic processing may also recruit cellular proteases for the production of some proteins; host enzymes generally mediate the processing of virion envelope proteins while nonstructural and capsid proteins are commonly processed by viral encoded proteases (Koonin and Dolja, 1993).
Structural Proteins
The protein coat, or capsid, of (+) strand RNA viruses consists of many copies of viral encoded protein molecules (usually only one or two distinct types in the case of plant viruses) which are assembled into highly symmetrical structures (reviewed in Harrison, 1983;
Lomonossoff and Wilson, 1985; Rossmann and Johnson, 1989). The capsid structures of (+) strand RNA viruses are generally either rod-shaped, with the protein subunits packed into a helical array, or spherical with the subunits arranged with icosahedral symmetry. Variations on these common themes include bacilliform or filamentous particles and the addition of an outer lipid envelope often containing viral-encoded glycoproteins (Harrison, 1983). Encapsidated inside the particle is the viral RNA which may be wound between the subunits of rod-shaped capsids (Lomonossof and Wilson, 1985), associated with basic residues of the icosahedral
shell, or stabilized by polyamines or hydrophobic interactions (Rossmann and Johnson, 1989). Although the capsids of most rod-shaped as well as spherical plant viruses consist of a single type of protein, some of these capsid structures (like those of the animal picornaviruses) have been found to contain more than one type of coat protein molecule (e.g. comoviruses, Wu and
Bruening, 1971; and beet yellows closterovirus, Agranovsky et al, 1995). The capsids of spherical viruses contain multiples of 60 protein subunits with the majority of plant virus particles consisting of 180 identical protein subunits bound in a quasiequivalent manner with a triangulation (T) number of 3 (where T x 60 is the number of protein subunits in the capsid).
There are exceptions, however, such as the capsids of nepoviruses which consist of a single copy of one large protein with three major domains and so are, like the picornaviruses, T=l icosahedrons (see Rossman and Johnson, 1989).
The crystal structures of a number of spherical plant RNA viruses, including tomato bushy stunt virus (TBSV; Harrison et al, 1978), turnip crinkle virus (TCV; Hogle et al, 1986), southern bean mosaic virus (Abad-Zapatero et al, 1980), cowpea mosaic virus (Stauffacher et al, 1987) and bean pod mottle virus (Chen et al, 1989), have been determined at high resolution. Although little or no amino acid sequence similarity is apparent among the coat proteins of these and other (+) strand RNA viruses (e.g. picornaviruses), they all share conserved structural domains based on the 'jelly roll' p" barrel conformation which likely reflects their common ancestry (Rossman and Johnson, 1989). This (3 barrel structure corresponds to the shell (S) domain of the coat proteins of these viruses with the TBSV and
TCV coat proteins containing an additional protruding (P) domain which projects outward from the virus particle (see section 1.2). The capsid proteins of many plant viruses also contain binding sites for divalent cations, particularly calcium, which are thought to function in maintaining the integrity of the particle until it reaches the low calcium environment of the host cytoplasm where the viral RNA is released (Durham et al, \911; Hull, 1978). Unlike the situation for some animal viruses where disassembly of the capsid and release of the RNA is associated with a conformational change due to receptor binding and/or fusion between membranes (or hydrophobic residues) in the low pH environment of endosomal vesicles, the entry of viruses into plant cells is not believed to be mediated by cell surface receptors (reviewed in Wilson, 1985). Instead, after entry through wounds or transmission via seed or vector, uncoating of virus particles and release of the viral RNA from destabilized capsids (due to a low calcium or hydrophobic environment) is proposed to occur through a cotranslational disassembly process mediated by host ribosomes (Wilson, 1985). The roles of the capsid protein in the life cycle of RNA viruses are therefore numerous and varied including both protection of the encapsidated nucleic acid against degradation as well as release of the viral
RNA into the cytoplasm of the host during infection. In addition, the coat proteins of RNA plant viruses have been found to be associated with vector specificity (reviewed in Harrison,
1987; see also Atreya et ai, 1991; McLean et al., 1994), host range and symptomatology
(reviewed in Dawson and Hilf, 1992), cell to cell and/or long distance movement (reviewed in
Leisner and Howell, 1993; see also Laasko and Heaton, 1993 and below) and, in at least one case, viral RNA replication (reviewed in Jaspars, 1985).
Movement Proteins
Specific to plant viruses is the production of movement proteins which mediate the spread of the virus within the infected plant (see Deom et al, 1992; Citovsky and Zambryski, 1993;
Goldbach et al, 1994 for recent reviews). Because the plant cell wall prevents entry of viruses via the membrane fusion or endocytic pathways exploited by animal viruses, plant viruses have evolved distinct strategies for movement between adjacent cells (Deom et al, 1992). It is generally accepted that many plant viruses utilize the plant intercellular connections, the plasmodesmata, for cell-to-cell spread and the vascular system for extended spread, thus dividing the process into short distance and long distance movement (reviewed in Atabekov and Taliansky, 1990). It is also presumed that most, it not all, viruses capable of systemic spread (i.e. a combination of both of the above processes) encode protein(s) which enable the movement of the virus between cells as well as in and out of the vascular system. While the mechanism(s) by which movement proteins operate is poorly understood, two patterns of plant virus movement have emerged. The first of these patterns is exemplified by the tobamoviruses
(e.g. TMV) and the other by the comoviruses (e.g. cowpea mosaic virus) (reviewed in Deom et al, 1992; Mushegian and Koonin, 1993; Goldbach et al, 1994). In tobamoviruses, the 30 kDa movement protein (the first definitively demonstrated to potentiate cell-to-cell movement;
Deom et al, 1987) has been shown to both modify the plasmodesmata size exclusion limit and to bind single stranded nucleic acid in a cooperative, although nonspecific, manner (Wolf etal,
1989; Citovsky et al., 1990). Based on these observations, a model has been proposed in which the movement protein, analogous to a molecular chaperone, binds to genomic RNA in order to transport it in an unfolded complex through the modified plasmodesmata (Koonin et al, 1991).
This type of movement process is additionally characterized by the potential for cell-to cell spread in the absence of viral capsid protein (Mushegian and Koonin, 1993). In comoviruses
(as well as nepoviruses), cell-to-cell movement requires both capsid protein as well as movement protein. In this case, the movement protein is associated with the formation of tubular structures protruding from the cell wall which subsequently associate with plasmodesmata (van Lent etal., 1991; Wieczorek and Sanfagon, 1993). Virus spread is thought to occur via the transport of intact particles (hence the requirement for coat protein) through the tubular structures and plasmodesmata into adjacent cells (van Lent et al, 1991).
Movement proteins have been tentatively identified in many plant virus groups based on the lack of a productive infection in whole plants associated with the absence of these proteins (see
Mushegian and Koonin, 1993). Amino acid sequence comparisons of known and putative movement proteins has identified a conserved motif which may represent a hydrophobic domain for interaction with cellular proteins (reviewed in Mushegian and Koonin, 1993).
Attempts to genetically map the functional domains of movement proteins have also revealed the presence of sequences which are required for virus infectivity, rate of movement, localization of movement protein to the plasmodesmata or cell wall, RNA binding and/or
altered phenotype (Berna et al, 1991; Calder and Palukaitis, 1992; Citovsky et al., 1992; Emy
etal., 1992; Gafney etal, 1992; Giesman-Cookmeyer andLommel, 1993). 1.1.3 Replication of genomic RNA
The multiplication cycle of (+) strand RNA viruses involves four basic steps: disassembly of the RNA from the capsid, translation of genomic RNA for the production of proteins required for subsequent transcription and expression, replication of RNA resulting in the synthesis of additional (+) strands, and encapsidation of the RNA genomes (reviewed in David et al, 1992).
Much of what is known about the strategy of genome replication in (+) RNA viruses is based on studies of the RNA bacteriophage Q(3 (reviewed in Blumenthal and Carmichael, 1979; see also Meyer et al, 1981; Barrera, et al, 1993 ). From this work, it was originally discerned that the replication process itself involves transcription of a complementary (-) strand RNA from the (+) strand RNA template followed by the synthesis of (+) strand progeny RNA from the (-) strand template. During the replication of (+) strand RNA viruses, RNA synthesis is highly asymmetric with (+) strands produced in great excess over (-) strands and production of the latter selectively ceasing earlier in the replication process. In the alphaviruses, Sindbis virus, brome mosaic virus (BMV), and TMV, such asymmetric replication has been found to reflect differences in the strategy of (+) and (-) strand RNA production indicating that different forms of the replicase complex are responsible for their synthesis (Sawicki et al, 1981; Ishikawa et al, 1991; Marsh etal, 1991; Sawicki and Sawicki, 1993).
The replicase complex
The isolation of replicase complexes from a number of (+) strand RNA viruses has contributed greatly to an understanding of the RNA replication process (reviewed in David et al, 1992). The replicase or replicase complex (which includes associated host and/or viral factors) has been purifed from tissues infected with plant viruses with tripartite genomes (e.g.
BMV, cowpea chlorotic mottle virus, cucumber mosaic virus, and alfalfa mosaic virus, A1MV)
as well as those with bipartite (e.g. cowpea mosaic virus) and monopartite genomes (e.g. TMV
and turnip yellow mosaic virus, TYMV). Although in most cases, the replicase complex is not
capable of faithfully producing full-length (+) RNA strands, the potential for (-) strand synthesis has lead to extensive characterization of (-) strand promoters located at the 3' terminus of the (+) strand template. By analogy with the replicase of QP RNA bacteriophage, it is predicted that the replicase complex of most (+) strand RNA viruses is composed of both viral and host encoded proteins. The Qp replicase, still among the best characterized, consists of the viral encoded RdRp, bacterial translation elongation factors EF-Tu and EF-Ts, the ribosomal protein SI (Blumenthal and Carmichael, 1979) and a 36 kDa ribosome-associated protein (identified as the host factor responsible for plus-strand initiation; Kajitani and Ishiama,
1991) . The replicase of BMV has been shown to contain the interacting, viral encoded la
(containing methyltransferase and helicase domains) and 2a (containing the polymerase domain) proteins (reviewed in Duggal et al., 1994; see also Kao et al., 1992). In addition, the replicase contains several host proteins, including a protein antigenically related to translation factor eIF3, and formation of the replicase complex is dependent upon coexpression of viral proteins and viral RNA (Quadt et al., 1993; 1995). Based on the replication strategy of Qp, it is speculated that one function of the recruited host factors is to bring the replicase in proximity with the 3' terminus following binding to internal sites on the (+) strand RNA (Meyer et al.,
1981). The emerging theme of an association of translation factors with (+) strand RNA replication lends speculation to the idea that such factors may have a general role in viral RNA replication (Duggal et al., 1994), a suggestion further supported by the presence of tRNA-like structures at the termini of several (+) strand RNA viruses.
Terminal replication structures
Several structures present in the genomes of (+) RNA viruses have been hypothesized to play a role in viral RNA replication due to both their conservation and location. The 3' termini of the genomic RNAs of a number of plant viruses contain highly conserved regions which both structurally and functionally mimic tRNAs (reviewed in David et al, 1992; Duggal et al,
1994). These tRNA-like structures are capable of interacting with enzymes normally specific to cellular tRNAs such as aminoacyl tRNA synthetase, elongation factor 1-a and nucleotidyl
transferase (Hall et al, 1972; Bastin, 1976; Bujarski et al, 1986), the latter of which may act to repair the terminal CCA in order to maintain sequence integrity (Rao et al, 1989). Specific aminoacylation of the 3' terminus is characteristic of a given virus group; bromo- and cucumoviruses accept tyrosine (Hall et al., 1972; Kohl and Hall, 1974), tobamoviruses accept histidine (Oberg and Philipson, 1972) or valine (Beachy etal, 1976), and tymoviruses accept valine (Yot et al, 1970; Pinck et al, 1972). Although the significance of these tRNA-like structures and their interaction with tRNA-associated enzymes is not entirely understood, they have been shown to be involved in initiation of (-) strand genomic RNA synthesis (Ahlquist et al, 1984a; Morch et al., 1987). In BMV and TYMV, in which the tRNA-like structures have been best characterized, deletion and mutational analyses have identified both sequence- specific and structural regions which are essential for adenylation and aminoacylation as well as for recognition by the replicase, (Dreher et al., 1984; Bujarski et al., 1985; Bujarski et al.,
1986; Morch etal, 1987; Dreher and Hall, 1988a,b). Tyrosylation of BMV genomic RNAs 1 and 2 (but not RNA 3) is essential for replication and has been suggested to function in sequestering host elongation factors (Bujarski et al, 1985; Dreher et al, 1989; Rao and Hall,
1991). Similarly, mutations which affect valylation in TYMV also debilitate replication (Tsai and Dreher, 1991; 1992). It has been proposed that tRNA-like endings also functioned in the ancient RNA world by tagging RNA molecules for replication and their presence at the termini of present day plant viruses represents a 'molecular fossil' still exploited for replication purposes (Weiner and Maizels, 1987).
In the absence of a tRNA-like structure, the 3' termini of genomic RNAs may contain a poly(A) tract (e.g. como- and potyviruses as well as furoviruses, see below) or simply end in a terminal hydroxyl group (e.g. ilarviruses and A1MV). Although these termini do not appear to be directly involved in replicase binding, the presence of RNA pseudoknots upstream of the poly(A) tail or tRNA-like structure appear in some cases to be necessary for efficient replication. In TMV, deletion and mutational analysis demonstrated the importance of the pseudoknot region upstream of the tRNA-like structure in both replication and systemic spread
(Takamatsu et al, 1990). Highly structured stem loops are also predicted to occur upstream of
the tRNA-like ending in barley stripe mosaic virus as well as the poly(A) sequence in cowpea mosaic comovirus and mutations in the latter were shown to severely affect replication (Rohll etal. ,1993; Duggal etal., 1994).
The 5' termini of the genomic RNAs of several plant viruses also contain stem loop or nonfunctional (i.e. non-aminoacylatable) tRNA-like structures which are proposed to function in viral replication (Marsh et al., 1988; Duggal etal., 1994; Pogueef al., 1994). In BMV, the 5' terminus of (+) strand RNA, as well as complementary bases at the 3' terminus of (-) strand
RNA, are predicted to fold into stable stem loop structures (Pogue and Hall, 1992). By analogy with poliovirus (Andino et al., 1990), the stem loop structure present at the 5' terminus of (+) strand BMV RNA is suggested to function in the initiation of (+) strand synthesis (Pogue and
Hall, 1992; Pogue et al., 1994). It is postulated that after the synthesis of a complementary (-) strand RNA, a region within the 5' (+) strand structure is recognized by a host factor which binds to the double stranded complex to yield a single stranded region at the 3' end of the (-) strand RNA. This newly exposed single stranded region in the (-) strand RNA may then interact with the replicase complex to promote synthesis of (+) strand genomic RNA (Pogue et al., 1994). The presence of conserved stem loop structures in beet necrotic yellow vein furovirus (Gilmer et al., 1993) as well as Sindbis virus, and demonstration of their importance for the promotion of (+) strands, suggests that this strategy may be used by many (+) strand
RNA viruses (Nesters and Strauss, 1990; Strauss and Strauss, 1994).
Cis-acting replication sequences
The discovery of sequence motifs at the 5' termini of many (+) strand RNA viruses that share a striking resemblance to eukaryotic tRNA sequences has stimulated investigation into the role of these regions in the promotion of (+) strand synthesis. First observed as tandem repeats in BMV, these motifs closely resemble the internal control regions (ICR) 1 and 2 (also referred to as box A and B) of RNA pol III promoters found within tRNA genes (French and
Ahlquist, 1987; Marsh et al., 1989). The ICR-like sequences are inherently also tRNA-like sequences and, due to their nearly palindromic nature, are found within the terminal stem loop
structures of both (-) and (+) strands (in the region corresponding or complementary to the T\j/C loop of the tRNA). The similarity of these motifs to the internally located promoters of tRNA genes and their presence on the 5' (+) strand (and thus the complementary 3' (-) strand) initially implicated them as promoters for (+) strand RNA synthesis. Subsequent deletion or mutation of the ICR-like motif present in the intercistronic region of BMV RNA3 (see below) resulted in a severe reduction of RNA 3 accumulation through effects on initiation of (+) strand synthesis
(Pogue et al, 1990; Pogue et al., 1992; Smirnyagina et al, 1994). The presence of ICR2-like motifs in most viruses having aminoacylated 3' endings (e.g. bromo-, cucumo-, tobamo- and tymoviruses; see above) as well as others lacking tRNA-like structures at their 3' termini (e.g. alphaviruses and A1MV; van der Vossen etal, 1993) is suggested to be reflective of a common ancestry of viruses in the alphavirus-like supergroup that is also shared with eukaryotic tRNAs
(French and Ahlquist, 1988; Marsh et al., 1988; 1989). The importance of these sequences in the initiation of (+) strand RNA synthesis has been most extensively investigated in the promotion of (+) strand subgenomic mRNAs.
1.1.4 Generation of subgenomic mRNAs
One feature shared by many members of the alphavirus-like supergroup is the expression of
at least some genes through subgenomic mRNA generation (Koonin and Dolja, 1993). The production of subgenomic mRNAs is one strategy by which internally located open reading frames (ORFs) of multicistronic eukaryotic RNA viruses may be expressed and regulated during replication. Two mechanisms for the synthesis of subgenomic RNAs have been proposed: the first, discontinuous leader RNA-primed transcription, is thought to occur during
the production of coronavirus subgenomic mRNAs (Spaan et al, 1983; Lai et al, 1984; Lai,
1990) and the second, internal initiation of transcription on (-) strand template RNA has been
shown to occur in vitro for BMV (Miller et al, 1985), A1MV (van der Kuyl et al, 1990) and in
vivo for TYMV (Gargouri et al, 1989). This latter mechanism first requires transcription of a
genomic-length (-) strand template from the (+) strand genomic RNA by the viral replicase
(and associated host factors). The viral replicase (in conjunction with the same or different factors) is thought to then bind to internally located promoter regions on the (-) strand template to produce one or more (+) subgenomic mRNAs (see Fig. 1.1). These subgenomic mRNAs may together form a 'nested set' of molecules which are 3' coterminal with genomic RNA but contain otherwise internally located ORFs at their 5' termini.
The promoter regions on the (-) strand template responsible for directing the synthesis of subgenomic mRNAs have been well studied in several members of the alphavirus-like supergroup, most notably Sindbis virus (Levis et al, 1990; Raju and Huang, 1991; Hertz and
Huang, 1992) and the plant tricornaviruses (French and Ahlquist, 1988; Marsh et al, 1988;
Allison et al, 1989; van der Kuyl et al, 1990; Pacha and Ahlquist, 1992; Boccard and
Baulcombe, 1993; Smirnyagina et al, 1994). The subgenomic promoter regions of these viruses do not share extensive nucleotide sequence similarity but do contain similar sequences in their promoters and upstream elements which resemble the ICR2 (box B) motif discussed above. The intercistronic region in BMV RNA3 contains both sequences that are required for efficient amplification of this RNA (which include the ICR2 motif) as well as core promoter and activating sequences (also ICR2-like) that direct the synthesis of the coat protein subgenomic mRNA (French and Ahlquist, 1987, 1988; Marsh et al, 1988). Included in this region is an oligo(A) activator element which is not present in the subgenomic promoters of other members of the alphavirus-like supergroup; however, recently second-site mutations have been discovered which compensate for its absence (Smirnyagina et al, 1994). The observation that the oligo(A) tract is not absolutely required, together with the conservation of sequence motifs between many members of the alphavirus-like supergroup, suggests possible parallels in the mechanism of subgenomic RNA transcription among viruses of this group (French and
Ahlquist, 1988; Marsh et al., 1988; Smirnyagina etal, 1994). In contrast to members of the
alphavirus-like supergroup, the promoter regions of members of the picornavirus-like
supergroup have not been extensively studied. Among those members of the picornavirus-like
supergroup that generate subgenomic mRNAs (including sobemoviruses, luteoviruses, and
coronaviruses), only coronavirus subgenomic mRNA production has been characterized and
this does not involve initiation from a promoter per se but instead from a leader RNA primer. (+) GENOMIC RNA 5'—[ .3'
REPLICASE
(-) RNA 3' | 5'
REPLICASE
5' .3'
(+) SUBGENOMIC MRNAs
5'. .3'
Fig. 1.1 Model for the generation of subgenomic mRNA by internal initiation of transcription (Miller et al., 1985). During replication, the viral RNA-dependent-RNA- polymerase (RdRp) and associated factors (i.e. the replicase complex) binds to the 3' end of (+) strand genomic RNA and generates a full-length complementary (-) strand. This (-) strand RNA acts as a template for the synthesis of one or more (+) strand subgenomic RNAs through the binding of the replicase to internal promoter elements along the (-) strand template. The subgenomic RNAs are 3' coterminal with genomic RNA but contain different ORFs at their 5' termini such that each member of a nested set of subgenomic RNA molecules may express a different product. The subgenomic promoter regions of applicable members of the flavivirus-like supergroup
(which includes dianthoviruses, necroviruses, carmoviruses and tombusviruses) have also not been characterized.
1.1.5 Production of viral proteins
The genomes of (+) strand RNA viruses have the capacity to function directly as mRNA and so have evolved to utilize the host translational machinery in order to be recognized and efficiently translated by eukaryotic ribosomes. Unlike eukaryotic mRNAs, which are capped, polyadenylated, and generally monocistronic (Shih and Kaesberg, 1973), many (+) strand RNA viruses lack these terminal structures and, moreover, contain more than one coding region (i.e. are multicistronic). According to the scanning model for the initiation of translation by eukaryotic ribosomes, however, usually only the 5' proximal coding region of a mRNA is expressed (for recent reviews see Kozak 1991a,b). The scanning model postulates that during
translation, the 40S ribosomal subunit, along with Met-tRNAimet and associated initiation factors, initially binds at the capped 5' end of the message and migrates linearly until it reaches the first AUG codon at which time it is joined by the 60S ribosomal subunit and translation is initiated. Translation is terminated when the 80S ribosome encounters a termination codon afterwhich one or both of the ribosomal subunits dissociate from the transcript and are ordinarily not able to reinitiate translation. To overcome the restriction of eukaryotic ribosomes efficiently translating only 5' proximal cistrons of capped cellular mRNAs, (+)
strand RNA plant viruses have developed a number of strategies which enable the expression of
downstream coding regions (for recent reviews see Kuhlemeier, 1992; Gallie, 1993; Rohde et
al, 1994).
Translational strategies
A general strategy for expressing all of the information of a multicistronic (+) strand RNA
virus is for each of the coding regions to be situated at the 5' termini of different viral RNA components. Such segmentation of the genome results in division of the genomic information between two or more RNA components, each serving as a monocistronic mRNA for a single protein. This translation strategy is exemplified by the bipartite como-, tobra-, and nepovirus groups (for review see Mayo, 1987) as well as the tripartite bromo-, cucumo-, and alfalfa mosaic virus groups (i.e. tricornaviruses; see Kaesberg, 1987 for review). Similarly, the generation of subgenomic mRNAs allows the genome of multicistronic viruses to be expressed through subgenomic molecules which contain the coding region for a different viral protein at their 5' terminus (see section 1.1.4). Because each subgenomic mRNA expresses only the 5' proximal cistron, each is functionally monocistronic. As an alternative strategy, all or a portion of the viral genome may be expressed as a single long polyprotein from a monocistronic mRNA. The polyprotein may then be processed into two or more functional gene products by viral or host encoded proteases or by autocatalytic cleavage. Best studied in the picornaviruses
(for review, see Palmenberg, 1990), proteolytic processing also occurs in the alphaviruses as well as the nepo-, como-, poty-, tymo- and sobemovirus groups of plant viruses.
During translation two independent processes, readthrough translation and ribosomal frameshifting, may occur which allow the expression of downstream portions of the viral genome without the necessity of reinitiation (for reviews see Atkins etal., 1990; Ten Dam et ai, 1990; Gesteland et ai, 1992). For readthrough translation, the RNA molecule contains a
'leaky' termination codon which is recognized some proportion of the time as a sense codon by eukaryotic ribosomes. This process is thought to be mediated, in the case of TMV, by naturally occurring tyrosine-specific suppressor tRNAs which contain a pseodouridine residue in the
G\|/A anticodon and therefore potentiate readthrough of leaky amber (UAG) codons (Bruening et al, 1976; Beier et ai, 1984a,b). The efficiency of suppression is also influenced by sequences flanking the stop codon; mutational analysis of downstream codons in TMV has revealed that the 3' context confers leakiness and represents part of the signal for suppression
(Skuzeski et al, 1991; Zerfass and Beier, 1992). In addition to members of the tobamovirus group, such readthrough translation has also been demonstrated, or is suspected to occur, in luteoviruses, tobraviruses, carmoviruses and tombusviruses (Rohde et al., 1994). Ribosomal frameshifting is an alternative strategy which also enables the by-pass of termination codons to express the downstream region of an overlapping cistron in a different translational reading frame. Frameshifting subsequent to translation of a portion of an upstream cistron (but prior to termination of this cistron) allows the ribosomes access to a second downstream cistron and results in the production of a fusion protein. Since only a certain proportion of the ribosomes change frame at the frameshift signal, such fusion proteins are produced in addition to, rather than instead of, the protein which is encoded exclusively by the first cistron. Important features associated with the frameshift region include a heptanucleotide sequence (termed the slippery sequence as it is proposed to allow the tRNA to slip from pairing with its correct in-frame codon) and a stem loop, or pseudoknot structure, which likely causes pausing of the ribosomes to facilitate the frameshift (Atkins et al, 1990;
Ten Dam et al., 1990; Gesteland et al, 1992). Translational frameshifting was first discovered in retroviruses (Jacks and Varmus, 1985; Jacks etal, 1988) and coronaviruses (Brierley etal,
1987), and is also used by several plant viruses including the luteoviruses, potato leafroll virus
(Prufer etal, 1992; Kujawa etal, 1993), barley yellow dwarf virus (Brault and Miller, 1992) and beet western yellows virus (Garcia et al, 1993), as well as the dianthovirus red clover necrotic mosaic virus (Xiong et al, 1993; Kim and Lommel, 1994) to express a portion of their genomes.
While the translation strategies described above all conform to the ribosomal scanning model for translation initiation, a number of viral mRNAs contain internally located AUG codons which are accessed instead of, or in addition to, the first AUG codon (Kozak, 1991a).
To explain the apparent deviations from the ribosome scanning model, several strategies and the features necessary for their operation have been proposed. These include: internal ribosome entry, first discovered in picornaviruses (Pelletier and Sonenberg, 1988; reviewed in
McBratney et al, 1993; Jackson et al, 1994) and since then in other systems including the plant cowpea mosaic virus (Macejak and Sarnow, 1991; Thomas et al, 1991; Liu and Inglis,
1992; Wang et al, 1994), nonlinear ribosome migration (termed the ribosome shunt mechanism; Fiitterer et al, 1993) and transactivation (Bonneville et al, 1989) both in cauliflower mosaic virus, termination-reinitiation in papillomavirus (Tan et al, 1994) and influenza B virus (Horvath et al,, 1990) and leaky ribosomal scanning in reovirus (Munemitsu and Samuel, 1988), simian virus 40 (Sedman and Mertz, 1988), hepatitis B (Lin and Lo, 1992;
Fouillot et al., 1993), retroviruses (Schwartz et al, 1992; Caroll and Derse, 1993), rabies virus
(Chenik et al, 1995), barley yellow dwarf luteovirus (Dinesh-Kumar and Miller, 1993) and peanut clump furovirus in vitro (Herzog et al, 1995). Leaky ribosomal scanning can be rationalized by a normal scanning mechanism if both AUG codon position and context as well as secondary structure and leader length are considered (Kozak, 1991a,b). For leaky scanning to occur, the first AUG codon usually lies in a suboptimal context allowing some ribosomes to scan past the first potential start site and initiate instead at the next downstream AUG codon.
This tendency to bypass the first start site may be promoted in mRNAs containing short 5' noncoding leader sequences, lacking considerable secondary structure downstream of the 5' proximal AUG codon (Kozak, 1991a,b), and with the second AUG codon in relatively close proximity to the first initiating codon (Kozak, 1995).
Initiation codon selection
The features influencing initiation codon selection have been well studied in animal systems and expanded to include those mRNAs which appear to deviate from the Kozak scanning model of translation initiation. The optimal context for initiation in animal systems is
CCAACCAUGG with the -3 position (relative to the AUG codon) being the most important mediator of translational efficiency (Kozak, 1991a,b). In cases where the -3 position is not a purine, the remaining positions (particularly the +4 position) exert their influence.
Comparisons of plant start site sequences suggest the consensus sequence for plants is
AACAAUGGC (Joshi, 1987; Liitcke etal, 1987; Cavener and Ray, 1991), however, there is
some uncertainty concerning the nucleotide positions which most strongly regulate initiation codon selection. In vitro studies using wheat germ extracts have indicated that, in contrast to
animal systems, the -3 position is not an important modulator in plants (Liitcke et al, 1987).
Also, modification of the start codon context in the -3 position of a plant viral mRNA did not result in an increase in gene expression in plants (Lehto and Dawson, 1990). However, the simultaneous modification of nucleotides in the -3, +4 and, in some cases, the +5 position, in other plant systems have provided evidence for the importance of one or more of these positions (Taylor et al, 1987; Jones et al, 1988; McElroy et al, 1991). In contrast, substitutions in both the -4 and -1 positions (those sites which differ between the consensus sequences of plants and animals) have demonstrated a negligible contribution of these positions in an otherwise consensus context (Guerineau et al, 1992). The disparities observed in translational efficiencies between various plant systems may be explained by differences in salt conditions (particularly magnesium; Kozak, 1989a) or the absence of certain translation factors in vitro or by the requirement for additional proteins or sequences present on constructs for transient or stable expression studies in vivo (Rohde et al, 1994). In general, the posttranscriptional regulation of plant gene expression, including the contributions made by codon context and leader sequence to translational control, have not been well characterized
(reviewed in Gallie, 1993).
1.2 The Tombusvirus Group
Cucumber necrosis virus is a member of the genus Tombusvirus which consists of 12 additional species and, together with the genus Carmovirus, form the family Tombusviridae
(reviewed in Morris and Carrington, 1988; Martelli et al, 1988,1989; Russo et al., 1994). All members of the tombusvirus group are small spherical viruses with a ca. 30 nm particle composed of a single type of capsid protein. The natural host range of these viruses is narrow and restricted to dicotyledons, however the artificial host range is wide and includes both monocotyledonous and dicotyledonous families (Martelli et al, 1988). The majority of tombusvirus species occur in temperate regions where they have been reported to occasionally cause diseases of economic importance (Martelli etal, 1988). Tombusviruses are stable and naturally transmitted in soil and water, however, fungus transmission has also been demonstrated to occur in CNV (Dias, 1970; Stobbs etal, 1982) as well as the closely related carmovirus, melon necrotic spot virus (Furuki, 1981) and unclassified cucumber leaf spot virus
(Campbell et al., 1991). The relationships between tombusviruses have been demonstrated using both serological techniques as well as nucleic acid hybridization analyses. While those examined appear to be extremely similar, CNV remains distinct in being serologically unrelated to a number of viruses within the group (Gallitelli et al., 1985; Koenig and Gibbs,
1986; Rochon and Tremaine, 1988; Rochon et al., 1991). The complete nucleotide sequence of several members of this group including CNV (Rochon and Tremaine, 1989), cymbidium ringspot virus (CymRSV; Grieco etal, 1989a,b), the cherry strain of TBSV, a close relative of the type member of the tombusvirus group (TBSV-ch; Morris and Carrington, 1988; Hillman et al, 1989; Hearne etal, 1990), and artichoke mottle crinkle virus (AMCV; Tavazza etal,
1989; Grieco and Gallitelli, 1990; Tavazza et al, 1994) have been determined and their genome organizations deduced. These viruses share identical genome organizations, considerable nucleotide sequence similarity in noncoding regions, and amino acid similarity in certain coding regions of the genomes (see Russo et al., 1994). In particular, extensive amino acid similarity was found in the putative RdRp genes, which together with those of the carmoviruses and more distantly related dianthoviruses, necroviruses, and the luteovirus,
BYDV-PAV, places these viruses in the flavivirus-like supergroup of (+) strand RNA viruses
(Habili and Symons, 1989; Riviere and Rochon, 1990; Koonin and Dolja, 1993; see section
1.1). Additional features of the tombusvirus group will be described below in relation to CNV.
1.2.1 Cucumber necrosis virus
CNV was originally characterized as a member of the tombusvirus group through comparison of the double-stranded RNA (dsRNA) intermediates generated upon infection and by nucleic acid hybridization analyses (Rochon and Tremaine, 1988). Systemic infection is
limited to cucumber in nature (McKeen, 1959) and, as noted above, natural spread of the virus
is facilitated by zoospores of the fungus, Olpidium bornovanus (Dias, 1970; Campbell et al,
1995). CNV was first isolated as the causative agent of a disease in greenhouse-grown cucumbers, where it caused severe foliar symptoms and serious stunting of growth, but has not been reported to cause any other disease of agricultural significance (McKeen, 1959; Rochon et al, 1991). Mechanical inoculation of most experimental hosts results in a localized infection on the inoculated leaves, and in at least two experimental hosts, Nicotiana clevelandii and N. benthamiana, results in systemic infection characterized by rapid and severe necrosis (Rochon et ai, 1991). Infection by CNV, as well as by other tombusviruses may be associated with the presence of defective interfering (DI) RNA molecules which interfere with genomic RNA replication and attenuate disease symptoms (reviewed in Russo et al., 1994). The generation of
DI RNAs during plant viral infection, first discovered in TBSV-ch (Hillman et al, 1987) and since reported in other plant virus systems including the tombusviruses, CNV (Rochon, 1991) and CymRSV (Burgyan et al, 1989), and the related carmovirus, TCV (Li et al., 1989; Li and
Simon, 1991) has recently become the subject of intensive study (reviewed in Russo et al,
1994; see also White and Morris, 1994; Chang et al., 1995; Dalmay et. al., 1995; Finnen and
Rochon, 1995 for recent work).
Particle Structure
The three-dimensional structure of the CNV particle and constituent subunits is inferred from that of TBSV which has been determined at 2.9 A resolution (Harrison et al, 1978). The particle is a T=3 icosahedron composed of 180 copies of a 41 kDa viral coat protein. Each coat protein subunit is divided into three functional domains: the basic random (R) domain which is thought to interact with the viral RNA inside the capsid structure, the shell (S) domain which comprises the surface of the virus particle, and the protruding (P) domain which projects outward from the virus shell. The R and S domains are connected by the arm (a) and the S and
P domains by a small five amino acid hinge (Harrison et al, 1978; Hopper et al, 1984). While the S and R domains are common structural features among spherical viruses, the P domain is
specific to members of the tombus- , carmo-, and dianthovirus groups (Harrison et al., 1983).
The P domain is thought to be associated with the immunological properties of the virions (Russo et al, 1994) and to play a role in particle stability and/or assembly as well as vector specificity.
CNV Genome Organization and Expression
Like other tombusviruses, the CNV genome is 4.7 kb and contains at least five, and possibly six ORFs with the capacity to encode proteins of 33, 92, 41, 21, 20 and 3.5 kDa (Rochon and
Tremaine, 1989; Boyko and Karasev, 1992). The ORFs for the 33 and 92 kDa proteins (i.e. p33 and p92) are 5' proximal, the ORF for the 41 kDa protein (p41) is internally located, and the ORFs for the 20 and 21 kDa proteins (p20 and p21) are located at the 3' terminus of the
CNV genome (the sixth ORF for the putative 3.5 kDa protein, designated pX, is located at the extreme 3' terminus; see Fig. 1.2). Infection by CNV genomic RNA results in the synthesis of two 3' coterminal subgenomic RNAs of 2.1 and 0.9 kb (Rochon and Tremaine, 1988; Johnston and Rochon, 1990); the significance of a possible third subgenomic RNA of 0.35 kb is currently under investigation. CNV genomic RNA serves as the template for the production of p33 and p92 (the latter predicted to arise via readthrough translation of the p33 amber terminator codon), the 2.1 kb subgenomic mRNA directs the synthesis of p41, and the 0.9 kb subgenomic mRNA directs the synthesis of both p20 and p21 (Johnston and Rochon, 1990;
Rochon and Johnston , 1991). Similar genome organization and expression strategies have been demonstrated for CymRSV, TBSV and AMCV (Burgyan et al, 1986; Hayes et al, 1988;
Russo etal, 1988; Grieco etal, 1989a,b; Hillman etal, 1989; Hearne etal., 1990; Tavazza et ai, 1989; 1994), with production of p92 observed in vitro from TBSV genomic RNA in the presence of calf liver tRNA (Hayes et al, 1988) and in vivo from TBSV-infected plants
(Scholthof et al, 1995b). Thus CNV and other tombusviruses utilize at least three strategies
for the expression of their gene products from internally located ORFs; these include the
generation of subgenomic mRNAs, readthrough translation of an amber codon, and a third
strategy for the production of both p20 and p21 from a single RNA template. 25
ORF 5 ORF 1 ORF 2 ORF 3 ORF 6 I HI IIIIIIIIIIIIIIIH^* • 4.7 kb genomic RNA ORF 4 p33
p92 (putative polymerase)
2.1 kb subgenomic RNA
p41 (coat protein)
0.9 kb subgenomic RNA
p20,
p21
-[[[|—0.55fcfo subgenomic RNA
pX
Fig. 1.2 Schematic representation of the organization and expression of the CNV genome. The CNV genomic and subgenomic RNAs are diagrammed and their sizes are shown on the right. The boxed regions represent open reading frames (ORFs) with different reading frames indicated by different shading patterns. The encoded protein products are diagrammed below their corresponding ORF and their known or putative functions are indicated. Functions of encoded proteins
CNV p33/p92. Although the functions of most of tombusvirus proteins have not been definitively demonstrated, they have been inferred from amino acid sequence comparisons with proteins of known function or through the effects of mutations introduced into infectious genomic-length transcripts. The amino acid sequence of p33 does not appear to contain conserved motifs (e.g. methyltransferase or helicase domains) from which to deduce its function, however, the analogous proteins of related viruses have been demonstrated to be essential for replication. In both CymRSV and TBSV and well as the related carmovirus, TCV, production of a truncated prereadthrough product or production of only a readthrough product
(through the introduction of deletions, frameshift mutations or the substitution of a sense codon for the amber terminator codon) completely abolished infectivity indicating a requirement for both p33 and p92 for replication (Hacker et al, 1992; Dalmay et al, 1993; Scholthof et al,
1995a). The p92 readthrough product is implicated as the viral replicase from the presence of the glycine-aspartate-aspartate (GDD) tripeptide and surrounding sequence characteristic of an
RdRp domain found in the known and putative replicases of other (+) strand RNA viruses
(Rochon and Tremaine, 1989; Hearne etal, 1990; Dalmay et al., 1993). The readthrough portion of p92 also does not appear to contain a helicase motif suggesting that helicase activity is unnecessary for the replication of tombusvirus genomes or that a cellular enzyme is recruited for this function (Koonin and Dolja, 1993). The importance of p92 in viral replication has been demonstrated in CymRSV and TBSV as discussed above, as well as by the ability of a
CymRSV deletion mutant capable of encoding only the replicase gene to accumulate in protoplasts and to support the replication of a coinoculated CymRSV DI RNA (Dalmay et al,
1993; Kollar and Burgyan, 1994; Russo etal, 1994). In addition, it has recently been shown that TBSV p33 and p92 proteins are coordinately expressed and associated with the membrane fraction of virus-infected plants as is predicted to be the case for components of the viral replicase (Scholthof etal, 1995b). CNV p41. The role of p41 as the virus coat protein has been determined for a number of tombusviruses including CNV; the predicted amino acid sequence of at least the shell domain of these proteins closely resembles that determined by chemical analysis of TBSV coat protein subunits and they are selectively immunoprecipitated with antisera prepared against intact virus particles (Hopper et al, 1984; Burgyan et al, 1986; Hayes et al, 1988; Riviere et al, 1989;
Johnston and Rochon, 1990). Mutations introduced into the coat protein coding region of several tombusviruses have been demonstrated to variously affect virion assembly, symptomatology, and systemic movement (Dalmay et al, 1992; Dalmay et al., 1993; McLean et al., 1993; Scholthof et al., 1993; Sit et al, 1995). In general, these studies have established that the coat protein is not required for replication and cell-to-cell movement but is necessary for wild type systemic spread and symptomatology (reviewed in Russo et al, 1994). For both
CymRSV and TBSV, the rate of spread-and severity of symptoms was more or less affected depending upon the host plant and type of mutation (Dalmay et al, 1992; Scholthof et al,
1993). In contrast, mutations introduced into the coat protein of the related carmovirus, TCV, resulted in decreased in cell-to-cell movement and abolished systemic spread (Laakso and
Heaton, 1993). For CNV, the coat protein has been demonstrated to be dispensible for both cell-to-cell and systemic movement (although it does enhance the rate of systemic spread; Sit et al, 1995) and to contain determinants for the specificity of transmission by the zoospores of its fungal vector (McLean etal, 1993; 1994).
CNVp20/p21. CNV p20 and p21, and the analogous proteins of other tombusviruses, are
encoded by extensively overlapping ORFs of a single subgenomic mRNA (Johnston and
Rochon, 1990). CNV p20 protein is suggested to play a role in viral RNA replication and
symptomatology as its absence leads to the rapid de novo generation of defective interfering
RNAs (thought to arise via a template switch during replication; Lazzarini et al, 1981) and
results in a dramatically attenuated phenotype (Rochon, 1991). The analogous (pi9) protein in
CymRSV is similarly associated with symptom development as its absence also results in a
milder phenotype, however this condition is not correlated with the appearance of DI RNAs in transcript-inoculated plants (Dalmay et al, 1993). These and other observations have suggested that pl9/p20 may also have an auxiliary role in systemic spread in some hosts (Russo et al, 1994; Scholthof et al, 1995a). CNV p21 is implicated as a cell-to-cell movement protein since it shares some amino acid sequence similarity with known and putative movement proteins of other plant RNA viruses (Rochon and Tremaine, 1989; Melcher, 1990, personal communication; Mushegian and Koonin, 1993) and is essential for infection in whole plants
(Rochon and Tremaine, 1989; Rochon and Johnston, 1991). The analogous (p22) protein of
CymRSV is also required for virus accumulation in plants but not protoplasts (Dalmay et al,
1993) and exogenously expressed TBSV p22 has been shown to trans - complement the movement of mutants defective in cell-to-cell spread (Scholthof et al, 1995a). In addition to their postulated roles in movement, the pl9 and p22 proteins of TBSV have also been shown to be important symptom determinants in a variety of host plants (Scholthof et al., 1995a).
1.3 Thesis Objectives
The present work was undertaken to investigate the generation of the CNV 0.9 kb subgenomic mRNA as well as its translation to produce the two proteins, p20 and p21, which it encodes. As alluded to previously, the subgenomic mRNA promoters of members of the flavivirus-like supergroup (to which CNV belongs) have not been characterized. In contrast, the subgenomic promoter regions in applicable members of the alphavirus-like supergroup have been well studied and found to contain similar sequence motifs suggesting potential parallels in subgenomic mRNA transcription. Delineation of the 0.9 kb subgenomic mRNA promoter of CNV could therefore provide useful information concerning the signals necessary for subgenomic mRNA production in one member of the flavivirus-like supergroup and possibly provide insight into sequences required for initiation of (+) strand genomic RNA synthesis in CNV. The production of two proteins from the 0.9 kb subgenomic mRNA of CNV suggests that this RNA may be bifunctional and, if so, indicates that CNV must utilize an alternate translation strategy for expression of the downstream ORF. One possible strategy for initiation of translation at the downstream p20 AUG codon is via leaky ribosomal scanning due to the potentially suboptimal context of the upstream p21 AUG codon (lacking a purine in the
-3 position but containing a G in the +4 position) as well as the unusually short 0.9 kb subgenomic mRNA leader. Analysis of the effect of selected nucleotide substitutions surrounding the p21 AUG codon could provide important information concerning which nucleotides most strongly regulate the efficiency of translation initiation at this AUG codon in plant protoplasts. The effect of these substitutions, as well as an increase in leader length, on expression from the downstream p20 AUG codon could then be determined and potentially provide an understanding of the strategy used for production of p20. The specific objectives of this thesis are therefore as follows:
1. To delineate the 5' and 3' borders of the promoter for the CNV 0.9 kb subgenomic mRNA
2. To determine the role of selected nucleotides surrounding the CNV p21 AUG codon in the efficiency of translation initiation at this AUG codon .
3. To assess the role of leaky scanning in the production of p20 from the bifunctional 0.9 kb
CNV subgenomic mRNA.
During the course of this research, several collaborative projects were also undertaken.
These include the investigation of a possible third subgenomic RNA generated during CNV infection and examination of the role of p21 in the CNV life cycle. The results of these collaborative projects and the conclusions drawn from them will be summarized briefly with the contributions made by J.C.J, described in detail and clearly distinguished from those of the other collaborators. Chapter 2 Materials and Methods
2.1 Plasmid construction
The plasmids listed below were generated at least in part from pK2/M5, a full-length CNV cDNA clone adjacent to the T7 promoter in Bluescribe (Stratagene) phagemid, the detailed synthesis of which is described in Rochon and Johnston (1991). All plasmids were constructed using commercially available vectors (unless otherwise stated) and standard recombinant DNA techniques as described in Sambrook et al. (1989). Restriction enzymes and modifying enzymes were obtained from Bethesda Research Laboratory (BRL), Pharmacia or Boehringer
Mannheim and used according to manufacturer's recommendations. Oligonucleotides were synthesized at the Nucleic Acid - Protein Service Unit (NAPS) at the University of British
Columbia and were purified as recommended. Sequenase was purchased from US Biochemical
(USB) and DNA sequencing carried out according to the dideoxynucleotide method of Sanger etal. (1977) as described in the USB handbook and the simplified procedure of Hsiao (1991).
Site-directed mutagenesis was performed based on the dut, ung~ method described by Kunkel et al. (1987) using a kit supplied by Bio-Rad Laboratories. Standard polymerase chain reaction
(PCR) conditions using DNA or RNA (i.e. RT-PCR with Superscript reverse transcriptase supplied by BRL) as initial templates for amplification were used and are described in detail in
McLean etal. (1993).
2.1.1 Construction of plasmids used to map the 0.9 kb subgenomic mRNA promoter
Large scale deletion constructs to map the 5' and 3' borders of the promoter
The large scale deletion mutants used to roughly define the 5' border of the 0.9 kb
subgenomic mRNA promoter were provided for use in this study and their detailed construction is described in McLean et al. (1993). The plasmid pK2/M5PD(-) contains a deletion of 316 nucleotides corresponding to the region between two Xhol sites introduced into the coat protein protruding domain coding sequence upstream of the 0.9 kb subgenomic mRNA start site (McLean et al, 1993). The plasmid pK2/M5CP(-) corresponds to an in-planta derived deletion mutant of PD(-) but which lacks almost the entire ca. lkb coat protein coding region (McLean et. al., 1993). A diagrammatic representation of these mutants is provided in section 3.2.1 of Results.
To initially map the 3' border of the 0.9 kb subgenomic mRNA promoter, two plasmids containing large scale deletions 3' of the subgenomic start site were constructed. Plasmid pK2/M5NcoI-HpaI was generated by digestion of pK2/M5 with Ncol followed by mung bean nuclease treatment, digestion with Hpal and religation to yield a mutant lacking a 286 nucleotide region encompassing CNV nucleotides 3830 to 4116. pK2/M5NcoI-AsuII was similarly constructed by digestion of pK2/M5 with Ncol and Asull followed by mung bean nuclease treatment and religation to produce a mutant lacking a 504 nucleotide region corresponding to CNV nucleotides 3830 to 4334 (see diagram in section 3.2.3 of Results).
Small scale deletion constructs to map the 5' and 3' borders of the promoter
To refine the borders of the 0.9 kb subgenomic mRNA, two series of deletion constructs were generated. The pK2/M5 X series was constructed from pK2/M5XhoI which contains a single introduced Xhol restriction enzyme site at CNV nucleotide position 3733 located 51 nucleotides upstream of the 0.9 kb subgenomic start site. pK2/M5XhoI was generated from a subclone of pK2/M5 containing the introduced Xhol site (see McLean et al, 1993) by restriction enzyme digestion with Bglll and Ncol which flanked the Xhol site. This fragment was purified following agarose gel electrophoresis using the Qiaex gel extraction kit (hereafter referred to as gel-purified) and ligated into similarly digested pK2/M5. To generate a series of deletions, pK2/M5XhoI was linearized with Xhol and then treated with 0.05 Units Bai 31 exonuclease (BRL) per u\g DNA at 25 °C which resulted in the removal of ca. 50 bp per termini in 10 min. During the 30 min reaction time, aliquots of the reaction were stopped at different time intervals by adjusting the mixture to 50 mM EDTA. The separate Bai 31-treated samples were phenol/chloroform extracted, ethanol precipitated, resuspended and treated with AsuU
(CNV nucleotide 4331) to yield fragments of between ca. 450 and 600 nucleotides. The samples were then gel-purified using Qiaex matrix and ligated into Xhol linearized pK2/M5XhoI which had been treated with mung bean nuclease, digested with Asull and dephosphorylated with calf intestinal phosphatase (CIP) followed by gel-purification. Ligation reactions were transformed into E. coli DH5a cells, the resulting colonies grown in LB media and the DNA extracted and screened by restriction enzyme digestion as follows. DNA was digested with Ndel and Kpnl which flanked the Xhol site in pK2/M5XhoI. Digested DNA was separated on a 4% GTG Agarose (NuSieve) gel to resolve the small NdellKpnl fragments (WT size being ca. 260 nucleotides) and allow the selection of appropriate plasmids for further screening by DNA sequencing. A series of 15 plasmids carrying deletions of between 4 and 74 nucleotides (designated pK2/M5XA4 through -XA74) were finally chosen for further analysis
(see diagram in section 3.2.1).
The pK2/M5 N series was generated by digestion of pK2/M5 with Ncol (corresponding to
CNV nucleotide 3835 located 50 nucleotides downstream of the 0.9 kb subgenomic mRNA start site) followed by treatment with Bai 31 as described above. Further digestion with BglU
(CNV nucleotide 3383) yielded fragments of between ca. 300 to 450 nucleotides which were purified as above and ligated into Ncol, mung bean nuclease, Bglll, CIP - treated and gel- purified pK2/M5 vector DNA. Ligation, DNA extraction and screening were also carried out as above and a series of nine plasmids designated pK2/M5NAl0 through -NA55, carrying deletions of between 10 and 55 nucleotides, were selected (see diagram in section 3.2.3).
2.1.2 Construct containing mutations flanking the 0.9 kb subgenomic mRNA start site
The plasmid pK2/M5BamHI was constructed to determine the effect of nucleotide
substitutions immediately surrounding the 0.9 kb subgenomic mRNA start site.
Oligonucleotide-directed in vitro mutagenesis (Kunkel et al., 1987) was used to introduce a
Bam HI site at nucleotide position 3784 which would result in the alteration of nucleotides at
positions 3784, 3787 and 3788 (where the start site is nucleotide 3785) . pSCHincl.55, a subclone containing a region corresponding to CNV nucleotides 2566 to 4116 (Johnston and
Rochon, 1990), was used to produce a single stranded DNA template for mutagenesis. The phosphorylated mutagenic oligonucleotide, 5ATTAGGGGCTTCTGGArCCTAACCAATTCATGGAT-
ACTGAATACGAAC3'(corresponding to CNV nucleotides 3771 to 3818; the introduced BamHI site is underlined and the modified nucleotides are italicized), was then used to introduce the
BamHI restriction enzyme recognition site which was confirmed by restriction enzyme digestion. A 447 nucleotide Bglll - Ncol fragment containing this site was subcloned into similarly digested pK2/M5 and the entire subcloned region was verified by DNA sequencing.
Transcripts corresponding to pK2/M5BamHI contain nucleotide substitutions in the -1 (U—>G),
+3 (A—>U) and +4 (U—>C) positions relative to the 0.9 kb subgenomic mRNA transcription initiation site (defined as +1; see diagram in section 3.3 of Results).
2.1.3 Construction of plasmids for transient expression in protoplasts
The plasmid pAGUS-1 (Skuzeski et al., 1990) was used to construct both pCGUS and pBGUS mutant vector series described in the following sections for transient expression in protoplasts. pAGUS-1 contains the (3-glucuronidase (GUS) reporter gene flanked by the
CaMV 35S promoter and the nopaline synthetase (NOS) termination signal and was kindly provided by J. M. Skuzeski and R. F. Gesteland (University of Utah School of Medicine, Salt
Lake City). The region between the CaMV 35S promoter and the GUS coding region in pAGUS-1 contains restriction endonuclease recognition sites for BamHI followed by Sal\
Ncol, Hindlll and Apal; the BamHI site corresponds to the transcription start site, the Ncol site contains the ATG initiation codon for GUS and the Hindlll and Apal sites are contained within a short extension upstream of the original coding sequence for GUS (Jefferson et al, 1986), enabling the construction of translational fusions (see section 5.1.4 of Appendix).
Constructs to determine the effect ofp21 codon context on translation
To determine the effect of selected nucleotide substitutions surrounding the CNV p21 start codon on the efficiency of translation initiation, the 5' untranslated leader region in pAGUS-1 was replaced with sequences corresponding to the CNV 0.9 kb subgenomic mRNA leader region. To do this, pAGUS-1 was first digested with BamHI, treated with mung bean nuclease and then digested with Apal. Linearized pAGUS-1 was then incubated with CNV/GUS oligonucleotide (oligo) #1, 5GAATCTAACCAATTCATGGAAAGCTTAGCGGGCC 3', which corresponds to the entire CNV leader region (nucleotides 3785-3804) including the p21 initiation codon (bold) and next two nucleotides, followed by pAGUS-1 GUS coding sequence
(italicized) from the VYmdIII site (underlined) to a partial Apal site (underlined) under conditions described by Edwards etal. (1991) for the ligation of oligonucleotides to single- stranded cDNAs. The resulting construct, designated pCGUS-wt, would give rise to transcripts containing a 5' leader sequence identical to that of the authentic CNV 0.9 kb subgenomic mRNA with the AUG codon for CNV p21 in-frame with GUS.
To obtain constructs containing sequences corresponding to the 0.9 kb subgenomic mRNA leader but with nucleotide substitutions surrounding the AUG codon for CNV p21, a portion of the above clone, pCGUS-wt, was used for the production of a ssDNA template for in vitro mutagenesis. A 100 nt EcoRV/Apal fragment of pCGUS-wt (containing a portion of the
CaMV35S promoter followed by the region corresponding to the CNV 5' untranslated leader) was inserted into similarly digested Bluescript II KS(+) (Stratagene) to give pJUNCTION 1.
This construct was then used to produce ssDNA for in vitro mutagenesis (Kunkel et al., 1987)
5 T G A using the degenerate CNV/GUS oligo #2 'CATTrGGAGAG^AtcCTAACCAA /aTCATG /t /cT-
AGCTTAGCGGG^ which contains different nucleotides surrounding the p21 translation start site.
Specifically, CNV/GUS oligo #2 corresponds to the 3' most 11 nts of the CaMV 35S promoter region ending with the first G of a BamHI site (underlined) introduced into the region corresponding to the CNV leader up to and including the initiation codon (bold) and first codon of p21. This is followed by sequence within the pAGUS-1 GUS coding region (italicized) including a partial Apal site (underlined) but lacking a Hindlll site (mutations are denoted in small case and subscripted where applicable). Appropriate pJUNCTION clones were sequenced then digested with BamHI and Apal and inserted into similarly digested pAGUS-1 to yield the pCGUS series 1 through 8 (see diagram in section 3.7.1 of Results). These constructs direct the synthesis of transcripts containing nucleotide substitutions surrounding the AUG codon for p21 which initiates the synthesis of GUS. In addition, the transcripts contain two nucleotide changes at the 5' end of the leader relative to wild type transcripts corresponding to a BamHl site introduced for cloning purposes.
Constructs to analyze the effect ofp21 codon context on production ofp20
To assess the effect of nucleotide substitutions downstream of the CNV p21 start site on production of CNV p20, the pBGUS mutant series was generated. This series contains sequences corresponding to the 0.9 kb subgenomic mRNA leader followed by the initiation sites for p20 and p21 but, unlike the above, with the p20 start site in-frame with GUS. To generate the pBGUS mutants, in vitro mutagenesis was carried out using an available pSC/2.1sg ssDNA template corresponding to CNV nucleotides 2566 to 4116 (which encompasses the region surrounding the p20 and p21 initiation sites) and the degenerate CNV
5 G A 3 Oligo #35 mixture, 'ATTAGGGGCTTCTGGAtcCTAACCAATTCATG /t /cTACTGAATACGAAC '
(which corresponds to CNV nts 3771 to 3818 ). Mutagenesis using this oligonucleotide would, in addition to introducing nucleotide substitutions (in small case) surrounding the CNV p21 initiation codon (bold), again result in the introduction of a Bam HI site (underlined) at a region corresponding to the CNV 0.9 kb subgenomic mRNA start site. The resulting p21 CONTEXT clones were confirmed by sequence analysis and the 45 nucleotide BamHUNcoI fragment (the
Ncol site overlaps the p20 initiation codon) from each was gel-purified and inserted into similarly digested pAGUS-1 to obtain the pBGUS constructs 1, 4, 5 and 7 (see section 3.7.2).
2.1.4 Construction of plasmids to generate subgenomic-length templates for in vitro translation
Constructs containing an altered pX translation initiation site
To investigate whether the pX subgenomic RNA (which corresponds to CNV nucleotides
4358 to 4701) has a coding function, plasmids were constructed which lack the putative initiation codon for pX (i.e. the 3.5 kDa protein that this RNA has the capacity to encode; Boyko and Karasev, 1992). To change the initiation codon for pX to a nonAUG codon, ssDNA corresponding to pHpa50 (which encompasses CNV nucleotides 3634 to 4639) was generated and used as the template for in vitro mutagenesis (Kunkel et al., 1987) using CNV oligo #36. CNV oligo #36, 5CTTCCCATACGATatCGAGTCAGGTC3' corresponds to CNV nucleotides 4417 to 4442 but contains an EcoRW site (underlined) which introduces two nucleotide substitutions (small case) at the translation initiation site (small case) and results in the alteration of the ATG start codon to a nonATG codon (i.e. ATA). Colonies were screened for mutants containing the EcoRV site by restriction enzyme digestion. The mutated region was verified by sequence analysis and a 116 nucleotide AsuWApal fragment which contains the mutated region was gel-purified and ligated into similarly digested pK2/M5. The resulting construct, pK2/M5AAUGpX, containing the entire CNV genome but with an altered pX initiation codon, was utilized for subsequent infectivity and host range studies (C.J. Riviere,
J.C.J, and D.M.R., manuscript in preparation; see diagram in section 3.4 of Results).
Construction of subgenomic-length constructs containing cDNA which corresponds to the pX subgenomic RNA was carried out in collaboration with C.J. Riviere at the Agriculture and
Agri-Food Canada PARC Vancouver Research Station. To generate constructs containing wild type sequence or constructs with an altered pX initiation site, a 370 nucleotide Asull/Smal fragment (which encompasses the pX subgenomic RNA coding region) from either pK2/M5 or pK2/M5AAUGpX was gel-purified and ligated into AccVSmal digested Bluescribe vector creating pSC/0.35 or pSC/0.35AAUGpX, respectively. (Note that the Asull site from pK2/M5
and the Accl site located in the multicloning region of Bluescribe have compatible sticky ends.)
Run-off transcription from Smal-linearized pSC/0.35 or pSC/0.35AAUGpX using the T3 promoter would generate transcripts corresponding to the 0.35 kb subgenomic RNA but containing an additional 38 nucleotides at the 5' end (10 viral nucleotides and 28 vector
nucleotides) not present in the authentic subgenomic RNA. Subclones with altered translation initiation sites for p20 and p21
Plasmids containing cDNA corresponding to the entire CNV genome but with nucleotide substitutions in the putative initiation codons for p20 or p21 were provided by D.M. Rochon for use in the present study. pK2/M5201 and pK2/M5215 contain nucleotide changes such that the initiation codons which start the translation of p20 and p21, respectively, are changed to non ATG codons (ATG -> TTG in the case of p21 and ATG -> ACG in the case of p21; note that the nucleotide substitution at the p20 initiation site did not result in an amino acid substitution in p21). To construct plasmids containing cDNA corresponding to the 3' terminus of CNV (and thus the p20 and p21 coding regions), a 1 kb Hpall fragment from pK2/M5201 or pK2/M5215 was gel-purified and inserted into AccI-digested, CIP-treated Bluescribe. Run• off transcription using the T7 promoter in the resulting plasmids, pSC/201sg and pSC/215sg, would produce transcripts similar to the 0.9 kb subgenomic mRNA (which normally directs the synthesis of these proteins) but which lack 62 nucleotides of noncoding sequence corresponding to the extreme 3' terminus of CNV RNA and contain an additional 151 viral nucleotides and 28 vector nucleotides not normally present upstream of the 0.9 kb subgenomic
start site. A similar plasmid, pHpa50 which contains wild type cDNA corresponding to the
same 1 kb region in the AccI site of Bluescribe was also provided by D.M. Rochon and used for
initial studies. In vitro transcription from linearized pHpa50 gives rise to transcripts containing wild type 0.9 kb subgenomic mRNA sequence but lacking the extreme 3' 62 nucleotides and containing the additional nucleotides described above. For subsequent studies, the plasmid pSC/0.9sg which contains cDNA exactly corresponding to the authentic 0.9 kb subgenomic
mRNA placed immediately downstream of the T7 promoter in pUC19 (see below for similar
constructions) was provided by T. Sit.
Subclones containing deletions in the p41 coat protein coding region
Two plasmids, pSC/CP(-)sg and pSC/ANM2sg were constructed to determine the potential
for their corresponding transcripts to direct the synthesis of CNV proteins. The plasmid
pSC/CP(-)sg was generated from pK2/M5CP(-), a previously described cDNA clone (McLean etal., 1993) derived from a naturally occurring CNV coat protein deletion mutant (also referred to in section 2.1.1 above). Sequences corresponding to the CP(-) "2.1 kb" subgenomic mRNA which contains a large ca. 1 kb deletion in the coat protein coding region beginning 49 nucleotides after the transcription start site (and so is actually only caX.X kb) were amplified using PCR and two oligonucleotides. Oligo #45 5' AACTGCAGAATTCTA4TACGACTCACTATAGA-
CCAAGCAAACACAAACAC3 contains a Pst I site (underlined) followed 5 nucleotides downstream by the T7 promoter (italicized) and then the first 20 nucleotides of the coat protein subgenomic mRNA leader. Oligo #24, 5'GGGAGTAATGGTACCTCC3', which is the complement of CNV nucleotides 3901 to 3918, corresponds to a region several bases downstream of the p20
AUG codon. The resulting 277 bp PCR product was then gel-purified and ligated directly into the pT7Blue T-tailed vector (Novagen) to produce an intermediate construct. This construct was then digested with PstI and Ncol (corresponding to CNV nucleotide 3830) and the resulting 286 bp product gel-purified. An available plasmid encompassing the entire CNV 0.9 kb subgenomic mRNA and upstream sequences (CNV nucleotides 3383 to 4701) in pUC19
(Pharmacia) was digested with PstI (upstream of the insert in the pUC19 multicloning site) and
Ncol (which overlaps the CNV p20 initiation codon) and ligated with the 286 bp fragment described above. The region obtained using PCR was subsequently confirmed by DNA sequencing. The resulting construct, designated pSC/CP(-)sg, would, upon run-off transcription using T7 RNA polymerase, give rise to transcripts exactly corresponding to the deleted form of the CNV CP(-) "2.1 kb" subgenomic mRNA. These transcripts are similar to those produced by run-off transcription of pSC/0.9sg using T7 RNA polymerase except that they contain an additional 5' 114 nucleotides corresponding to the 5' 49 nucleotides of the WT
2.1 kb coat protein subgenomic mRNA, followed by 52 nucleotides of non-contiguous coat protein coding sequence fused to 13 nucleotides normally present upstream of the 0.9 kb subgenomic mRNA transcription start site.
The plasmid pSC/ANM2sg was generated from a previously described cDNA clone
(pK2/M5ANM2) derived from another naturally occurring CNV coat protein deletion mutant
(Sit et al., 1995). The 5' portion of the 0.9 kb subgenomic mRNA coding sequence (CNV nucleotides 3785 to 3918) along with 33 nucleotides of upstream sequence of pK2/M5ANM2
(which corresponds to the 5' 20 nucleotides of the coat protein subgenomic mRNA leader fused to the 13 nucleotides lying immediately upstream of the 0.9 kb subgenomic mRNA) was amplified using PCR and CNV oligo #45 and #24 (described above). The resulting 196 bp
PCR product was then gel-purified and ligated as above into the intermediate vector, pT7Blue, followed by sequence analysis. The remaining steps were as described for pSC/CP(-)sg. T7 polymerase derived transcripts produced from pSCANM2sg are similar to those produced from pSC/0.9sg except that they contain an additional 33 nucleotides of leader sequence which corresponds to the 5' 20 nucleotides of the 2.1 kb (coat protein) subgenomic mRNA followed by 13 nucleotides corresponding to the region immediately upstream of the 0.9 kb subgenomic mRNA. Because of the similarity between transcripts derived from pSC/0.9sg and pSCANM2sg, the latter containing what amounts to a 33 nucleotide 5' extension, transcripts produced from these two plasmids were also used to analyze the importance of leader length for production of p20 and p21.
Subclones containing nucleotide substitutions downstream of the p21 AUG codon
A series of four pSC/0.9 sg plasmids was generated to determine the effect of nucleotide substitutions downstream of the p21 initiation codon on the relative amounts of p20 and p21 produced in vitro. To construct this series, p21 CONTEXT clones resulting from the in vitro mutagenesis described in section 2.1.3.2 involving pSC/2.1 ssDNA (corresponding to CNV nucleotides 2566 to 4166) and CNV oligo #35 were utilized. Initially, mutations were introduced into the genomic-length CNV cDNA clone, pK2/M5. To accomplish this, a 447 nucleotide Bglll/Ncol fragment (which includes an introduced BamHl site at position 3785 corresponding to the 0.9 kb subgenomic mRNA start site followed by the p20 and p21 initiation sites and downstream nucleotides) from each of the four p21 CONTEXT clones was ligated into similarly digested pK2/M5. This resulted in the generation of four pK2/M5BamHI
mutant clones with the clone containing wild type nucleotides downstream of the p21
translation initiation site being identical to pK2/M5BamHI described in section 2.1.2. To construct plasmids containing subgenomic-length cDNA corresponding to the above four pK2/M5BamHI mutants, a 916 bp BamHUSmal fragment from each which corresponds to the
0.9 kb subgenomic mRNA (but with nucleotide changes resulting from the introduced BamHl site as well as changes downstream of the p21 start site) was ligated into similarly digested
Bluescribe/BamHI. This vector was derived by in vitro mutagenesis using ssDNA prepared from Bluescribe and the phosphorylated oligo 5'GCATGCAAGCTTTpGaTCCCTTTAGTGAG 3'. (The
BamHl restriction enzyme recognition site is underlined with nucleotide changes shown in small case and sequences complementary to the T3 promoter are italicized). Run-off transcription of Smal-linearized plasmids from the above ligation using T3 RNA polymerase would result in transcripts which exactly correspond to the 0.9 kb subgenomic mRNA but contain two nucleotide changes (due to the introduced BamHl site) in the 5' leader region as well as nucleotide substitutions downstream of the p21 start site. Plasmids in which the p21 initiation site is followed by GA, TA, GC or TC are designated pSC/0.9sgSl, -S4, -S5 and -S7, respectively, in keeping with the terminology adopted in section 2.1.3.
2.1.5 CaMV 35S promoter-based constructs to map the promoter for the 0.9 kb subgenomic mRNA
An initial approach to mapping the promoter for the 0.9 kb subgenomic mRNA involved the generation of constructs which contain the GUS coding region downstream of sequences corresponding to putative promoter elements for 0.9 kb subgenomic mRNA synthesis. This entire region was placed in antisense orientation downstream of the CaMV 35S promoter and upstream of the NOS terminator sequence from pAGUS-1 (Skuzeski et ai, 1990) such that transcription would result in the synthesis of capped, polyadenylated antisense RNA.
Recognition of promoter elements in trans by the replicase of a helper virus could then potentially result in transcription of the (-) sense template to (+) sense RNA and enable production of GUS. Constructs to map the 0.9 kb subgenomic mRNA promoter by complementation assay
In order to generate plasmids containing putative 0.9 kb subgenomic mRNA promoter elements upstream of the GUS coding region placed in antisense orientation under the control of the CaMV 35S promoter, a series of four intermediate pBTPro plasmids were initially constructed (see diagram in Supplement section 3.8 of Results). These intermediate constructs were generated using different BamHI (or BgM)INcol fragments containing progressively smaller regions corresponding to sequences upstream of the 0.9 kb subgenomic mRNA start site. The first, pBTProBglll, contains a 447 nucleotide BgHUNcol fragment from pK2/M5
(note that the Ncol site overlaps the p20 start site and the Bglll site is 447 nucleotides upstream) fused to a NcoVSacl (mung bean nuclease treated) fragment of pAGUS-1 which encompasses the GUS coding region. The GUS coding sequence was then followed by an
Asull (mung bean nuclease treated)/ Sail fragment of pK2/M5 which corresponds to the extreme 3' 380 nucleotides of the CNV genome. These three fragments were inserted into
BamHUSall digested Bluescript in a one step ligation procedure and the junction sequences of the resulting clones confirmed by sequence analysis. The plasmid pBTProHpall was similarly constructed however using a 196 BamHIINcol fragment from pHpa50 (see section 2.1.4 and note that the BamHI site of pHpa50 is actually contained within the multicloning site of
Bluescribe). The plasmid pBTProXhoI was constructed using a 122 nucleotide BamHIINcol fragment from pXGUS-1 which contains a Xhol/Ncol fragment from pK2/M5XhoI (see section
2.1.1) in the BamHVSall site of pAGUS-1 (see section 2.1.3). Finally, pBTProBamHI was generated using a 45 nucleotide BamHIINcol fragment from pK2/M5BamHI (see section
2.1.2) . Thus, pBTProBglll, -Hpall, -Xhol and -BamHI contain regions of 402, 151, 50 and 0 nucleotides, respectively, corresponding to sequences upstream of the 0.9 kb subgenomic
mRNA transcription initiation site placed upstream of the GUS coding region. The above
CNV/GUS/CNV sequences were then placed in antisense orientation flanked by the CaMV
35S promoter and NOS termination sequence in pUC19. This was accomplished by the
insertion of a ca. 2.2 to 2.5 kb Sall/Sacl fragment from each (note that the Sacl site is located
upstream of the BamHI site in the multicloning region of Bluescript) into similarly digested pAGUS-1, generating pSGProBglll, -HpaTI, -Xhol and -BamHl (see diagram in section 3.8 of
Results).
Constructs containing genomic-length CNV cDNA behind the CaMV 35S promoter
The plasmid, p35SCNV, was generated to act as a helper virus in protoplasts transfected with the above antisense promoter constructs. For use in its construction, pK2/M5RI, which contains cDNA corresponding to the entire CNV genome (with the exception of the first two nucleotides) was provided by D.M. Rochon. pK2/M5RI contains CNV cDNA in the
EcoRI/Smal location of Bluescribe such that run-off transcription using T7 RNA polymerase gives rise to (+) sense genomic RNA. Since direct cloning of the 4.7 kb CNV sequence behind the 35S promoter was problematic due to the presence of a second internal EcoRI site as well as an imperfect 5' terminus, an intermediate plasmid was constructed. Two fragments, a 1.3 kb
EcoRUAatll fragment corresponding to the 5' terminus of CNV and a 3.4 kb AatWSall fragment corresponding to the remainder of the CNV genome, were inserted into EcoRUSall digested Bluescript. Excision of cDNA corresponding to the entire CNV genome using Pstl
(upstream of the EcoRI site in Bluescript) followed by treatment with mung bean nuclease and digestion with Smal (located at the 3' junction of the CNV insert and vector sequence) and insertion into BamHVSstI digested, mung bean nuclease and CIP-treated pAGUS-1 resulted in the generation of p35SCNV. (Note that the BamHl site corresponds to the CaMV 35S promoter transcription initiation site and the SstI site delineates the 5' border of the NOS termination signal in pAGUS-1; see section 2.1.3 and 5.1.4 of Appendix).
Constructs containing cDNA corresponding to the CNV genome but in which the p20/21 coding regions were replaced by the GUS coding region were also generated in order to assess
GUS activity in protoplast experiments. Two such plasmids were constructed,
p35SCNV/GUSHpaI and p35SCNV/GUSAsuII, containing the GUS coding sequence in the
NcoUHpal site or the NcoI/AsuU site, respectively, of p35SCNV (described above; note that the
Ncol site of p35SCNV corresponds to the p20 start site and the Hpal and ASMII sites are located
256 and 41 nucleotides, respectively, upstream of the p21 stop codon.). To obtain a fragment containing the GUS coding sequence, pAGUS-1 (described in section 2.1.3) was digested with
Sstl (which is located downstream of the GUS stop codon) followed by mung bean nuclease treatment and then digestion with Ncol (which overlaps the GUS start codon). This fragment was then ligated into p35SCNV which had been digested either with Ncol and Hpal (the latter which leaves a blunt end) or ASMII, followed by mung bean nuclease, and digestion with Ncol such that the GUS coding sequence essentially replaced that of CNV p20. The resulting clones were screened by restriction enzyme digestion and the CNV/GUS junctions verified by sequence analysis.
2.2 In vitro transcription
Run-off transcripts used for plant inoculation, protoplast transfection, and in vitro translation were produced from plasmids which were first linearized by restriction enzyme digestion. Full- length transcripts from pK2/M5 or pK2/M5 constructs containing deletions or mutations were synthesized using i'mal-linearized templates and the bacteriophage T7 RNA polymerase
(BRL). Subgenomic-length transcripts were synthesized using templates also linearized with
Smal (unless otherwise indicated) and either T3 or T7 RNA polymerase (BRL). Transcription
reactions contained 40 mM Tris-HCl, pH 7.6, 10 mM NaCI, 6 mM MgCl2, 10 mM DTT, 2 mM spermidine, 0.5 mM each of ATP, CTP, GTP, and UTP, 20 units RNAguard (Pharmacia), 5 (Xg linearized DNA and 100 units of T3 or T7 RNA polymerase in a 100 ul reaction volume.
Reactions were incubated at 37 °C for 1 hr afterwhich time the transcripts were treated differently depending upon their intended use. Since infectivity studies (Rochon and Johnston,
1991) previously determined that a 7-methyl G cap is not required on genomic-length transcripts for inoculation onto plants and because CNV subgenomic RNAs appear to be naturally uncapped (D.M. Rochon, personal communication), a cap analog was not included in the transcription reactions. For plant inoculations, 10 ul of 100 mM sodium phosphate buffer, pH 7, was added to the 100 ul volume transcription reaction and then used immediately for inoculation. For protoplast transfection, transcripts were ethanol precipitated and subsequently taken up in a 10 u.1 volume of sterile H2O immediately before use. Transcripts to be used for in vitro translation were routinely stored at -70 °C (due to the greater amount synthesized; see below) and therefore were phenol-chloroform extracted, ethanol precipitated and dissolved in sterile H2O. Where appropriate (i.e. for time course studies or where different amounts of transcript were used to program cell-free extracts), the amount of RNA was estimated by agarose gel electrophoresis and ethidium bromide staining of a transcript dilution series. In general, only about 5 u.g genomic-length transcript RNA was synthesized from 5 |ig of linearized pK2/M5 based plasmids due to the presence of only a single G residue following the
T7 promoter in these constructs (see Rochon and Johnston, 1991). Since the complete T3 or
T7 promoter region in vectors containing subgenomic-length cDNA was maintained (i.e. the promoter is followed by several G residues), the amount of transcript RNA synthesized was five to 10 fold greater (depending on the amount of polymerase used) than from the above.
2.3 Transcript inoculation
To determine the effect of mutations in the CNV genome on symptomatology, Nicotiana clevelandii plants were inoculated with transcript RNA contained in transcription buffer plus 10 mM sodium phosphate, pH 7 (see above). Four Carborundum-dusted leaves of ca. six week old plants were rub-inoculated with approximately 5 u.g of uncapped transcript (i.e. 1.25 ug transcript RNA/leaf). Two weeks after inoculation, virus was passaged by grinding a systemically infected leaf in 10 mM sodium phosphate buffer, pH 7, using a mortar and pestle.
The plant extract was then used to rub-inoculate the leaves of additional N. clevelandii plants.
Symptoms were monitored for up to two months and RNA was extracted from systemically infected leaves at 6 to 18 days post-inoculation as indicated. 2.4 Protoplast isolation and transfection
Protoplasts from either Cucumis sativus (variety Straight 8) or Nicotiana plumbaginofolia were prepared, isolated and transfected essentially as described by Wieczorek and Sanfacon
(1995). Briefly, cucumber cotyledons from plants grown under sterile conditions were
incubated in CMI medium [250 mM mannitol, 100 mM glycine, Aoki salts (0.2 mM KH2P04>
1.0 mM KNO3, 1.0 mM MgS04-7H20, 10 mM CaCl2-2H20, 0.1 |lM CuS04-5H20, and 1.0 uM KI; Aoki and Takebe, 1969)], 3 mM 2[N-morpholino]ethanesulphonic acid (MES) at pH
5.8 with the addition of 1% cellulase (Onozuka R-10) and 0.1% pectinase (Macerozyme) from
Yakult Honsha Co. After overnight digestion, the protoplasts were released by disruption with a glass rod, filtered through cheesecloth to remove large debris and the filtrate centrifuged at
250 x g for 10 min. The pellet was resuspended in 10 ml 15% w/v Ficoll mw 500,000
(Pharmacia) in CMI (above) and a two-step gradient formed using an overlay of 12% w/v
Ficoll followed by CMI. After centrifugation at 500 x g for 15 min, the protoplasts were collected, washed three times with 0.4 M mannitol and kept at 4 °C for 30 min to 1 hr before transfection. The number and viability of recovered protoplasts was determined by counting a sample of fluorescein diacetate stained (Widholm, 1972) protoplasts using a hemocytometer.
Immediately before transfection, protoplasts were transferred to MaCa medium [0.5 M
mannitol, 0.02 M CaCl2 and 0.1% MES pH 7.0 (Negrutiu et al., 1987)] and the concentration adjusted to 3.3 x 106 viable protoplasts per ml. For transfection, approximately 5 u.g uncapped transcript RNA was mixed with 0.3 ml protoplasts (i.e. 1 x 106) and 0.3 ml polyethylene glycol
w (PEG) solution containing 20% ( /v) PEG 3250 (Sigma) in MaCa (see above). After the
addition of PEG solution, the protoplasts were immediately diluted in 10 ml CMI and
transferred to ice for 15 min followed by centrifugation at 250 x g for 5 min and resuspension
in 5 ml CMI. Transfected protoplasts were incubated in the dark at 20 - 25 °C for the times
indicated and then harvested by centrifugation as above. The supernatant was decanted leaving
a pellet with an approximate volume of 100 u.1 for RNA extraction as described below. N. plumbaginofolia protoplasts were prepared and transfected essentially according to
Negruitiu etal. (1987) with several modifications described in Wieczorek and Sanfacon (1995).
N. plumbaginofolia leaves from plants grown under sterile conditions were incubated overnight in NT media (see Pollard and Walker, 1990) to which 1% cellulase 'Onozuka R-10' and 0.1%
Macerozyme was added. The protoplasts were isolated in a procedure similar to that described above except, instead of using Ficoll, the protoplasts were floated on the sucrose contained in
the filtrate overlaid with W5 solution (150 mM NaCI, 125 mM CaCl2-2H20, 5 mM KC1, and 6 mM glucose, pH 5.8; Negrutiu et al, 1987). Protoplasts were collected from the interface, washed twice with W5, resuspended in W5 and kept at 4 °C for 30 min to 1 hr prior to transfection. As above, immediately before transfection, protoplasts were resuspended in
MaCa such that their concentration was adjusted to 2 x 106 protoplasts per ml and 0.3 ml added to an equal volume of PEG solution together with 20 ag cesium gradient-purified supercoiled plasmid DNA. The mixture was immediately diluted in K3 medium (see Vankan et al, 1988) and incubated for 24 hr at 26 °C afterwhich time the protoplasts were harvested by centrifugation (as above) and the protein extracted for analysis of GUS activity as described below.
2.5 RNA extraction
RNA was purified from systemically infected leaves after first freezing them in liquid nitrogen and grinding them to a powder using a mortar and pestle. Total nucleic acid was then extracted in phenol/chloroform and TNE buffer (100 mM Tris-HCl, pH 7.5, 100 mM NaCI, 10
mM EDTA) containing 5 mM p-mercaptoethanol and 0.1% SDS. The nucleic acid was
precipitated from the twice-extracted aqueous phase following chloroform extraction using
ethanol and then dissolved in sterile H2O. The RNA was analyzed on a nondenaturing agarose
gel and an appropriate amount used for northern blot analysis (described below). RNA was
isolated from infected protoplasts by first collecting the protoplasts by centrifugation for 5 min at 225 x g. Total nucleic acid was then extracted from the pellet as described above and one- tenth of the sample used for northern blot analysis.
2.6 Northern blot analysis
RNA purified as above was denatured in 10 mM methyl mercuric hydroxide and separated by electrophoresis through a 1% agarose gel (unless otherwise indicated) containing 5 mM methyl mercuric hydroxide (Bailey and Davidson, 1976). RNA was blotted onto Zeta-Probe
GT membrane (BioRad) under alkaline conditions (Vrati et al., 1987) and placed in
32 hybridization solution containing 0.25 M Na2HP04, pH 7.2 and 7% SDS at 65 °C. P-labeled
DNA probes containing sequences corresponding to either the 3' terminal 1005 nucleotides of the CNV genome (except for the last 62 nucleotides; designated pHpa50 ) or the 3' terminal
370 nucleotides of the CNV genome (designated pX) were generated by nick-translation
(Sambrook et al., 1989). Radiolabeled RNA probes for the detection of virion sense RNA were prepared by in vitro transcription of EcoR I - linearized pK2/M5RI (a plasmid which contains sequences corresponding to the entire CNV genome with the exception of the second nucleotide) using the bacteriophage T3 polymerase (Sambrook etal., 1989).
2.7 In vitro translation and SDS-PAGE
From 0.5 to 2.0 u.g subgenomic-length transcript RNA was used to program a wheat germ extract cell free translation system (Promega) in which magnesium and potassium concentrations were adjusted to 2.0 mM and 120 mM, respectively. In vitro translation was carried out in the presence of [35S]methionine (ca. 1000 Ci/mmol; New England Nuclear)
essentially according to manufacturer's recommendations. The translation products were
analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) through
a 15% (unless otherwise indicated) separating gel (Laemmli, 1970) and subsequent
fluorography using Entensify (New England Nuclear). 2.8 Determination of relative GUS activity
GUS activity was measured using a kinetic spectrophotometric assay modified from
Jefferson etal. (1986). Approximately 100 ul transfected protoplasts were lysed in 100 pi 2X
GUS extraction buffer (IX buffer consists of 50 mM sodium phosphate, pH 7.0, 1 mM ethylenediamine tetraacetic acid (EDTA), 1 mM dithiothreitol, 0.1% Triton X-100, and 0.1% sarkosyl) during repeated freeze/thaw cycles. The soluble fraction was then collected by centrifugation at 14,000 rpm in an Eppendorf benchtop centrifuge and the protein concentration determined by the Bradford (1976) method using a BioRad protein determination kit. For each sample, 5 pg soluble protein was combined with IX GUS extraction buffer (described above) to a final volume of 900 pi followed by the addition of 100 pi 10X GUS spectrophotometric solution (10X solution contains 1 mg/ml bovine serum albumin and 10 mM p-nitrophenyl p-D- glucuronide in IX GUS extraction buffer). The 1 ml reactions were incubated at 37 °C over a 4 hr period during which time 80 pi of each sample was transferred to a microtitre plate and the reaction terminated by the addition of 20 pi 2.5 M 2-amino-2-methylpropanediol. The absorbance of p-nitrophenol was measured at 415 nm and simple regression analysis was used to determine the slope of p-nitrophenol absorbance over time. The average of three replicates of each sample was taken as the level of GUS activity and the value for pCGUS-wt or pBGUS-
1 arbitrarily assigned the value of one to which the rest of the constructs were compared, giving the relative GUS activities for each (see also section 5.1 of Appendix). Chapter 3 Results
3.1 Analysis of CNV 0.9 kb subgenomic mRNA production
During infection, CNV generates two subgenomic mRNAs of 2.1 and 0.9 kb which serve as templates for the synthesis of the p41 coat protein and the p20/p21 proteins, respectively
(Johnston and Rochon, 1990). The presence of these RNA species in CNV-infected plants was initially demonstrated using both dsRNA and northern blot analysis (Rochon and Tremaine,
1988; 1989) and the transcription initiation sites for both RNA species subsequently mapped by primer extension analysis (Rochon and Johnston, 1991; D.M. Rochon, unpublished observations). A third less than genomic-length RNA of 0.35 kb is also generated during CNV infection and the ability of this RNA species to serve as a subgenomic mRNA or, alternatively, a regulatory RNA is still under investigation (see section 4.2). In order to investigate the generation of one of these subgenomic RNAs, the 0.9 kb subgenomic mRNA, the kinetics of
CNV subgenomic mRNA accumulation in cucumber cotyledon protoplasts was determined and deletion and mutational analysis was then used to characterize the promoter for 0.9 kb subgenomic mRNA synthesis.
3.1.1 Kinetics of CNV subgenomic RNA production in protoplasts
To determine the kinetics of genomic and subgenomic RNA accumulation during CNV infection, cucumber protoplasts were inoculated with wild type (WT) transcripts of a full length
CNV cDNA clone (pK2/M5; Rochon and Johnston, 1991). The relative abundance of CNV- specific RNAs was then examined at 6 hr to 44 hr post-inoculation by northern blot analysis using a CNV specific probe (Fig. 3.1). Although both the 2.1 kb and 0.9 kb subgenomic mRNAs were observed at all time points tested, the 0.9 kb subgenomic mRNA was most abundant (relative to genomic RNA and 2.1 kb subgenomic mRNA) at earlier time points (i.e., M 2 (N oo O \Q 3; £, \o r-i i-H (N co co
•genomic
•2.1 kb
•0.9 kb ^+ m*»
•0.35 kb
Fig. 3.1 Kinetics of the accumulation of CNV subgenomic RNAs in protoplasts. Cucumber protoplasts were inoculated with equal amounts of WT CNV transcripts for the indicated times and and one tenth of each sample was analyzed by northern blotting using a 32P-labeled RNA probe complementary to the entire CNV genome. Bands corresponding to CNV genomic RNA and the 2.1, 0.9 and 0.35 kb subgenomic RNAs are indicated. 6 and 12 hr post-inoculation) whereas the 2.1 kb subgenomic mRNA was relatively more abundant at later time points (i.e., 18 to 44 hr post-inoculation). The early accumulation of the
0.9 kb subgenomic mRNA and the later accumulation of the 2.1 kb subgenomic mRNAs are consistent with the postulated and known roles of their translation products in cell-to-cell movement and virus assembly/long distance movement, respectively (see section 3.5)
3.2 Deletion analysis of the CNV 0.9 kb subgenomic mRNA promoter
To initially map the location of the promoter for 0.9 kb subgenomic mRNA synthesis, the accumulation of subgenomic-length RNA species from transcripts with large deletions both upstream and downstream of the 0.9 kb subgenomic mRNA start site was investigated in protoplasts. Characterization of these large scale deletion mutants enabled the location of sites flanking the 0.9 kb subgenomic mRNA transcription initiation site from which progressively longer deletions towards the subgenomic mRNA start site could be made. The level of accumulation of subgenomic-length RNA species from transcripts corresponding to the resulting deletion mutants was then observed, allowing a more refined delineation of both the 5' and 3' borders of the 0.9 kb subgenomic mRNA promoter.
3.2.1 Large scale deletion analysis of sequences 5' of the CNV 0.9 kb subgenomic mRNA
start site
To initially determine which sequences might be important for 0.9 kb subgenomic mRNA
promoter function, mutants carrying deletions in the CNV coat protein coding region, which
lies upstream of the 0.9 kb subgenomic mRNA start site, were made available for
characterization in cucumber protoplasts (McLean et al, 1993). One such mutant, PD(-),
contains a 316 nucleotide deletion corresponding to the protruding domain of the CNV coat
protein. This deletion ends at an introduced Xhol site located 51 nucleotides upstream from the
0.9 kb subgenomic mRNA transcription initiation site (see Fig. 3.2A). To determine whether I P20 I p41 p21 pK2/M5PD(-) A316nt m— Xhol Xhol p20 pK2/M5CP(-) DO- A1057 nt ym— Xho I Nco I
B 0.9 kb subgenomic start site p41 stop codon p21 start codon p20 start codon EI£CiA£CAATCACTGAAAATGCGGTGCAGGTTGTGlMATTAGGG Nco I Xhol XA4 GCAATCACTGAAAATGCGGTGCAGGTTGTGIAAATTAGGGGCTTCTTGAATCTAACCAATTCAICiGATACTGAATACGAACAAGTCAATAAACCA!!!^ XA11 CTGAAAATGCGGTGCAGGTTGTGIAAATTAGGGGCTrCITGAATCT^^ XA18 TGCGGTGCAGGTTGTGTAAATTAGGGGCTTCTTGAATCTAACCAATTCATGGATACTGAATACGAACAAGTCAATAAACCATGG XA22 GTGrAGGTTOTGTAAATTAGGGGCrrCTTOAATCTAACCAATrCATOGATACTGAATACOAACAAGTCAATAAACCATGG XA23 TGCAGGTTGTGTAAATTAGGGGCITCTTGAATCTAACCAATrCATOGATACTGAATACGAACAAGTCAATAAACCATGG
XA25 CAGGTTGTGIMATTAGGGGCTTCTTGAATCTAACCAATTCAiaGATACTGAATACGAACAAGTCAATAAAC£AJSa XA27 GGTTGTGTAAATTAOGGGCTTrTTGAATCTAACCAATTrATOOATACTGAATAGOAACAAOTCAATAAACCATGG
XA30 TGTGIAAATTAGGGGCTTCTTGAATCTAACCAATTCAiaGATACTGAATACGAACAAGTCAATAAAfXAISe XA31 GTGIAAATTAGGGGCTTCTTGAATCTAACCAATTCATiiGATACTGAATACGAACAAGTCAATAAACXAIGIi XA41 GGGGCTTCTTGAATCTAACCAATTCAJIiGATACTGAATACGAACAAGTCAATAAAa^IGii XA42 GGGCTTCTTGAATCTAACCAATrCAlQGATACTGAATACGAACAAGTCAATAAACI^ICiQ XA43 GGCTTrTTGAATCTAACCAATTCATGOATACTGAATACGAACAAOTCAATAAACCATGG XA51 GAATCTAACCAATTCAOSGATACTGAATACGAACAAGTCAATAAAC£AI(jG XA64 TCAjnGATACTGAATACGAACAAGTCAATAAAJXAIGii
XA74 TGAATACGAACAAGTCAATAAACCAim
Fig. 3.2 Description of deletion mutants used to analyze the 5' border of the CNV 0.9 kb subgenomic mRNA. A. Diagrammatic representation of the two large scale deletion mutants used to delineate the 5' border. The structure of the WT CNV genome is shown in the upper portion of the diagram and relevant portions of the two deletion mutants are shown below. Restriction enzyme cleavage sites used to generate pK2/M5PD(-) are shown along with the sizes of the deletions in nucleotides for both pK2/M5PD(-) and pK2/M5CP(-). B. CNV sequences remaining in the pK2/M5 X series following digestion of Xho I cleaved template with Bai 31 exonuclease are shown. Sequences surrounding the WT CNV 0.9 kb subgenomic mRNA are shown in the upper line. The 0.9 kb subgenomic start site as well as the location of the p41 (coat protein) stop codon and the p21 and p20 start codons are indicated. the deletion in PD(-) affects 0.9 kb subgenomic mRNA production, protoplasts were inoculated with WT CNV and PD(-) transcripts and the levels of 0.9 kb subgenomic mRNA (relative to genomic RNA) were analyzed by northern blotting at 12, 24 and 40 hr post-inoculation. Fig.
3.3 shows that the levels of 0.9 kb subgenomic mRNA in PD(-) infected protoplasts are similar to those in WT CNV infected protoplasts at each time point analyzed. The 40 hr sample of
PD(-) is faint in this experiment due to a problem during loading of the sample. In other experiments the level of the viral RNA species and 0.9 kb subgenomic mRNA was similar to the 40 hr WT level. These studies therefore indicate that the 0.9 kb subgenomic mRNA promoter in PD(-) is not appreciably affected by the large upstream deletion.
The second mutant made available for these studies, CP(-), was derived de novo from PD(-) during infection in whole plants (see McLean et al., 1993) and lacks nearly the entire ca. 1 kb
CNV coat protein coding region. The location of the deletion in CP(-) are shown in Fig. 3.2A.
It can be seen that the 3' border of the deletion is the same as that of PD(-) but that the 5' border is far upstream near the 5' terminus of the coat protein gene. In addition, a small internal portion of the coat protein coding region is retained in CP(-). As above, cucumber protoplasts were inoculated with CP(-) transcripts and the levels of 0.9 kb subgenomic mRNA accumulated over time analyzed by northern blot. Fig. 3.3 shows that the amount of the 0.9 kb subgenomic mRNA (relative to genomic RNA) in CP(-)-infected cucumber protoplasts, like that of PD(-), is not substantially affected compared to that observed in WT CNV-infected protoplasts. In addition, the overall levels of CP(-) viral RNA appear to be higher possibly due to an increase in the replication rate of this smaller template and/or a lack of encapsidation.
These results suggest that the 0.9 kb subgenomic mRNA core promoter begins no farther than
51 nucleotides upstream from the subgenomic mRNA start site and further suggest that important auxiliary promoter elements do not lie within the deleted portions of CP(-) or PD(-).
In addition, these studies show that mutations which affect coat protein synthesis (and thus viral RNA encapsidation) do not appear to inhibit the ability of genomic RNA to be stably replicated. WT PD(-) CP(-) M & M J* J3 J3 J3 £ J3 CN Tt O CM CN o (N ^ *t O CN TT CN
Fig. 3.3 Accumulation of PD(-) and CP(-) 0.9 kb subgenomic mRNAs in cucumber protoplasts. Cucumber protoplasts were inoculated with equal amounts of WT, PD(-) or CP(-) transcripts for the indicated times and one tenth of each sample was analyzed by northern blotting using a 32P-labeled RNA probe complementary to the entire CNV genome. Bands corresponding to CNV genomic RNA and the 2.1, 0.9 and 0.35 kb subgenomic RNAs are indicated. The multiple arrowheads indicate the different sizes of the "genomic" and "2.1 kb subgenomic" RNAs affected by the 316 and 1057 nucleotide deletions in PD(-)and CP(-), respectively (see Fig. 3.2). 3.2.2 Deletion analysis of the 5' border of the 0.9 kb subgenomic RNA promoter
The above analysis of PD(-) and CP(-) RNA accumulation in protoplasts indicates that the promoter for the 0.9 kb subgenomic mRNA lies downstream of the deleted region, the 3' border of which corresponds to an introduced Xho I restriction enzyme recognition site at CNV nucleotide position 3733 (McLean etal., 1993). This Xho I site, located 51 nucleotides upstream of the 0.9 kb subgenomic mRNA start site (Rochon and Johnston, 1991) was used as a convenient site from which to make further downstream deletions toward the start site from a similarly positioned Xho I site in a full-length CNV cDNA clone. A schematic representation of the deletion constructs used to map the 5' border (with respect to virion sense RNA) of the
0.9 kb subgenomic mRNA promoter is shown in Fig. 3.2B.
For initial analyses, transcripts were synthesized from selected mutants (pK2/M5XA4, A18,
A22, A41, A64 and A74), transfected into cucumber protoplasts and the resulting levels of subgenomic RNA relative to genomic RNA were determined by northern blot analysis. Fig.
3.4A demonstrates that 0.9 kb subgenomic mRNA levels are not substantially affected by deletions of up to 22 nucleotides downstream of the Xho I site. However, a deletion of 41 nucleotides is associated with decreased levels of 0.9 kb subgenomic mRNA and deletions of
64 nucleotides or more appear to abolish 0.9 kb subgenomic mRNA production.
For subsequent more refined promoter analyses, transcripts with deletions of between 22 and 51 nucleotides downstream of the Xho I site were analyzed as above. Fig. 3.4B indicates that deletions of up to 31 nucleotides do not noticeably affect the level of 0.9 kb subgenomic mRNA, but as before, a deletion of 41 nucleotides is associated with reduced 0.9 kb
subgenomic mRNA levels. In addition, a deletion of 51 nucleotides appears to completely
inhibit 0.9 kb subgenomic mRNA synthesis. The reduced levels of subgenomic RNA
associated with XA41 suggests that the promoter for 0.9 kb subgenomic mRNA lies upstream
of the 3' border of XA41. However, the levels of 0.9 kb subgenomic mRNA appear to be
relatively unaffected in XA43 which contains two additional deleted nucleotides compared to
XA41. To examine this apparent anomaly in more detail, the levels of 0.9 kb subgenomic o 5 a a T3 CL) c CN d Sf o PH CO > ccj c3 CJ VH T3 O u cj ~ewx I a U CO T3 43 ts CJ AS a vol ZPVX o g CO c -C PQ 73 CL) CJ en _iwx OJ CL> _ a I 45 ^ 03 o ~ewx •4—I 43 4o3« O bO CN e Tt iwx CN O .S UH iwx ti >» o i- bOTJ U 45 £ c3 3g 'o5 pq* & cj 1A\ I 0 e to CL) 43 ISVX < O z C T3 Z CJ >>43 ewx 42 Tt- CJ iwx .g o cO T3 CN CJ xevx cj 42 3 5 CJ LO oevx 3 cu V) Z * .5 ti rv> 43 « 3 « CJ LZVX H CN ^ ON &1 cj I u SS bJj Z 42 £ZVX 43 JS 3 I u •s f B5 CJ 43 a ZZVX
3.2.3 Large scale deletion analysis of sequences 3' of the CNV 0.9 kb subgenomic mRNA start site
To determine whether large scale deletions downstream of the 0.9 kb subgenomic mRNA start site affect promoter function, transcripts were synthesized from constructs with deletions in the p20 and p21 coding regions. ANcoI-Hpal and ANcoI-AsuII (see Fig. 3.5A) contain deletions of 286 and 504 nucleotides, respectively, downstream of the Nco I site at CNV nucleotide position 3830 (which forms part of the p20 start codon and is located 50 nucleotides downstream of the transcription start site). Fig. 3.6A shows that ANcoI-AsuII-infected protoplasts accumulate near WT levels of the 0.4 kb deleted form of the "0.9 kb subgenomic mRNA" over time (i.e. 20, 30 and 40 hr post-inoculation). Similarly, Fig. 3.6B indicates that protoplasts inoculated with ANcoI-AsuII or with ANcoI-Hpal accumulate near WT levels of
deleted forms of the "0.9 kb subgenomic mRNA" (i.e. 0.4 kb and 0.6 kb, respectively) at 24 hr
post-inoculation. These results demonstrate that the 3' border of the core promoter for the 0.9
kb subgenomic mRNA core promoter lies within 50 nt downstream of the start site for
transcription. In addition, the accumulation of both these mutants to WT levels in protoplasts
suggests that RNA accumulation is not drastically affected by the absence of either p21 or p20. pK2/M5ANcoI-Hpal ~1 p41 I—TI ^ | h-E53 Hpal
PK2/M5ANcoI-AsuII ~1 p41 ^ — TJ-@
Xfto/ Ncol AsuII
J} 0.9 kb subgenomic start site
p41 stop codon f^" />27 jfcirt codon /?20 start codon CTCGAGCAATCACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGCTrCTTGAATCTAACCAATTCATGGATACTGAATACCiAACAAGTCAATAAACCATGG
Xho I Ncol
Om^CAATCACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGCTTCTTGAATCTAACCAATTCAIQGATACTGAATACGAACAAGTCA NA10
CTC£A£CAATCACTGAAAATGCGGTGCAGGTTGTGIAAATTAGGGGCTTCTTGAATCTAACCAATTCAIG.GATACTG NA16
CJIIG^iiCAATCACTGAAAATGCGGTGCAGGTTGTGIAAATTAGGGGCTTCTTGAATCTAACCAATTCAIG.GATAC NA27
amAiCAATCACTGAAAATGCGGTGCAGGTTGTGlAAATTAGGGGCTTCTTGAATCTAACCAATTCAI^ NA32
CTX^A^CAATCACTGAAAATGCGGTGCAGGTTGTGTAAATTAGGGGCrTCTTGAATCTAACCAATTCA NA34
CjmAiiCAATCACTGAAAATGCGGTGCAGGTTGTGlAAATTAGGGGCTTCTTGAATCTAACC NA40
CTmA£CAATCACTGAAAATGCGGTGCAGGTTGTGIAAATTAGGGGCTTCTTGAATCT NA44 CICGAGCAATCACTGAAAATGCGGTGCAGGTTGTGIAAATrAGGGGC NA55
Fig. 3.5 Description of deletion mutants used to analyze the 3' border of the CNV 0.9 kb subgenomic mRNA. A. Diagrammatic representation of the two large scale deletion mutants used to delineate the 3' border. The structure of the CNV genome is shown in the upper portion of the diagram and relevant portions of the two deletion mutants are shown below. Restriction enzyme cleavage sites used to generate the two mutants (pK2/M5ANcoI-HpaI and pK2/M5ANcoI- AsuU) are shown with the number of nucleotides (nt) deleted indicated. B. CNV sequences remaining in the pK2/M5 N series following digestion of Ncol cleaved template with Bai 31 exonuclease are shown. Sequences surrounding the WT CNV 0.9 kb subgenomic mRNA are shown in the upper line. The 0.9 kb subgenomic start site as well as the location of the p41 (coat protein) stop codon and the p21 and p20 start codons are indicated. 59
A 20 hr 30 hr 40 hr B 24 hr l-H 1 C3 3 P P p CO oa in < < i
-"0.9 kb"
— tr 4 -0.35kb
Fig. 3.6 Large scale deletion analysis of the sequences 3' of the CNV 0.9 kb subgenomic mRNA start site. Cucumber protoplasts were inoculated with equal amounts of the indicated transcripts and then analyzed in A. at 20, 30 and 40 hr post-infection or in B. at 24 hr post• infection by northern blot. Total RNA was separated on a 2% agarose gel and CNV-specific RNA was detected using a nick-translated cDNA probe corresponding to the 3' terminus of CNV RNA. Bands corresponding to CNV genomic RNA and the 2.1, 0.9 and 0.35 kb subgenomic RNAs are indicated. The multiple arrowheads indicate the different sizes of the "genomic" and "2.1 kb" and "0.9 kb" subgenomic RNAs affected by the 286 and 504 nucleotide deletions in ANcoI-Hpal and ANcoI-AsuII, respectively (see Fig. 3.5). Reports by others have similarly indicated the lack of requirement for p21 and p20 in protoplast infections by other tombusviruses (Dalmay et al, 1993; Scholthof etal., 1993 ).
3.2.4 Deletion analysis of the 3' border of the 0.9 kb subgenomic mRNA promoter
A schematic diagram of the deletion constructs used to further define the 3' border of the 0.9 kb subgenomic mRNA promoter is shown in Fig 3.5B. The Nco I site at the 5' border of the deletion constructs described above was used as the site from which to make further deletions toward the 0.9 kb subgenomic mRNA start site located 50 nucleotides upstream. Transcripts with deletions of between 10 and 55 nucleotides were used to inoculate cucumber protoplasts and the resulting subgenomic mRNA levels were determined by northern blot analysis. Fig.
3.7 demonstrates that deletions of up to 44 nucleotides do not noticeably affect 0.9 kb subgenomic mRNA levels but that a deletion of 55 nucleotides completely inhibits 0.9 kb subgenomic mRNA synthesis. These results indicate that the 3' border of the 0.9 kb subgenomic mRNA extends no further than 6 nucleotides downstream of the transcription start site.
3.3 Mutational analysis of the core promoter for the 0.9 kb subgenomic mRNA
Once the location of the core promoter for the 0.9 kb subgenomic mRNA was established by deletion analysis, it was of interest to introduce mutations within the promoter element and investigate their effect on promoter function. As an initial step, a restriction endonuclease recognition site was introduced into the region corresponding to the 0.9 kb subgenomic mRNA promoter and the effect of the mutation on the accumulation of subgenomic RNA analyzed in both protoplasts and plants. Plants inoculated with mutant transcripts or infected tissue which exhibited changes in symptomatology and rate of systemic spread were also examined for the presence of genotypic revertants. M o i es m cn "t io , o -genomic -2.1 kb •0.9 kb •0.35 kb Fig. 3.7 Deletion analysis of the 3' border of the CNV 0.9 kb subgenomic mRNA. Cucumber protoplasts were inoculated with the indicated transcripts and then analyzed 24 hr post-infection by northern blotting using a nick-translated cDNA probe corresponding to the 3' terminus of CNV RNA. Bands corresponding to CNV genomic RNA and the 2.1, 0.9 and 0.35 kb subgenomic RNAs are indicated. 3.3.1 Effect of mutations in the 0.9 kb subgenomic core promoter on RNA accumulation in protoplasts To investigate the effect of mutations immediately surrounding the 0.9 kb subgenomic mRNA transcription initiation site, a BamH I site was introduced into CNV cDNA (pK2/M5) resulting in the alteration of nucleotides in the -1, +3 and +4 positions (where the transcription start site is +1; see Fig. 3.8). These changes led to the substitution of a G for a U at position -1 (nucleotide 3784), a U for an A at position +3 (nucleotide 3787) and a C for a U at position +4 (nucleotide 3788) in CNV RNA. Northern blot analysis of cucumber protoplasts transfected with transcripts of this mutant (pK2/M5BamHl) indicates a substantially reduced level of 0.9 kb subgenomic mRNA as compared to WT levels (Fig. 3.9). This suggests the involvement of any or all of the mutated nucleotides in the regulation of 0.9 kb subgenomic mRNA synthesis. 3.3.2 Effect of mutations in the core promoter on 0.9 kb subgenomic mRNA production in plants To determine if the lower level of subgenomic mRNA synthesis observed in protoplasts (see Fig. 3.9) would affect the symptoms produced in whole plants, transcripts of the pK2/M5BamHI mutant were inoculated onto N. clevelandii leaves. Plants developed symptoms but the symptoms were delayed and considerably attenuated in comparison to WT infected plants (Fig. 3.10; compare B and C with F). In addition, analysis of viral RNA from systemically infected leaves 18 days post-inoculation indicated that the 0.9 kb subgenomic mRNA accumulates over time (Fig. 3.11) but not to the same high levels as seen in WT infections. Thus, the mutations surrounding the subgenomic RNA start site affect subgenomic RNA levels in protoplasts as well as in plants and lead to the production of an attenuated phenotype. p20 _p41_ p2l CNV WT GUGUAAAUUAGGGGCUUCUUGAAUCUAACCAA * * * M5Bam GUGUAAAUUAGGGGCUUCUGGAUCCUAACCAA Revertant GUGUAAAUUAGGGGCUUCUUGAUCCUAACCAA Fig. 3.8 Nucleotide sequence of the region surrounding the 0.9 kb subgenomic start site in CNV WT RNA and original M5Bam mutant and revertant RNAs. A diagrammatic representation of relevant regions in the CNV genome is shown above. M5Bam RNA contains three introduced nucleotide substitutions surrounding the 0.9 kb subgenomic mRNA start site. M5Bam revertant RNA was isolated from plants inoculated with M5Bam passaged material (see text). The 0.9 kb subgenomic start site is denoted by an arrow (in CNV WT RNA) and nucleotide changes are indicated by asterisks. WT M5Bam Ja M & J3 M M N t 9 rs rf Q '—I CS Tt -H 2.1 kb 0.9 kb 0.35 kb m Fig. 3.9 Accumulation of WT and M5Bam 0.9 kb subgenomic mRNAs in cucumber protoplasts. Cucumber protoplasts were inoculated with equal amounts of WT or M5Bam transcripts for the indicated times and one tenth of each sample was analyzed by northern blotting using a 32P-labeled RNA probe complementary to the entire CNV genome. Bands corresponding to CNV genomic RNA and the 2.1, 0.9 and 0.35 kb subgenomic RNAs are indicated. Fig. 3.10 Comparisons of infections produced by CNV WT and M5Bam transcript RNA and M5Bam passaged RNA. Three leaves of N. clevelandii were inoculated with A. buffer control, B. M5Bam transcript RNA (1.5 ug per leaf), C. M5Bam transcript RNA (1.5 ug per leaf); see below, D. sap from a 2 week post-inoculation M5Bam-infected plant (first passage), E. sap from a 2 week post-inoculation M5Bam first passage-infected plant (second passage) or F. CNV WT transcript RNA (2ug per leaf). All plants are shown 2 weeks after inoculation except for C. which is shown 4 weeks post-inoculation. protoplast plant I a OQ o o •genomic •2.1 kb •0.9 kb •0.35 kb Fig. 3.11 Effects of mutations surrounding the 0.9 kb subgenomic mRNA transcription start site on subgenomic RNA levels in protoplasts and plants. Cucumber protoplasts or plants were inoculated with WT or M5Bam transcripts and then analyzed by northern blotting using a nick-translated 32P-labeled cDNA probe corresponding to the CNV 3' terminus. Protoplasts were analyzed 24 hr post-inoculation. WT CNV- and M5Bam- infected plants were analyzed 6 or 18 days post-inoculation, respectively. 3.3.3 Isolation of 0.9 kb subgenomic mRNA promoter revertants from plants The accumulation of 0.9 kb subgenomic mRNA in plants inoculated with M5BamHI transcripts suggested the possibility that the viral RNA species present in systemically infected leaves no longer contained one or more of the mutations corresponding to the introduced BamH I site. RT-PCR amplification of RNA isolated from systemically infected leaves of transcript inoculated plants followed by sequence analysis, however, revealed no reversion of the sites corresponding to the BamH I mutations or any second-site reversions within a ca. 150 nucleotide region (not shown). Inoculation of plants with extract from systemically-infected leaves of transcript-inoculated plants (i.e. first passage inoculum; see Materials and Methods) or extract from plants infected with first passage material (i.e. second passage inoculum) resulted in the production of symptoms which appeared progressively more severe (Fig. 3.10D and E, respectively). RT-PCR amplification of RNA from plants inoculated with passaged material followed by sequence analysis of individual clones revealed a single nucleotide reversion upstream of the region corresponding to the 0.9 kb subgenomic mRNA start site, that is, the substitution of a U (as in WT) instead of a G (of the original M5Bam mutant) at the -1 position (see Fig. 3.8). This reversion occurred in four out of six clones sequenced and therefore suggests that the -1 position is important for 0.9 kb subgenomic mRNA promoter function. In addition, it appears that the severity of symptoms are directly correlated to the amount of 0.9 kb subgenomic mRNA produced, presumably due to a reduction in the synthesis of p20 and/or p21 which this subgenomic mRNA encodes (see section 3.5). 3.4 Characterization of a CNV 0.35 kb subgenomic RNA species In addition to detecting subgenomic mRNAs of 2.1 and 0.9 kb, northern blot analyses carried out in the above study demonstrated the existence of an additional RNA species of 0.35 kb (see Figs 3.1, 3.3 and 3.6). This RNA species was initially detected in both virions and CNV-infected plants and was demonstrated by northern blot analysis to represent a third p20 p21 pX 0.35 kb subgenomic RNA start ^ ^ (CNV nt 4358) GACTCTTCAGTCTGACTTGGTGGAATCTTGCGAATTTAACTGTTA pX start codon t ^CNVnt4428) p21 stop codon (CNV nt 4370) CTCTTCATGGGTTCCTTCCCATACGATGACGAGTCAGGTCGGG... ATAT (pX AUG codon mutant) Fig. 3.12 Nucleotide sequence surrounding the putative translation initiation site of CNV pX. The 0.35 kb subgenomic RNA transcription initiation site is denoted by a bent arrow and the p21 stop codon and putative pX start codon are underlined with their correponding CNV genomic positions indicated. The broken arrow shows the location of nucleotide substitutions introduced into the pX AUG codon mutant. subgenomic RNA corresponding the extreme 3' terminus of the CNV genome (see Fig. 3.12 for a diagram of its location on the CNV genome); duplicate blots probed with radiolabeled cDNA corresponding to the CNV 5' or 3' terminus distinguished 3' co-terminal subgenomic RNAs from similar-sized defective interfering RNAs which contain both 5' and 3' termini (D.M.R., personal communication). Primer extension analysis indicated that transcription of the 0.35 kb subgenomic RNA likely initiates at CNV nucleotide 4358 (i.e. 70 nucleotides upstream of an AUG which is predicted to initiate synthesis of a 3.5 kDa protein; see Fig. 3.12) however two additional less prominent primer extension products were also detected (D.M.R., personal communication). The presence of more than one RNA species in the 0.35 kb size range was also observed by northern blot analysis of total leaf RNA separated on a 2% agarose gel (see WT lanes in Fig. 3.6). Also indicated in this analysis is the greater accumulation of the 0.35 kb RNA subgenomic RNA relative to genomic RNA late in infection (see also WT lanes in Fig. 3.3), the implications of which will be discussed. The recent observation of a high degree of sequence similarity (both at the nucleotide as well as the predicted amino acid level) based on computer assisted comparisons of the genomes of several tombusviruses indicates that a region near the 3' terminus of tombusvirus genomes may have an important function in the life cycle of these viruses (Boyko and Karasev, 1992). Therefore, as part of a collaborative project with D.M. Rochon and C.J. Riviere, the region of the CNV genome corresponding to the 0.35 kb subgenomic RNA was investigated for its functional significance as well as its ability to encode a sixth small protein (designated pX). 3.4.1 In vitro translation of wild type and mutant 0.35 kb subgenomic RNA transcripts The conservation of a small ORF revealed upon computer translation of the 3' terminal regions of the tombusviruses TBSV, CymRSV, AMCV and CNV suggests that this region has a coding function (Boyko and Karasev, 1992). This possibility is supported by the presence of a conserved AUG codon in a favorable context for translation initiation, an optimal distribution of guanosine residues within the codons for pX, and an identical amino acid motif found in all four sequences (Boyko and Karasev, 1992). In addition, the absence of sequences corresponding to this ORF in the defective interfering RNAs associated with several tombusviruses (e.g. Knorr et al, 1991; Finnen and Rochon, 1993) suggests that conservation of this region is not due to the necessity of maintaining cz's-acting replication sequences (Boyko and Karasev, 1992). To determine if the 0.35 kb subgenomic RNA can direct the synthesis of the predicted 32 amino acid protein (pX) in vitro, synthetic subgenomic RNA corresponding to the 3' terminal 370 nucleotides of the CNV genome was translated in wheat germ extracts. Fig. 3.13 demonstrates that a protein of ca. 3.5 kDa, the predicted size of pX, is synthesized suggesting that pX may also be produced in vivo during CNV infection. The synthesis of in vitro translation products by endogenous RNA, also indicated in Fig. 3.13, has previously been observed in certain batches of wheat germ extracts in the absence of exogenous RNA or when programmed with mRNA which is not efficiently translated. In vitro translation of a synthetic 0.35 kb subgenomic transcript RNA in which the AUG codon for pX was changed to a nonAUG codon (AUA; see Fig. 3.12) resulted in the synthesis of two proteins of ca. 3.5 and 1.5 kDa. It is possible that the 3.5 kDa product arises from initiation at the modified AUA codon and the 1.5 kDa product is the result of initiation at a downstream AUG codon corresponding to CNV nucleotides 4482 to 4484 within the pX ORF. Therefore, even with the synthesis of some pX product from the AUG codon mutant, these results suggest that the AUG codon predicted on the basis of computer comparisons is that which is used to initiate synthesis of pX at least in vitro. 3A.2 Effect of mutations in the pX ORF on infectivity of CNV transcripts To determine the effect of mutations in the pX ORF in vivo, mutations were introduced into genomic length transcripts and these were used to inoculate N. clevelandii plants or protoplasts. As this work was not conducted by the author of this thesis, it will be only briefly described in order to summarize the results. A genomic length mutant carrying an altered pX initiation codon (as described above) replicated to high levels in N. clevelandii plants but the symptoms C/l i 3 H oID <3 7^ ° X X PH OH •c w IT) in ro cn ? ° O o > Q m z z s p41" p33- p21- -(endogenous) p20- 3.5 kDa 1.5 kDa Fig. 3.13 /n v^Vro translation of synthetic pX subgenomic-length transcripts. CNV virion RNA (6 ug) or synthetic transcripts (6 ug) corresponding to the 3' terminus of either WT CNV RNA (M5/0.35pXWT) or the pX start codon mutant (M5/0.35pXAUG) were translated in wheat germ extracts in the presence of 35S-methionine. Protein products were electrophoresed through an 18% polyacrylamide gel containing SDS and analyzed by subsequent fluorography and autoradiography. The bands present in the endogenous lane are believed to represent products directed by an endogenous message when no RNA or poorly translated RNA is added exogenously (see text). The sizes of the in vitro translation products directed by synthetic subgenomic-length transcripts are indicated on the right (in kDa) and the CNV in vitro translation products are indicated on the left. produced were distinctly mild compared to those produced by WT transcripts. A second mutant carrying a deletion located 14 nucleotides downstream of the initiation codon resulting in a frameshift failed to produce symptoms or replicate to detectable levels in N. clevelandii plants or protoplasts. As indicated in the above described in vitro translation experiments, some production of a pX-sized protein from the AUG codon mutant is possible which could account for the difference in symptomatology and replication of the two pX mutants. These results are consistent with the hypothesis that an inability to produce pX leads to an absence of both symptoms and detectable RNA accumulation in vV. clevelandii plants and protoplasts inoculated with the frameshift mutant. However, the possibility remains that ds-acting regulatory sequences which are essential for viral replication have been partially or completely disrupted in the start codon and frameshift mutants, respectively, and it is this alteration of sequence that is responsible for the changes in symptomatology and RNA accumulation (CJ. Riviere and D.M. Rochon, personal communication). 3.5 Production of p20 and p21 from wild type and mutant 0.9 kb subgenomic RNA transcripts Previous studies using both sucrose gradient purified CNV virion RNA and synthetic subgenomic RNA transcripts corresponding to the 3' terminus of CNV demonstrated that the 0.9 kb subgenomic directs the synthesis of both p20 and p21 in vitro (Johnston and Rochon, 1990) . Primer extension analysis indicated the presence of only one RNA species in this size range in infected plants and therefore, as predicted from the nucleotide sequence, both p20 and p21 likely arise from different but extensively overlapping ORFs of the 0.9 kb subgenomic mRNA in vivo. To determine whether both proteins are, in fact, produced during viral infection, and if so, whether they are translated from different ORFs or by incorrect initiation or premature termination of the same ORF, point substitutions were introduced into the ATG codons which define the initiation sites for the p20 and p21 ORFs (Rochon and Johnston, 1991) . Plasmids containing cDNA corresponding to the entire CNV genome and incorporating 73 p20 p41 p21 pX pl9' AUG codon ^ (nt 3890) p21 AUG codon p20 AUG codon yCNVnt 3800) yCNVnt 3832) 0.9 kb subgenomic start (CNV nt 3785) .G AAUCUAACCAAUUCAUGGAUACUGAAUACGAAC AAGUC AAUAAACCAUGGAA...//.. .GGGAUGGAA AUG toACG (M5215) AUG to UUG (M5201) B Animal consensus sequence CACCAUGG Plant consensus sequence AACAAUGGC CNVp21 AUUCAUGG CNVp21 AUUCAUGGA gh jb CNVp20 AACCAUGG CNVp20 AACCAUGGA Fig. 3.14 Nucleotide sequences surrounding the translation initiation sites for CNV p20 and p21. A. The 0.9 kb subgenomic start site and location of the AUG codons for the p20 and p21 ORFs as deduced from the nucleotide sequence (Rochon and Tremaine, 1989). The subgenomic start site and initiation codons are denoted by arrows and/or underlined with the corresponding CNV genomic positions indicated. The broken arrows show the locations of nucleotide substitutions used to produce CNV mutants M5201 and M5215 (Rochon and Johnston, 1991). The shaded bars indicate the different reading frames for the p20 and p21 ORFs. The location of the AUG codon for the putative pi9 (see text) within the reading frame for p21 is also shown. B. Comparison of the initiation sites for CNV p20 and p21 with the consensus sequence for translation initiation in animals (Kozak, 1986) and plants (Lutcke, 1987). Asterisks indicate identity with the corresponding consensus sequence. these nucleotide changes in the p20 and p21 initiation codons (i.e. pK2/M5201 and pK2/M5215, respectively) were then provided by D.M. Rochon for use in the studies outlined in the following section; see Fig. 3.14 3.5.1 In vitro production of p20 and p21 from CNV AUG codon mutants Synthetic subgenomic-length transcripts containing mutations in the p20 and p21 AUG codons were prepared from subclones, pSC/M5201 and pSC/M5215, of the above plasmids and translated in wheat germ extracts. As reported previously (Johnston and Rochon, 1990), WT subgenomic-length transcripts derived from pK2/M5 direct the synthesis of two proteins which comigrate with the p20 and p21 translation products synthesized from CNV virion RNA or from sucrose gradient purified virion derived subgenomic RNA (see Fig. 3.15). In addition, subgenomic-length transcripts which carry an altered AUG codon for the p21 ORF (M5215 sg) direct the synthesis of p20 but only very minor amounts of p21 and subgenomic-length transcripts which carry the altered start codon for the p20 ORF (M5201 sg) direct the synthesis of p21 but not p20 . It is noted in Fig. 3.15 that M5201 subgenomic RNA directs increased synthesis of ca. 19 and 18 kDa proteins which are also produced at a low level by all of the other RNAs tested. The precise genomic origins of pl9 and pl8 translation products are not known at this time but it seems likely that one of them is due to initiation at a downstream AUG codon which is in-frame with the p21 ORF (see Fig. 3.14A) and that the other due to initiation of translation at a nearby non-AUG codon which may be in- or out-of-frame with the p20 and p21 coding sequence. The possible relevance of this observation to the conclusions drawn in this study will be discussed. In summary, these in vitro studies strongly suggest that both the p20 and p21 products predicted from the CNV genomic sequence are produced during CNV infection, that they are derived from distinct ORFs, and that they are translated from the same 0.9 kb subgenomic mRNA species. 150 _S -S 00 40 ^ OS ON 40 1? "3 © d '5 7 o H in in L—' =3 u s £ s S £ p21 p20 pl8/19 Fig. 3.15 /« v*7ro translation of natural and synthetic CNV subgenomic mRNAs containing the p20 and p21 ORFs. CNV virion RNA, wheat germ endogenous RNA, synthetic 0.9 kb subgenomic transcript RNA, subgenomic-length transcripts with an altered p20 initiation codon (M5201 0.9 kb sg), subgenomic-length transcripts with an altered p21 initiation codon (M5215 0.9 kb sg) or authentic sucrose gradient fractionated 0.9 kb subgenomic mRNA (6 ug exogenous RNA per each in vitro translation reaction) were translated in wheat germ extracts in the presence of [35S]methionine. In vitro translation products were then analyzed by SDS-polyacrylamide gel electrophoresis (through a 15% separating gel) and fluorography. The numbers on the right refer to the CNV proteins which correspond to each in vitro translation product. 3.5.2 Effect of mutations in the start codons of p20 and p21 on infectivity The above conclusions are supported by further in vivo studies, however, as these were not conducted by the author of this thesis, they will be only briefly described insofar as they relate to the present work. Genomic-length transcripts carrying altered initiation codons for the p20 or p21 ORFs or carrying a termination codon in the p20 ORF, were inoculated onto N. clevelandii plants (note that the nucleotide substitutions in the p20 ORF are silent mutations with respect to the p21 ORF). Transcripts which lacked the AUG codon for p21 did not produce symptoms or replicate to detectable levels in whole plants and transcripts unable to produce p20 accumulated to high levels but the symptoms were dramatically attenuated (data not shown) and were associated with the appearance of de novo generated defective interfering RNAs (Rochon, 1991). These observations provided further evidence for the independent synthesis of p20 and p21 as distinct proteins and indicate that both are normally produced in vivo (Rochon and Johnston, 1991). 3.5.3 Accumulation of CNV p21 and p20 AUG codon mutants in cucumber protoplasts The phenotypic changes resulting from the above mutations point to an involvement of p20 in some aspect of virus replication and suggest that p21 is associated either with replication or movement of the virus throughout the infected plant. The functions of movement and replication in these mutants cannot be distinguished in whole plants since mutant transcripts might still replicate efficiently yet be unable to spread and therefore not accumulate to detectable levels. Therefore, to assess the replication of these mutants, full-length transcripts in which the AUG codons for p20 and p21 were changed to nonAUG codons (M5201 and M5215, respectively), were transfected into cucumber protoplasts and the accumulation of genomic and subgenomic RNAs analyzed by northern blot. Fig. 3.16 demonstrates that the M5201 mutant (which lacks the AUG codon for p20) accumulates to WT levels in protoplasts whereas the M5215 mutant (which lacks the AUG codon for p21) accumulates in protoplasts Fig. 3.16 Northern blot demonstrating replication of WT, M5215 and M5201 mutant RNA in cucumber protoplasts. Protoplasts were infected with equal amounts of WT and full-length mutant transcripts for the indicated times and one tenth of each sample was analyzed by northern blot using a 32P labeled RNA probe complementary to the entire CNV genome. Bands corresponding to CNV genomic RNA, and the 2.1 and 0.9 kb subgenomic mRNAs are indicated. but not to WT levels (as assessed from repeated experiments; data not shown). The observation that M5215 RNA accumulates to detectable levels in protoplasts but not in whole plants indicates that p21 is involved in cell-to-cell movement of the virus throughout the infected plant. In addition, the accumulation of the p20 AUG codon mutant to WT levels in protoplasts confirms earlier work in whole plants that showed production of p20 does not affect viral RNA accumulation. 3.6 Investigations into the restoration of systemic movement by coat protein deletion derivatives Previous studies have analyzed the role of the p41 coat protein during CNV infection and in particular the requirement for this protein for long-distance movement in plants. In this work, which is also briefly described in section 3.2, CNV mutants PD(-) or NM2, containing deletions in the carboxy-terminal portion of the coat protein gene which encodes the CP protruding domain, were found to be infectious on N. clevelandii but caused a delayed systemic reaction and a smaller lesion phenotype compared to WT virus (McLean et al., 1993; Sit et al., 1995). Passaging of these mutants in plants, however, led to a partial restoration of the rate of systemic movement and to the accumulation of the respective deletion derivatives, CP(-) and ANM2, in which varying amounts of almost the entire coat protein coding region has been deleted (McLean et ai, 1993; Sit et ai, 1995). Transcripts corresponding to both CP(-) and ANM2 replicate and subsequently move systemically in N. clevelandii, establishing that the coat protein is not required for either cell-to-cell movement or systemic spread of CNV. The coat protein is also dispensible for systemic movement in the closely related tombusvirus, TBSV-ch, however, systemic spread of another tombusvirus, CymRSV as well as the more distantly related dianthovirus, red clover necrotic mosaic virus, was demonstrated to be considerably impaired or restricted to certain hosts in the absence of a functional coat protein (Dalmay et ai, 1992; Xiong et al, 1993). An attractive explanation for the accumulation of CNV CP(-) and ANM2 coat protein deletion derivatives and the corresponding restoration of lesion size and systemic movement rate is the production of increased levels of p21 movement protein which could compensate for less efficient systemic spread in these coat protein-less mutants. 3.6.1 Production of p41, p20 and p21 from coat protein deletion mutants As part of a collaborative effort, the potential for the production of restored or increased levels of p21 movement protein in the deletion derivatives, CP(-) and ANM2, was investigated through the in vitro translation of subgenomic transcripts corresponding to these mutants (see Fig. 3.17). The generation of two subgenomic RNAs of ca. 1.0 and 0.9 kb in cucumber protoplasts has been demonstrated for CP(-) (see section 3.2.1) and is again shown in N. clevelandii protoplasts (see Fig. 3.17A). The larger ca. 1.0 kb RNA species, corresponding to the deleted form of the coat protein subgenomic mRNA (designated the "2.1 kb" subgenomic mRNA), is produced in abundance and is hypothesized to be capable of directing the synthesis of the p20 and p21 proteins in addition to those normally synthesized from the smaller 0.9 kb subgenomic mRNA. To assess whether p21 can be produced by the CP(-) and ANM2 deletion derivatives in plants, synthetic subgenomic-length transcripts analogous to those produced by these mutants during infection were generated and translated in wheat germ extracts. Fig. 3.17B shows that WT 2.1 kb subgenomic transcripts (WT2.1sg) direct the synthesis of p41 but not p21 or p20 as expected. In addition, PD(-) subgenomic transcripts initiated from the 2.1 kb subgenomic start site [PD(-)"2.1sg"] gave rise to the predicted 30 kDa sized product corresponding to the deleted form of p41 and also do not produce p21 or p20. Subgenomic transcripts initiated from the 2.1 kb subgenomic start site in the deletion derivatives, CP(-) and ANM2 [CP(-)"2.1sg" and ANM2"2.1sg", respectively] result in the synthesis of products which are identical in size to the p20 and p21 products directed by WT 0.9 kb subgenomic mRNA transcripts. These results indicate that the deleted versions of the coat protein subgenomic mRNA as well as the WT 0.9 kb subgenomic mRNA generated by CP(-) and ANM2 deletion derivatives, all act as templates for the production of p20 and p21 in vitro . Since the 80 Fig. 3.17 Characterization of WT, PD(-) and CP(-) subgenomic RNAs and their in vitro translation products. A. Northern blot analysis of infected N. clevelandii protoplasts. RNA was extracted from mock-inoculated protoplasts or protoplasts infected with WT, PD(-) or CP(-) transcripts 48 hr post transfection. Blots were probed with 32P-labeled nick translated DNA corresponding to the 3' terminus of the CNV genome. The locations of WT and deleted forms of genomic RNA,"2.1 kb" subgenomic and 0.9 kb subgenomic mRNAs are indicated. B. In vitro translation products directed by WT and deletion mutant subgenomic mRNAs. Wheat germ extracts containing [35S]methionine were programmed with endogenous RNA, WT 2.1 kb subgenomic transcript RNA, PD(-) "2.1 kb" subgenomic transcript RNA, WT 0.9 kb subgenomic transcript RNA, ANM2 "2.1 kb" subgenomic transcript RNA, or CP(-) "2.1 kb" subgenomic transcript RNA. The CNV proteins corresponding to the in vitro translation products are indicated on the right. accumulation of PD(-) and CP(-) 0.9 kb subgenomic mRNA in N. clevelandii protoplasts is not obviously lowered compared to WT levels (Fig. 3.17A; see also Fig. 3.3 for results in cucumber protoplasts) it seems likely that the original PD(-) mutation does not affect 0.9 kb subgenomic mRNA production and, therefore, p21 levels, in plants. However, the generation of CP(-) "2.1 kb" and 0.9 kb subgenomic mRNA species, which can both direct the synthesis of p20 and p21 in vitro, may lead to increased production of p21 by the deletion derivative in vivo. It is possible, then, that the selection pressure for the accumulation of CP(-) and ANM2 deletion derivatives in plants is increased production of p21. Higher levels of 21 cell-to-cell movement protein may compensate for the lack of a coat protein and enable an increased rate of "systemic" movement (see Discussion). 3.7 Analysis of translational regulation in the production of p20 and p21 The work described above (see section 3.4) demonstrates that CNV generates a subgenomic mRNA species which is capable of producing two distinct proteins, p20 and p21 from different AUG codons. Since most mRNAs are monocistronic and express only the 5' proximal cistron, however, it was of interest to investigate the production of both p20 and p21 from the same subgenomic mRNA. Other cases in which two proteins are synthesized from extensively overlapping ORFs have been reported and a number of these conform best to translation via Kozak leaky ribosomal scanning. According to this model, during translation some ribosomes scan past the 5' proximal AUG codon due to its unfavorable context and instead initiate translation at a downstream AUG codon. This strategy is also likely for translation of the internally located CNV p20 ORF since the upstream AUG codon for p21 lies in a potentially unfavorable context (lacking a purine in the -3 position but containing a G in the +4 position) for translation by eukaryotic ribosomes (see Fig. 3.14B). The effect of selected nucleotide substitutions surrounding the AUG codon for p21 were therefore investigated to determine which nucleotides most strongly regulate the efficiency of translation in our protoplast system and the effect of these substitutions on expression from the downstream AUG codon for p20 subsequently determined. 3.7.1 Effect of mutations surrounding the AUG codon for p21 To investigate the influence of selected nucleotides flanking the AUG codon for CNV p21 on the efficiency of translation initiation in plant protoplasts, a series of pCGUS constructs was generated which contain a sequence corresponding to the 5' untranslated leader region of the CNV 0.9 kb subgenomic mRNA. The 5' leader sequence, representing the region from the subgenomic start site up to and including the initiation codon for p21 (plus 3 downstream residues) was placed downstream of the CaMV 35S promoter and upstream and in-frame with the p-glucuronidase (GUS) reporter gene (Fig. 3.18A). The nucleotides surrounding the p21 start codon were then modified in the -3, +4 and +5 positions resulting in the generation of 8 pCGUS clones which would give rise to mRNA containing either an A or U in the -3 position, a G or a U in the +4 position and an A or a C in the +5 position (Fig. 3.18B). The ability of the different pCGUS constructs to transiently express GUS in N. plumbaginifolia protoplasts was measured using a kinetic spectrophotometric assay (see section 5.1 of Appendix). The GUS activity directed by pCGUS-wt (which contains sequences corresponding to the authentic 0.9 kb subgenomic mRNA leader) and pCGUS-1 (which contains 2 nucleotide changes corresponding to the extreme 5' end of the leader RNA) were similar (see Fig. 3.19) indicating that the two nucleotide changes introduced for cloning purposes and present in the remaining pCGUS constructs had little impact on expression. pCGUS constructs containing an A in the -3 position (i.e. pCGUS 3 and 8), a G in the +4 position (i.e. pCGUS 1 and 5) or both an A and a G in the -3 and +4 positions (i.e. pCGUS 2 and 6) directed GUS activity levels greater than or equal to WT levels. The GUS activity levels obtained from these constructs ranged from ca. 110% to 150% (+/- 20%) of the levels directed by pCGUS-wt. Repeated experiments using independently prepared plasmid DNA resulted in similar trends in relative GUS activity with pCGUS 1,2,3,5,6, and 8 constructs ranging from ca. 90% to 130% /HVA CNV 0.9 kb Genomic —I p33 | p92 1 P41 H p21 Subgenomic RNA pCGUS-1 CaMV 35S promoter DNA — — — ^ I 3-glucuronidase gene |— — - pCGUS RNA 1 -3 +4+5 pCGUS-wt GAAUCUAACCAAUUCAUGGAA pCGUS-1 GAUCCUAACCAAUUCAUGGAU pCGUS-2 GAUCCUAACCAAAUCAUGGAU pCGUS-3 GAUCCUAACCAAAUCAUGUCU pCGUS-4 GAUCCUAACCAAUUCAUGUAU pCGUS-5 GAUCCUAACCAAUUCAUGGCU pCGUS-6 GAUCCUAACCAAAUCAUGGCU pCGUS-7 GAUCCUAACCAAUUCAUGUCU pCGUS-8 GAUCCUAACCAAAUCAUGUAU Fig. 3.18 Diagrammatic representation of pCGUS constructs used to analyze nucleotides which regulate p21 translation initiation. A. Structure of the CNV genome with the 0.9 kb subgenomic mRNA untranslated leader sequence expanded below. Sequences corresponding to the 0.9 kb mRNA leader including the p21 initiation codon and following codon were placed downstream of the CaMV 35S promoter and upstream and in-frame with the coding region for P-glucuronidase (GUS) in pAGUS-1. B. Sequence of the 5' leader regions of pCGUS-wt and pCGUS 1-8 series transcripts carrying nucleotide sustitutions surrounding the p21 AUG codon which starts the synthesis of GUS. pCGUS-wt transcripts, containing the authentic 0.9 kb subgenomic leader sequence, were used as a reference against which the pCGUS 1-8 series was compared. pCGUS 1-8 transcripts contain sequences corresponding to the 0.9 kb subgenomic leader but with 2 nucleotide substitutions at positions -12 and -13 upstream of the AUG codon (where A is +1) introduced for cloning purposes. Nucleotide substitutions were introduced at the -3, +4 and +5 positions surrounding the initiation sites in the pCGUS 1-8 series such that the transcripts contain either an A or a U in the -3 position, a G or a U in the +4 position and an A or a C in the +5 position. The bent arrow denotes the transcript start sites and the p21 AUG codon is underlined. < < < u < u u u < o o o P p o o p p o o o o a o o o o p p p p P p P p p u< u< u< u< u< u< u< u< u< p p p p p p p p p p P < < p p < p < > •4—" cod -X. 00 D O CD _> '+-> CCj X 0 r 1 i r mock wt 12345678 pCGUS contructs Fig. 3.19 Relative GUS activity directed by pCGUS construct series in protoplasts. N. plumbaginofolia protoplasts were transfected with 20 ug of each pCGUS construct. GUS activites for each construct were measured using a kinetic spectrophotometric assay. The GUS activity directed by pCGUS-wt was arbitrarily assigned the value of 1 and the acitivites for the remaining constructs made relative to 1. The levels shown here represent the means obtained from three replicates of each contract using the same batch of protoplasts. Repeated experiments using independently prepared constructs demonstrated similar trends in expression with the values obtained discussed in the text (see section 5.1.3 of Appendix). The AUG context of each contract is shown above the bar graph with the pCGUS construct number indicated on the x axis. (+/- <20%) of the activity directed by pCGUS-wt (and with less dramatic differences between pCGUS wt and pCGUS 3 and 6; see section 5.1.3 of Appendix). In contrast to the above, pCGUS constructs which did not contain a purine in either the -3 or +4 position (i.e. pCGUS 4 and 7) gave rise to significantly lower levels of GUS activity relative to WT. These levels ranged from ca. 40% (+/- < 10%) of WT levels for pCGUS 4 (which would generate mRNA containing a U in the -3 position and a UA dinucleotide following the AUG) to ca. 60% (+/- <10%) of WT levels for pCGUS 7 (which would direct mRNA with a U in the -3 position and a UC dinucleotide following the AUG codon). In repeated experiments, the GUS activity from pCGUS 4 and 7 was always below that of the other constructs, directing an average of ca. 30% and 50% (+/- <10%) of WT levels, respectively (see Fig. 5.2 in Appendix). pCGUS constructs 3,5, and 6 containing purines in the -3 and +4 positions and a C in the +5 position (the latter postulated to be favorable based on nucleotide sequence comparisons of initiation sites in plant mRNAs; Joshi, 1987; Liitcke et al, 1987; Cavener and Ray, 1991), repeatedly gave rise to GUS activities similar to their counterparts containing an A in the +5 position (i.e. pCGUS 8, 1, and 2). However, pCGUS 7, which contains pyrimidines in the -3 and +4 positions and a C in the +5 position, consistently directed higher levels of GUS activity compared to pCGUS 4 which also contains pyrimidines in the -3 and +4 positions but which has an A in the +5 position (see also Fig. 5.2 in Appendix). Together with the above data, these results indicate that efficient codon selection requires the presence of a purine in either the -3 position or the +4 position (but that it is not necessary for a purine to occupy both positions) and, in addition, indicate that the absence of purines in either of these positions may be partially compensated for by the presence of a C in the +5 position. 3.7.2 Effect of mutations surrounding the p21 AUG codon on initiation from the downstream p20 initiation codon To investigate whether nucleotide substitutions flanking the upstream p21 AUG codon modulate expression from the downstream p20 AUG codon, a sequence corresponding to the leader region of the 0.9 kb subgenomic mRNA (see Fig. 3.20 legend), extending past the p21 AUG codon and including the p20 AUG codon, was used for the generation of a series of pBGUS constructs. This region was placed adjacent to the CaMV 35S promoter, as above, however in this case with the p20 AUG codon in-frame with the GUS coding sequence (Fig 3.20A). Downstream nucleotide substitutions, shown above to modulate translation initiation at the p21 AUG codon, were again introduced such that the resulting transcripts would contain either a G or a U in the +4 position and a C or an A in the +5 position of the p21 AUG codon and the relative GUS activity again determined for each construct (Fig. 3.20B). Fig. 3.21 shows that pBGUS 5 which contains a GC pair in the +4 and +5 positions (shown above to direct slightly higher than WT levels of GUS activity from the p21 AUG codon above; see Fig. 3.19) gave rise to GUS activity levels ca. 10% (+/- 5%) lower than the levels obtained from pBGUS 1 (representing the WT construct). pBGUS 4 and 7 (whose transcripts contain a U in the +4 position and either an A or a C in the +5 position, changes which gave rise to p21-directed GUS activity levels below those of WT above; see Fig. 3.19) each produced GUS levels ca. 30% (+/- 7% or less) higher than the levels produced by pBGUS 1. Again, separately prepared plasmid DNA gave rise to similar levels of GUS activity relative to WT (see Fig. 5.3 in Appendix). The trend in GUS activity obtained from the pBGUS mutant series is inversely related to the variation in GUS levels produced by the pCGUS mutants, above, indicating that the context of the upstream p21 initiation codon influences the efficiency of translation from the downstream p20 initiation codon. A CNV | p20 Genomic —T p33 p92 p41 "H p21 RNA c/vv 1 0.9 to GAAUCUAACCAAUUCAUGGAUACUGAAUACGAACAAGUCAAUAAACCAUGGAA pBGUS-1 CaMV 35S promoter DNA p-glucuronidase gene B pBGUSRNA |-^ +4+5 pBGUS-1 (w/) GAUCCUAACCAAUUCAUGGAUACUGAAUACGAACAAGUC AAUAAACCAUGGAA pBGUS-4 GAUCCUAACCAAUUCAUGUAUACUGAAUACGAACAAGUCAAUAAACCAUGGAA pBGUS-5 GAUCCUAACCAAUUCAUGGCUACUGAAUACGAACAAGUCAAUAAACCAUGGAA pBGUS-7 GAUCCUAACCAAUUCAUGUCUACUGAAUACGAACAAGUCAAUAAACCAUGGAA Fig. 3.20 Diagrammatic representation of pBGUS constructs used to analyze p20 expression. A. Structure of the CNV genome with the 0.9 kb subgenomic mRNA untranslated leader sequence and downstream coding region expanded below. Sequences corresponding to the 0.9 kb mRNA leader extending past the p21 initiation site and including the p20 initiation site and following codon were placed downstream of the CaMV 35S promoter and upstream (with the p20 start site in-frame) of P-glucuronidase (GUS) in pAGUS-1. B. Sequence of the 5' regions of pBGUS 1-4 series transcripts carrying nucleotide sustitutions surrounding the p21 AUG codon . pBGUS 1-4 transcripts contain sequences corresponding to the 0.9 kb subgenomic leader but with 2 nucleotide substitutions at positions -12 and -13 upstream of the AUG codon (where A is +1) introduced for cloning purposes. Nucleotide substitutions were introduced at the +4 and +5 positions following the initiation site in the pBGUS 1-4 series such that the transcripts contain either a G or a U in the +4 position and an A or a C in the +5 position. The bent arrow denotes the transcript start sites and the p20 and p21 AUG codons are underlined. u u < < o o o o o 2 < < < < X cod O x > cd 0^ mock 14 5 7 pBGUS contructs Fig. 3.21 Relative GUS activity directed by pBGUS construct series in protoplasts. N. plumbaginofolia protoplasts were transfected with 20 ug of each pBGUS construct containing nucleotide substitutions surrounding the CNV p21 initiation codon and with the downstream initiation codon for p20 in-frame with GUS. GUS activity resulting from synthesis from the p20 start site of each construct were measured using a kinetic spectrophotometric assay. The GUS activity directed by pCGUS-1 was assigned the value of 1 and the activites for the remaining constructs calculated relative to 1. The levels shown here represent the means obtained from three replicates of each construct using the same batch of protoplasts. Repeated experiments using independently prepared plasmid DNA demonstrated similar trends in expression (see section 5.1.3 in Appendix). The AUG context of each contract is shown above the bar graph with the pBGUS construct number indicated on the x axis. 3.7.3 Effect of codon context on relative production of p20 and p21 in vitro To determine the effect of nucleotide substitutions downstream of the p21 AUG codon on the production of both p20 and p21, subgenomic-length transcripts containing nucleotide changes in the +4 and +5 positions (as in the above) were translated in a cell-free system. Fig. 3.22 shows the in vitro translation products directed in wheat germ extracts programmed with equal amounts of M521/S1, -S4, -S5 or -S7 (M521/S1 represents WT subgenomic RNA transcripts whereas the remaining three constructs correspond respectively to transcripts with mutations UA, GC or UC in the +4 and +5 positions of the p21 AUG codon). The subgenomic-length transcripts carrying these mutations directed similar proportions of p20 and p21 in vitro translation products, with the exception of S4, which consistently directed more p20 compared to p21 translation product. The p21 AUG codon in S4 is followed by a UA dinucleotide which was shown above to be detrimental to expression from the p21 AUG codon but lead to an increase in expression from the downstream p20 AUG codon (see Fig. 3.21). The results presented in the above section are therefore in agreement with the in vitro translation data, however the latter system appears less responsive to changes in codon context. 3.7.4 Effect of leader length of the 0.9 kb subgenomic mRNA on production of p20 and p21 To examine whether the length of the CNV 0.9 kb subgenomic mRNA leader affects the relative production of p20 and p21, constructs were generated which would give rise to transcripts corresponding to either the 0.9 kb subgenomic mRNA (0.9 sg) or would contain an additional 33 nucleotides of 5' non-coding sequence (corresponding to the 5' untranslated leader sequence of the CNV 2.1 kb subgenomic mRNA and a region immediately upstream of the 0.9 kb subgenomic mRNA leader), designated ANM2 sg RNA. In vitro translation products directed by synthetic 0.9 kb subgenomic mRNA and ANM2 extended leader subgenomic mRNA in wheat germ extracts were compared by programming the extracts with increasing < < V u o o o o o O O < < 1—1 in r-- a co 00 co -> O CN (N CN CN a in in m in PQ Fig. 3.22 Zn v/fro translation of 0.9 kb subgenomic RNA transcripts containing mutations downstream of the initiation codon for p20. Wheat germ extracts were programmed with no added RNA (endogenous) or 2 ug each of M521/S1 0.9 kb subgenomic RNA (WT), M521/S4 0.9 kb subgenomic RNA, M521/S5 0.9 kb subgenomic RNA or M521/S7 0.9 kb subgenomic RNA transcripts in the presence of [35S] methionine. The context of the p21 AUG codon is indicated in brackets for each transcript. In vitro translation products were analyzed by SDS-polyacrylamide gel electrophoresis (through a 15% separating gel) and subsequent fluorography. The CNV proteins corresponding to the in vitro translation products are shown on the right. 0.9kbsgRNA ANM2 sg RNA " ANM2 sg RNA 1 i 1 i 1 00 00 oo 00 00 00 00 00 00 3 3 3 3 3 3 3 3 m O O in O o in O o d r-H CN d i-H CN d i—i CN Fig. 3.23 In vitro translation of wild type 0.9 kb subgenomic mRNA transcripts and extended leader ANM2 subgenomic length mRNA transcripts. Wheat germ extracts (Promega) were programmed with 0.5, 1.0 and 2.0 ug synthetic subgenomic transcript RNA in the presence of [35S]methionine. In vitro translation products were electrophoresed through a 15% SDS-PAGE gel and subsequently analyzed by fluorography and autoradiography. A and B are the results of separate in vitro translation experiments. amounts of each transcript (from 0.5 fxg to 2 (ig RNA). Fig. 3.23A indicates that both 0.9 kb subgenomic mRNA and ANM2 subgenomic mRNA direct the synthesis of p20 and p21. However, the relative amounts of products directed by both transcripts differs. While in repeated experiments, 0.9 kb subgenomic mRNA consistently gave rise to ca. equal proportions of p20 and p21 at all three RNA concentrations used, ANM2 subgenomic mRNA consistently directed the synthesis of more p21 relative to p20; further, in Fig. 3.23B, the amount of p20 directed by ANM2 subgenomic mRNA at low RNA concentrations was nearly negligible. 3.8 Trans-complementation assay This section will briefly describe an alternative approach for mapping the promoter for the 0.9 kb subgenomic mRNA, however,.as this proved unsuccessful, only those results which may be useful for future studies will be described. This approach involved the coinfection of (-) strand RNA containing putative subgenomic promoter element(s) with helper virus for the production of a replicase for (-) strand promoter recognition. The putative (-) strand promoter element(s) were fused with sequences corresponding to the coding region for GUS placed in antisense orientation in an attempt to provide an easily assayable system for promoter activity. Since production of GUS could occur only through recognition by the replicase of promoter element(s) on the (-) strand template, and subsequent transcription of (+) sense GUS RNA, quantitation of any resulting GUS activity could provide an indication of promoter activity. A series of four pBTPro constructs were generated which contain varying lengths of sequence corresponding to putative promoter element(s) extending in the 5' direction from the CNV p20 initiation codon (see Fig. 3.24). These sequences were placed upstream and in-frame with the coding region for GUS and the entire region introduced in antisense orientation downstream of the CaMV 35S promoter and upstream of the NOS termination sequence to create the pSGPro series. The rationale behind using CaMV 35S promoter-based constructs was to provide a continuous supply of in vivo generated antisense transcript RNA necessary for the subsequent production of detectable levels of GUS by helper virus complementation. For 93 _£20_ pK2/M5 — p41 p21 i r i r BglU HpallXhoI tA, Ncol Hpal AsuII Sail (BamHl) pBTProBglll GUS X SacIBamHI/BgUI Ncol SacUAsuII Sail pBTProHpall GUS X BamHl Hpall Ncol SacI/AsuII Sail pBTProXhoI GUS X BamHl Xhol Ncol SacI/AsuII Sail pBTProBamHI GUS X BamHl Ncol SacU AsuII Sail pSGPro series CaMV 35S promoter BamHl Sail SacI/AsuII Ncol BamHl SacI Fig. 3.24 Diagrammatic representation of constructs generated for the purpose of mapping the CNV 0.9 kb subgenomic mRNA promoter. A series of four constructs were generated which contain varying amounts of pK2/M5 sequence corresponding to the region upstream and including the CNV p20 initiation codon. These sequences were placed upstream of the coding region for GUS in Bluescript to create pBTPro constructs - BglU, -Hpall, -Xhol and -BamHl. A BamHI-Sall cassette from each pBTPro construct was placed downstream of the CaMV35S promoter and upstream of the NOS termination signal in pAGUS-1 (see section 6.2 in Appendix) creating the pSGPro construct series. Transcripts generated from the pSGPro series would contain putative CNV (-) sense promoter elements upstream of sequences complementary to the coding region for GUS. use as a helper virus, cDNA corresponding to the entire CNV genome was placed downstream of the CaMV 35S promoter and upstream of the NOS termination signal such that the transcripts generated would correspond to capped, polyadenylated CNV RNA. Although these transcripts contain an imperfect 5' terminus (i.e. 5'GGAATTC3' instead of 5GAAATTC3' initiated by pK2/M5 RNA) and presumably a poly(A) tail not normally present on CNV RNA, they were infectious on N. clevelandii plants and accumulated in cucumber protoplasts. Fig. 3.25A demonstrates that p35SCNV in vivo transcribed RNA was able to replicate (as indicated by the presence of subgenomic RNA species) and accumulate in protoplasts over time. However, these results also indicate an apparent lag in the time of appearance of replicatable RNA generated from transfected p35SCNV DNA as compared to that generated from pK2/M5 transcript RNA. Nevertheless, the CaMV 35S promoter-based CNV constructs may prove useful for further studies as they circumvent the need to generate in vitro transcribed RNA. Coinoculation of either pK2/M5 transcript RNA or p35SCNV DNA with the above pSGPro constructs did not result in detectable GUS activity for reasons which are not known at this time. While this approach to mapping the subgenomic was considered worthwhile to investigate, particularly since it would appear to have potential application in mapping a promoter for transcripts encoding essential gene products, additional approaches could be used in the case of the CNV 0.9 kb subgenomic promoter. To determine whether quantitation of GUS activity from genomic length (+) strand RNA would be useful in conjunction with deletion analysis for mapping the 0.9 kb subgenomic mRNA promoter, the CNV p20/p21 coding regions were replaced with that of GUS in one of two locations (resulting in either p35SCNV-GUS/HpaI or /AsuII). Fig. 3.25B shows that the RNA generated from these CaMV35S promoter-based constructs was able to replicate in protoplasts but, as noted above, the accumulation of RNA appeared delayed in comparison to that of pK2/M5 transcript RNA in which the p20/p21 coding regions were similarly replaced with that of GUS. These constructs were also found to direct detectable levels of GUS in protoplasts however due to concerns regarding the stability of the gus gene in CNV RNA, the approach outlined in section 3.2 was eventually favored. Fig. 3.25 Accumulation of CNV RNA from T7- and CaMV 35S promoter-based constructs in protoplasts. A. Accumulation of RNA in protoplasts transfected either with RNA derived from the T7-based pK2/M5 construct or with p35SCNV DNA. B. Accumulation of RNA in protoplasts transfected with p35SCNV constructs or pK2/M5 RNA in which varying amounts of the CNV p20 and p21 protein coding regions were replaced with the coding region for GUS (i.e. p35SCNV-GUS/HpaI, -AsuII, or pK2/M5- GUS/Hpal transcript RNA). Protoplasts were transfected with 5 ug RNA or 20 ug DNA for the indicated times (in hr) and one tenth of each sample was analyzed by northern blotting using a 32P labelled RNA probe complementary to the 3' end of the CNV genome. The arrowheads denote the bands corresponding to CNV genomic or subgenomic mRNAs; RNA species containing the GUS coding region are labelled as "genomic", "2.1 kb" or "0.9 kb" to denote the additional ca. 1.5 kb of coding sequence . Chapter 4 Discussion The work presented herein describes the location of cis-acting signals necessary for the promotion of CNV 0.9 kb subgenomic mRNA synthesis. The bifunctional nature of this subgenomic mRNA was also established and the strategy for the production of two proteins which it encodes, p20 and p21, was investigated. During the course of this work, a third subgenomic RNA of 0.35 kb was identified and, as part of a collaborative study, the function and expression of this RNA species were examined. In addition, studies are described which suggest a function for p21 in the life cycle of CNV and, in combination with the work of other collaborators, aid in the formulation of a hypothesis concerning the restoration of systemic spread and the accumulation of mutants lacking the CNV coat protein. 4.1 Delineation of the promoter for 0.9 kb subgenomic mRNA synthesis 4.1.1 The 0.9 kb subgenomic mRNA core promoter is located between nucleotides -20 and +6 relative to the subgenomic start site Deletion mapping of the promoter region for the CNV 0.9 kb subgenomic RNA has established the location of the promoter to be within a 26 nucleotide region surrounding the subgenomic RNA initiation site (+1). The 5' border of the promoter is situated within a short AU-rich region between nucleotides -10 and -20 and the 3' border extends no further than 6 nucleotides downstream of the transcription start site (see Fig. 4.1). This region was determined to be essential for subgenomic RNA synthesis and from examination of coat protein deletion mutants, sequences upstream of this "core" promoter region do not appear to dramatically influence the strength of the promoter. For comparison, subgenomic RNA production in the alphavirus-like BMV requires a minimum of 20 bases upstream and 16 bases downstream of the subgenomic RNA initiation site. However , WT levels of RNA production CNV 0,9 ggugcagguuGUGUAAAUUAGGGGCUUCUUGAAUCUaac A TBSV 0.9 UAAUUUAGUGUGUCCUGCGAGGGGCCUCUUGAACAAGAC A CymRSV 1.0 GUAGUUGCAUUGCACAGGAAGGGGCUUCUUGAACCUAAC A AMCV 0.9 UAAUUUAGUGAGUCCUGUGAGGGGCCUCUUGAACUAGAC CNV 5' AGAAAUUCU * * * * * * * CNV 0.9 ggugcagguuGUGUAAAUUAGGGGCUUCUUGAA--UCU * *** * ***** * *** CNV 2.1 AGCCCAGCAUCCUUGACUCCGCCGUAGCAUGACCAAGC Fig. 4.1 Sequences surrounding the CNV 0.9 kb subgenomic mRNA promoter and comparison with other putative promoters. A. The CNV 0.9 kb subgenomic promoter and comparison to sequences surrounding the subgenomic start site of the analogous region of other tombusviruses. The subgenomic start site for each viral RNA is indicated with a caret. Sequences which comprise the CNV core promoter as defined in this study are shown in upper case. The underlined sequences correspond to the stop codon for the coat protein. Double asterisks indicate identity between all four sequences and single asterisks identity at three of four positions. B. Comparison of the CNV 0.9 kb subgenomic promoter with sequences surrounding the CNV 2.1 kb coat protein subgenomic mRNA start site and sequences at the 5' terminus of CNV genomic RNA. The caret corresponds to the start sites for the 0.9 kb (Rochon and Johnston, 1991) and 2.1 kb subgenomic mRNA (unpublished data) and the position of the CNV genomic RNA 5' nucleotide. Asterisks indicate nucleotide identity between the 0.9 kb subgenomic promoter and either of the other two sequences. The italicized AG in the CNV 5' sequence are the presumed first and second nucleotides based on analyses of dimer junctions in CNV DI RNAs (Finnen and Rochon, 1995). require sequences extending to at least 74 nucleotides upstream including a poly (A) sequence immediately upstream of the -20 to +16 core promoter; further upstream sequences including an ICR2-like motif (see below) influence RNA3 accumulation (French and Ahlquist, 1987; French and Ahlquist, 1988; Marsh et al, 1988). Likewise, the promoter for the related cucumovirus, cucumber mosaic virus, is located between 70 nucleotides upstream (which includes the ICR2-like motif) and 20 nucleotides downstream of the initiation site (Boccard and Baulcombe, 1993). The sequences necessary for basal subgenomic promoter activity in A1MV are located between nucleotides -26 and +1 relative to the initiation site with additional upstream (extending to nucleotide -136 and including an enhancer element) as well as downstream sequences required for full activity (van der Kuyl et al, 1990; 1991; van der Vossen et al, 1995). One exception to the observation that alphavirus-like subgenomic promoters lie primarily upstream of the transcription initiation site is noted for beet necrotic yellow vein virus RNA 3sub which is situated largely downstream, extending only to position -16 in the 5' direction and to between +100 and +208 in the 3' direction (Balmori et al., 1993). It was noted that a deletion of 41 nucleotides (leaving nine intact nucleotides immediately upstream of the subgenomic mRNA start site) nearly abolishes 0.9 kb subgenomic mRNA synthesis whereas a deletion of 43 nucleotides (leaving seven nucleotides upstream of the start site) appears to partially restore mRNA production (see Fig. 3.4). Comparison of sequences remaining after the XA41 deletion and the XA43 deletion reveals no obvious homology between the area upstream of the deletion site and the 0.9 kb subgenomic mRNA promoter region aside from a G in the -20 position relative to the initiation site which is present in XA43 but not in XA41. However, it is still possible that the partial restoration of 0.9 kb subgenomic RNA promoter activity for XA43 could be explained by a fortuitous juxtaposition of sequence upstream of the deleted region with those contained in the 0.9 kb subgenomic RNA promoter, or alternatively, by an alteration in secondary structure due to the deletion. In addition, the 0.9 kb subgenomic RNA appears to be heterogeneous in length in XA41, XA42 and XA43 infected protoplasts suggesting that the deleted nucleotides are affecting the site at which transcription initiation occurs. Primer-extension studies would be useful to assess this interesting possibility. 4.1.2 The 0.9 kb subgenomic mRNA promoter shares little homology with ICR2-like sequences or other CNV putative cis-acting sequences Extensive analysis of the intercistronic regions of several members of the alphavirus-like supergroup has revealed sequence motifs analogous to the downstream portions of internal control regions (ICR2 or box B regions) of RNA polymerase III promoters located within tRNA genes suggesting fundamental similarities between certain members of this group (French and Ahlquist, 1988; Marsh et al, 1988; Smirnyagina et ai, 1994; see section 1.1.4). The CNV 0.9 kb subgenomic mRNA core promoter was examined for elements or features in common with the ICR2-like motifs found in the cis-acting replication sequences of several members of the alphavirus-like supergroup and obvious similarities were not apparent. The 0.9 kb subgenomic promoter also shares little homology with other putative cis-acting sequences within the CNV genome (i.e., sequences at the 5' terminus of genomic RNA and those surrounding the 2.1 kb subgenomic RNA; see Fig. 4.1). The lack of similarity between the 0.9 kb subgenomic RNA promoter and the region surrounding the transcription initiation site for the 2.1 kb subgenomic RNA may reflect their independent regulation by different trans-acting factors within the replicase complex as has been suggested to be the case for the TMV subgenomic mRNAs (Lehto et al., 1990). Some homology is predicted to occur between subgenomic RNA promoters and sequences at the 5' terminus of the genome since the viral replicase is expected to recognize and interact with specific (-) strand signals for (+) strand RNA synthesis (Pacha et al., 1990; Pogue et al, 1990). Similarities between the transcription start sites of the subgenomic mRNAs and the 5' end of genomic RNA within individual viruses have been noted for other members of the flavivirus-like supergroup, e.g., BYDV-PAV (Kelly et al., 1994) and maize chlorotic mottle virus (Lommel etal., 1991) as well as the alphavirus- like BMV (Marsh and Hall, 1987; Marsh etal, 1989), cowpea chlorotic mottle virus (Allison etal., 1989), cowpea mosaic virus (Boccard and Baulcombe, 1993), A1MV (van der Kuyl et al, 1990) and tobacco rattle virus (Cornelissen etal., 1986; Goulden etal., 1990). 100 4.1.3 The 0.9 kb subgenomic mRNA promoter shares considerable sequence similarity with the putative promoter region in other tombusviruses The core promoter for CNV 0.9 kb subgenomic mRNA synthesis contains significant nucleotide sequence homology to analogous regions in the genomes of other members of the tombusvirus group (see Fig. 4.1). The regions surrounding the 0.9/1.0 kb subgenomic mRNA transcription initiation site of TBSV (Hillman etal, 1989), CymRSV (Grieco etal, 1989a) and U AMCV (Tavazza et al, 1994) each contain a 14 nucleotide AGGGGC /CUCUUGAA element which is identical or near-identical (with the exception of one nucleotide) to nucleotides -11 to +3 relative to the transcription start site of CNV. The 5' border of this region of near-identity between the viral sequences is located one nucleotide upstream of the region remaining after the XA41 deletion, the smallest deletion to noticeably alter 0.9 kb subgenomic RNA accumulation (see Fig. 3.4). This latter observation suggests that the core promoter may be even smaller than the 26 nucleotide region determined by deletion analysis. 4.1.4 Nucleotides immediately surrounding the 0.9 kb subgenomic mRNA start site regulate promoter activity The importance of the core promoter was further demonstrated by the drastically reduced levels of 0.9 kb subgenomic mRNA directed by an M5Bam mutant carrying nucleotide substitutions in the -1, +3 (and +4) positions relative to the transcription start site in protoplasts. In addition, plants inoculated with transcripts containing these nucleotide changes developed only very mild symptoms and were only occasionally systemically infected. Examination of RNA extracted from systemically infected leaves revealed the presence of a substantial amount of 0.9 kb subgenomic mRNA, indicating the ability of this RNA species to accumulate in M5Bam-inoculated plants over time. Subsequent passaging of extract from M5Bam infected plants resulted in the development of symptoms which were less delayed and more severe than those observed in transcript inoculated plants. This partial restoration of systemic symptoms in plants inoculated with passaged material was correlated with the presence of viral RNA carrying a single nucleotide reversion in the 0.9 kb subgenomic promoter region (the presence of which was not detected in transcript inoculated plants). It therefore appears that the presence of a U in the -1 position relative to the transcription start site is important for 0.9 kb subgenomic promoter activity and that its absence is correlated with an altered phenotype and delayed systemic spread. These observations are in agreement with those predicted for a mutant affected in its ability to produce products associated with replication and cell-to-cell movement as is suggested for p20 and p21, respectively (see section 4.3). However, the basis for the restoration of systemic symptoms awaits further investigation in order to exclude the possible contribution of additional mutations as well as to examine the effect of the individual mutations by placing them back into a WT context. 4.2 Characterization of the 0.35 kb subgenomic RNA 4.2.1 A third subgenomic RNA of 0.35 kb is generated during CNV infection Examination of the RNA species generated during CNV infection in protoplasts has identified a third subgenomic RNA of 0.35 kb in addition to the previously characterized 2.1 and 0.9 kb subgenomic mRNAs. Northern blot analyses of RNA extracted from CNV infected leaves and virions demonstrated the 0.35 kb subgenomic RNA contains sequence corresponding exclusively to the 3' terminus of the genome thus excluding the possibility that this RNA species might correspond to a de novo generated defective interfering RNA (C.J. Riviere and D.M. Rochon, personal communication). Primer extension analysis indicated that the transcription initiation site for the 0.35 kb subgenomic RNA is located 70 nucleotides upstream of an AUG codon which may initiate synthesis of a small 32 amino acid protein (pX), however, additional sites were mapped to 87 and 91 nucleotides upstream of the putative pX start site (D.M. Rochon, personal communication). The potential for these upstream sites to be used for transcription initiation in addition to the downstream site is reinforced by the presence of more than one RNA band in the 0.35 kb size range in protoplasts inoculated with WT transcripts (see Fig. 3.6) . In addition, the 0.35 kb subgenomic RNA appears to accumulate late in infection suggesting that this subgenomic RNA, or its potential protein product, may have a role late in CNV replication. However, an alternative explanation, that this RNA species might represent a specific degradation product, has not been excluded. 4.2.2 0.35 kb subgenomic transcripts direct the synthesis of pX in vitro Synthetic transcripts corresponding to the 0.35 kb subgenomic RNA can direct the synthesis of a ca. 3.5 kDa product in vitro which suggests that a protein of this size can also be synthesized in vivo. The 3.5 kDa product corresponds to the size of a protein predicted to occur on the basis of computer assisted comparisons of the 3' terminal regions of several tombusviruses (Boyko and Karasev, 1992). It was noted that a pX-sized protein product was absent in wheat germ extracts programmed with CNV virion RNA (see Fig. 3.13), however, previous in vitro translation experiments using both synthetic transcripts and sucrose gradient fractionated CNV virion RNA have indicated that low molecular weight CNV RNA is capable of directing the synthesis of a pX-sized protein (Johnston and Rochon, 1990). Synthetic transcripts corresponding to the 0.35 kb RNA but lacking the AUG codon for pX also produced a ca. 3.5 kDa in vitro translation product as well as a smaller product of ca. 1.5 kDa. The ca. 3.5 kDa product likely arises from initiation at the nonAUG codon, as demonstrated to occur in animal (Kozak, 1989a; Mehdi et ai, 1990; Boeck and Kolakofsky, 1994) as well as plant cells (Gordon et al., 1992), and the 1.5 kDa product may be initiated from a downstream AUG codon present in the pX ORF. 4.2.3 Mutations in the pX ORF alter infectivity of CNV genomic transcripts Infectivity studies using mutant genomic transcripts indicate that the pX ORF contains either important cw-acting sequences required for replication and/or encodes a protein whose function is essential for replication in plants and protoplasts (C.J. Riviere and D.M. Rochon, personal communication). Synthetic genomic transcripts carrying an altered pX initiation codon accumulated in N. clevelandii plants and protoplasts but produced very mild symptoms on plants compared to WT transcripts whereas genomic transcripts carrying a frameshift mutation failed to replicate in both N. clevelandii plants and protoplasts (C.J. Riviere and D.M. Rochon, personal communication). The difference in the ability of the start codon and frameshift mutants to replicate in N. clevelandii may be explained by the low level of production of the pX protein by the start codon mutant as indicated by the above in vitro translation studies. The possibility that all or part of the pX ORF may have exacting effects on replication cannot be excluded by these data, however, and actually appears likely in light of recent work on a related tombusvirus. As in the present study, the infectivity of mutant CymRSV transcripts was analyzed in order to assess whether or not pX is normally produced during infection (Dalmay et al., 1993). It was found that a CymRSV pX stop codon mutant created by site-directed mutagenesis was capable of replication and produced WT symptoms indicating that the pX protein is not necessary for replication of CymRSV. This result contrasts markedly with the results obtained with the CNV pX frameshift mutant and raises the possibility that the loss of infectivity in this CNV mutant is due to effects in an essential c/s-acting sequence rather than to effects on the production of pX protein. It is also possible that these two functions are not mutually exclusive and that the pX ORF, being located at the extreme 3' terminus of the genome, contains important regulatory elements as well as encodes a protein which is required for some aspect of the CNV infection cycle. 4.3 Functional analysis of CNV proteins 4.3.1 CNV p21 is associated with viral cell-to-cell movement The CNV p21 protein has been suggested to be involved in virus transport based on the detection of limited amino acid sequence similarity with other known or putative movement proteins (Melcher, 1990, personal communication; Mushegian and Koonin, 1993) as well as by the requirement for a movement protein in most plant viruses capable of systemic infection (reviewed in Atabekov and Taliansky, 1990; Citovsky and Zambryski, 1991; Deom et al, 1992). A role for p21 in CNV movement is also consistent with previous studies which demonstrated that genomic transcripts unable to express p21 caused no apparent symptoms and were unable to replicate to detectable levels when inoculated onto plants (Rochon and Johnston, 1991). Because the functions of movement and replication cannot be distinguished in whole plants, genomic transcripts in which the p21 AUG codon was changed to a nonAUG codon were use to inoculate cucumber protoplasts. The accumulation of RNA in protoplasts inoculated with the p21 AUG codon mutant indicate that this protein is not involved in replication and imply that the absence of infection in whole plants inoculated with this mutant is due to a deficiency in cell-to-cell spread of the virus. Thus, p21 meets the only two criteria established so far for plant virus movement proteins, namely that (i) the protein is not a capsid protein and (ii) disruption of the coding sequence of the protein abolishes infection in whole plants but has no effect on virus replication in protoplasts (Mushegian and Koonin, 1993). Recently, the analogous p22 proteins of the related tombusviruses, TBSV and CymRSV, were also reported to be involved in cell-to-cell transport of the virus based on the results of similar analyses (Dalmay et al., 1993; Scholthof et al., 1993). In addition, the movement protein of the distantly related dianthovirus, red clover necrotic mosaic virus, has been demonstrated to cooperatively bind single-stranded nucleic acid (Osman et al., 1992; Xiong et al., 1993), indicating that this protein may form a complex with the viral RNA for passage through the plasmodesmata as has been proposed for the for TMV and other viruses (see introduction; Citovsky et al, 1990). 4.3.2 CNV p20, p21 and p41 are dispensible for RNA accumulation in protoplasts In addition to contributing to the delineation of the 0.9 kb subgenomic mRNA core promoter, the large scale deletion mutants used in this study also demonstrate the dispensable nature of the CNV p41 coat protein, the p21 movement protein, as well as the p20 protein for replication and accumulation of genomic and subgenomic RNAs in protoplasts. The absence of coat protein and movement protein genes might be expected to affect RNA accumulation since their products either encapsidate (in the case of coat protein) or possibly bind viral RNA (if p21 is indeed analogous to other cell-to-cell movement proteins) and therefore function to protect the RNA. However, inoculation of CP(-), lacking almost the entire coat protein coding region, or ANcoI-AsuII, lacking all of the p20 and most of the p21 coding regions, into cucumber protoplasts indicated that these proteins are not essential for RNA accumulation over the time periods used. In addition, experiments in which the AUG codons for either p20 or p21 (Rochon and Johnston, 1991) were changed to non AUG codons demonstrate that, in the absence of these proteins, overall RNA accumulation is not drastically reduced in protoplasts. The results of these experiments, which establish the dispensible nature of particularly p21 and p41 in protoplasts, is in contrast to the requirement for p21 in cell-to-cell movement and coat protein in WT systemic movement. 4.3.3 CNV mutants lacking the coat protein coding region have the potential to overexpress the p21 movement protein Previous studies in which the viability of mutants carrying deletions corresponding in the protruding domain of the CNV coat protein was assessed described the accumulation of deletion derivatives lacking almost the entire coat protein coding region (McLean et al., 1993; Sit et al., 1995). The appearance of the CP(-) and ANM2 coat protein deletion derivatives in PD(-) and NM2 infected plants was associated with a restoration in lesion size and partial restoration in systemic movement rate. The ability of CNV coat protein deletion mutants to move systemically in plants demonstrated the dispensible nature of the coat protein in systemic spread however the small lesion size and reduced rate of systemic movement observed with the original mutants suggested these were defective in some function necessary for efficient replication or movement. During investigation into the basis for the accumulation of these coat protein deletion derivatives, northern blot analysis demonstrated that production of the 0.9 kb subgenomic mRNA relative to genomic RNA appeared unaffected in the original PD(-) mutant, suggesting that the synthesis of movement protein in this mutant was not diminished. Northern blot analysis also indicated that RNA synthesis in the CP(-) deletion derivative was increased relative to that of PD(-), possibly due to an increase in replication rate and/or lack of encapsidation, and revealed the abundant production of a ca. 1.0 kb subgenomic mRNA corresponding to the deleted form of the 2.1 kb coat protein subgenomic mRNA. In vitro translation of the ca. 1.0 kb as well as the 0.9 kb subgenomic mRNA normally generated during CP(-) as well as ANM2 infection (D.M. Rochon, personal communication) demonstrated that both of these RNA species were capable of directing the synthesis of p20 as well as p21 movement protein indicating a potential for these proteins to be overproduced in vivo during CP(-) and ANM2 infections. It is therefore tempting to speculate that the selection pressure for the preferential accumulation of coat protein deletion derivatives in plants is due to their greater capacity for cell-to-cell movement. In addition, it seems possible that the increased rate of systemic movement seen with the deletion derivatives as compared to the original coat protein mutants may actually correspond to increased cell-to-cell movement (via stem cells) rather than true "systemic" movement through the plant vasculature, however, this conclusion awaits further experimentation. An explanation for the initial small lesion size and reduced rate of systemic movement seen with the original mutants also remains unclear. It may be that these mutants are affected in their ability to replicate early in infection (i.e. but eventually accumulate to the WT levels indicated by northern blot analysis) possibly due either to the absence of a ds-acting element necessary for RNA accumulation which is normally present in the protruding domain coding region or to a deleterious effect on RNA accumulation of the nonfunctional form of the coat protein (see Sit etal., 1995). As an interesting aside, it is noted that the 3' border of the coat protein deletion site in ANM2 (Sit et al., 1995) corresponds exactly to the start of the 0.9 kb subgenomic mRNA core promoter region that shares striking similarity with the analogous regions found in other tombusviruses (see Fig. 4.1). This observation further supports the suggestion that the core promoter for the 0.9 kb subgenomic mRNA may be smaller than that determined by deletion analysis (see section 4.1.3). In addition, further in vitro translation studies have established that production of CNV p21 is higher, relative to that of p20, from the deleted form of the coat protein subgenomic mRNA due to the presence of a longer 5' untranslated leader (see section 4.4.4). 4.4 Translation control of CNV p20 and p21 production 4.4.1 The 0.9 kb subgenomic mRNA is bifunctional Previous studies using both authentic and synthetic subgenomic transcripts have demonstrated that the CNV 0.9 kb subgenomic mRNA can serve as the template for the synthesis of both p20 and p21 in vitro (Johnston and Rochon, 1990). This result is in agreement with the nucleotide sequence of CNV which predicts the synthesis of both p20 and p21 from different but extensively overlapping ORFs located at the 3' terminus of the genome. In vitro translation of synthetic 0.9 kb subgenomic transcripts which lack the putative AUG codon for either p20 or p21 established that both proteins are independently initiated from AUG codons in different reading frames and do not arise, for example, by premature termination following initiation at the same AUG codon. In addition, genomic transcripts unable to produce p20 or p21 gave rise to distinctly different phenotypes when inoculated onto plants (Rochon and Johnston, 1991). The demonstration that p20 and p21 are directed from the same subgenomic mRNA in vitro, combined with the observed alteration in symptomatology and RNA accumulation attributed to the absence of these proteins in vivo, provides convincing evidence that they are both produced from a single bifunctional subgenomic mRNA during normal CNV infection. The production of proteins from different but extensively overlapping reading frames has been demonstrated or proposed to occur in a number of viruses (reviewed in Kozak, 1991a) including carnation mottle virus (Guilley etal, 1985), southern bean mosaic virus (Wu et al., 1987), maize chlorotic mottle virus (Nutter et al, 1989), turnip yellow mosaic virus (Keese et al, 1989; Weiland and Dreher, 1989), the plant luteoviruses (reviewed in Martin etal, 1990; see also Tacke etal, 1990; Dinesh-Kumar et al, 1992), cucumber mosaic virus (Ding et al, 1994) and peanut clump furovirus (Herzog et al, 1995). Several hypotheses have been.developed to explain the origin of such overlapping genes; these include gene duplication followed by a merging of coding sequence or the translation of an out-of-frame sequence (termed 'overprinting') to yield a new protein leading to selection of the encoding molecule (Keese and Gibbs, 1992). An interesting consequence for the creation of overlapping genes is the possibility that both genes may be limited in their capacity to become optimally adapted for their functions (Keese and Gibbs, 1992). Thus, for many viruses containing overlapping coding regions, both the presence and maintenance of such an arrangement likely reflects constraints placed on their genomes due to the small size of their capsids. 4.4.2 Efficient initiation codon selection requires purines in either the -3 or +4 position A likely strategy for the production of CNV p20 from the bifunctional 0.9 kb subgenomic mRNA is via leaky ribosomal scanning which would involve some ribosomes scanning past the upstream AUG codon for p21 and initiating translation instead at the downstream AUG codon for p20. The production of p20 appears to conform well to this strategy since the context of the upstream p21 initiation site does not include a purine in the -3 position relative to the AUG codon (Kozak, 1991a). The -3 position has been determined to be an important modulator of translational efficiency in animal cells (Kozak, 1991a,b), however a number of studies have reported variable importance of this position depending upon the nucleotide sequence, the stage of development and the system examined (for examples see Cigan et al, 1988; Bairn and Sherman, 1988; Feng etal, 1991). While the substitution of a purine for a pyrimidine in the -3 position of the preproinsulin initiation site decreased its production by as much as 20 fold in mammalian cells (depending on the remaining context, Kozak, 1984; 1986), the effects of substitutions in the -3 position reported in plant systems have generally been somewhat minor in comparison. Changing the sequence context around the initiation site of a plant viral gene to contain an A instead of a U in the -3 position did not increase expression of that gene in plants to a detectable level (Lehto and Dawson, 1990). However, a simultaneous replacement of nucleotides in the -3 and +4 positions resulted in a 4 fold stimulation of GUS activity in transformed rice cells as well as transgenic tobacco (Taylor etal, 1987; McElroy et al, 1991) and as much as a 9 fold increase in GUS activity in oat protoplasts (Dinesh-Kumar and Miller, 1993). In addition, it has been argued on the basis of in vitro data that it is not the -3 position, but instead the +4 position relative to the AUG codon, which regulates translational efficiency in plants (Liitcke et al., 1987). Therefore, to determine whether the AUG context of CNV p21 influences initiation of translation from the AUG codon for p20 (as would expected for its accession via leaky scanning), the effect of codon context on p21 synthesis was investigated. Examination of the effect of selected nucleotide substitutions surrounding the CNV p21 AUG codon on translational efficiency required the generation a series of pCGUS constructs containing the 0.9 kb subgenomic mRNA leader and p21 AUG codon in-frame with the GUS reporter gene. Transfection of the pCGUS construct series into N. plumbaginofolia protoplasts and determination of the resulting GUS activity indicated a ca. 2 fold increase in production with the substitution of a pyrimidine for a purine in either the -3 or +4 position relative to the p21 initiation codon. For constructs lacking a purine in either the -3 or +4 position, a slight increase in GUS activity was found with the introduction of a C in the +5 position. The independent substitution of nucleotides in these positions demonstrates the similar contributions of the -3 and +4 positions to translational efficiency in plant cells and, in addition, establishes the importance of a C in the +5 position in the absence of a purine in either of these positions. The similar impact on translational activity resulting from purine to pyrimidine changes in the -3 and +4 positions correlates well with statistical analyses of nucleotide frequencies flanking the AUG codons of plant mRNAs (Cavener and Ray, 1991). The frequency of a purine in the -3 position of dicot plant mRNAs is 87% (with an A being 70%) while the preference for a G in the +4 position and a C in the +5 position are 70% and 63%, respectively. While the frequency of a purine in the -3 position upstream of the start codon in vertebrate mRNAs is similar to that for plants (91%), the preference for a G in the +4 position and a C in the +5 position are a considerably lower, 46% and 37%, respectively. It has recently been demonstrated in a rabbit reticulocyte lysate system that only a G in the +4 position is stimulatory suggesting that it is not a purine per se but specifically a G that is necessary for efficient codon selection in plant systems as well (Grunert and Jackson, 1994). An investigation of the contribution of nucleotides downstream of the initiation codon to translational efficiency requires, in some constructs, a change of the second codon. The initial codon in all of the GUS fusions is methionine which is of the stabilizing class of amino acids according to the N-end rule (Bachmair et al, 1986), however, removal by amino-terminal processing could potentially expose different residues. While the residues that could be exposed by such processing may confer different stabilities, the presence of similar amino acids in the second positions of both high and low expressing constructs argues against the GUS activities obtained being due to differences in protein stability. (For example, both pCGUS 4 and pCGUS 8 encode proteins which could contain tyrosine at their amino termini yet the GUS activity directed by these constructs is significantly different; likewise, the proteins encoded by both pCGUS 7 and pCGUS 3 could contain at their amino termini a serine residue which would be expected to confer a long half life as predicted by the N-end rule). In addition, amino-terminal methionines are generally retained in long-lived proteins with destabilizing second residues (Tsunasawa etal, 1985). The second codons chosen were also not unfavorable for their use in plants, for example the UCU codon specifying serine is the most preferred of the six possible codons (with a 25% occurrence) and UAU codon for tyrosine is almost as equally common as UAC in dicots (with occurrences of 43% and 57%, respectively) (Murray et al, 1989). 4.4.3 Accession of the CNV p20 ORF is consistent with leaky ribosomal scanning During the accession of a downstream coding region through leaky ribosomal scanning, the propensity for the second AUG codon to be recognized is increased the further the first AUG codon deviates from the optimal context (Kozak, 1991a,b). Although, from the above analysis of expression from the CNV p21 AUG codon it would appear that in this case the first AUG codon is in a near optimal context (with a G in the +4 position), the effect of changes at this site on expression from the second AUG codon for p20 were examined to investigate the strategy of p20 production. The results indicate a trend of increased translation from the downstream p20 AUG codon when the upstream p21 AUG codon is in an unfavorable context (i.e. followed by a UA or UC pair) as opposed to a favorable context (i.e. followed by a GA or GC pair). Initiation from the internally located AUG codon for p20 is therefore in accordance with its accession by leaky ribosomal scanning, demonstrated to occur in a number of animal viruses (reviewed in Kozak, 1991a) and recently in the plant viruses, barley yellow dwarf luteovirus (Dinesh-Kumar and Miller, 1993) and in vitro for peanut clump furovirus (Herzog et al., 1995). 4.4.4 Leader length of the 0.9 kb subgenomic mRNA contributes to production of p20 via leaky ribosomal scanning In addition to a suboptimal context of the first AUG codon, leaky ribosomal scanning may also be promoted in mRNAs containing a relatively short 5' non-coding leader sequence which might impair the ability of scanning ribosomes to recognize and respond to nucleotide changes surrounding the first AUG codon (Kozak, 1991c) . While the average length of the 5' leaders of both plant and vertebrate mRNAs has been estimated to ca. 80 or 90 nt (Joshi, 1987; Kozak, 1987), many RNA viruses (including a number of plant viruses) have considerably shorter leader sequences of 15 nucleotides or less (Kozak, 1991a,b,c). The effect of leader length on translation from our bifunctional mRNA was analyzed by increasing the 5' non-coding sequence from the authentic 15 nucleotide leader to a longer yet similarly structured 48 nucleotide leader lacking upstream AUG codons or minicistrons which might be deleterious to efficient recognition of the initiating AUG codon (see Fig. 4.2). Expression from the first AUG codon is not only increased with a longer leader but there is a relative decrease in expression from the second AUG codon (see Fig. 3.23). Moreover, in some cases (see Fig. 3.23B), the 112 A B 10 20 30 10 20 30 GAAUCUAACCAA -A C AUACG GACCAAGCAAACACAAACACUUAGG -C AAU UUCAUGG UA UGA GGCUU UUG C AGGUACC AU ACU A UCGAG AAC U ACAUAUCGAGCA AA A GAACA ACAUA CA CAA 60 50 40 3 90 IGUAC C g ICAUG G 40 Ul 8 0 o ^ A G A U AC U A UG > O 7 0 Ch > AC G > o O > >- Fig. 4.2. Predicted secondary structure of the 5' untranslated leader and initial coding region of CNV subgenomic length transcripts. A. Secondary structure of the 15 nt leader and following 49 nucleotide coding region (including the AUG codons for CNV p20 and p21) of wild type 0.9 kb subgenomic mRNA determined by the method of Zuker (1989) using the Wisconsin Sequence Analysis Package by Genetics Computer Group, Inc. (Version 8.0-UNIX). The structure diagrammed has a Gibbs free energy value of -7.7 kcal/mol. B. Secondary structure of the 48 nt leader and following 49 nucleotide coding region (as in A) of an extended leader subgenomic- length mRNA (ANM2). The structure shown has a Gibbs free energy value of -11.3 kcal/mol. The arrow in each diagram indicates the transcription initiation site and the AUG codons for p20 and p21 are underlined. The first 20 nucleotides correspond to the first 20 nucleotides of the CNV coat protein subgenomic mRNA. The following 13 nucleotides correspond to the 13 nucleotides immediately upstream of the 0.9 kb subgenomic start site. The remaining sequence corresponds to the 5' terminus of the 0.9 kb subgenomic mRNA. tendency to scan past the first AUG is almost completely suppressed. A similar phenomenon was shown for a synthetic CAT mRNA with a lengthening of the leader from 3 to 32 nucleotides (Kozak, 1991c) as well as SV-40 16S mRNA and yeast MOD5 mRNA when the leader was increased to greater than 44 and 47 nucleotides, respectively (Sedman et al., 1990; Slusher etal, 1991). It is noted that a portion of the extended leader sequence corresponds to the 5' untranslated region of the viral coat protein subgenomic mRNA which is expected to be a highly efficient messenger. Other plant viral 5' leader regions (including the tobacco mosaic virus "omega" fragment and the leaders of A1MV RNA 4 and potato virus X genomic RNA) have been demonstrated to significantly enhance the efficiency with which a homologous or heterologous mRNA is translated (Gallie et al, 1987a,b; Jobling and Gehrke, 1987). These observations raise the issue of whether the changes in translational efficiency are due to sequence effects, however any increases in synthesis due to this additional nucleotide sequence should be reflected in the overall efficiency with which the encoded protein(s), in this case both p20 and p21, are translated. It is postulated by Kozak (1991c) that the effect of a longer leader is due to a greater capacity to load and/or an ability to slow the movement of scanning 40S ribosomal subunits leading to increased recognition of the first AUG codon. The presence of additional leader sequence is thus expected to affect the frequency with which 40S ribosomal subunits scan past the upstream initiation site rather than to solely affect the efficiency with which both initiation sites are recognized. In addition, in this case the 5' portion of the ANM2 leader does not appear to disrupt the secondary structure of the authentic 0.9 kb subgenomic mRNA leader, as an identical stem loop is retained (see Fig. 4.2). It is therefore proposed that the increase in production of p20 in our longer leader transcripts is due to the influence of leader length rather than of primary sequence. Other factors which promote leaky ribosomal scanning include the absence of appreciable secondary structure downstream of the first AUG codon which might otherwise slow the movement of scanning ribosomes and thus increase its recognition (Kozak, 1990), as well as a second AUG codon in close proximity to the first which is thought to minimize masking of the second AUG codon by elongating ribosomes (Kozak, 1995). The CNV 0.9 kb subgenomic mRNA has only very moderate secondary structure downstream of the first AUG codon although, interestingly, it appears that both AUG codons are sequestered within the stem of a hairpin structure (see Fig. 4.2). Similar sequestering has been proposed for the first of two utilized AUG codons in the mRNAs of other plant viruses including kennedya yellow mosaic tymovirus (Ding et al, 1990) and barley yellow dwarf luteovirus (Dinesh-Kumar and Miller, 1993), the latter of which is predicted to contain an extremely stable stem loop structure. The low stability of the CNV 0.9 kb subgenomic mRNA stem loop structure (with a Gibbs free energy value of -7.7 kcal/mol) would probably not be expected to hinder access to the downstream AUG codon by scanning ribosome complexes as they have been demonstrated to melt moderately stable duplexes (e.g. -30 kcal/mol) but only when located some distance from the cap (Kozak, 1989b). In addition, the first AUG codon is separated from the second AUG codon by less than 30 nucleotides and therefore it is likely that these features, along with a short unstructured region upstream of the first AUG codon, play a role in the efficient expression of the internally located p20 coding region. Such features may then compensate for the 5' proximal p21 AUG codon being in a favorable context for initiation of translation which is not usually the case in mRNAs that employ leaky scanning (Kozak, 1991a). It appears that CNV, with its limited coding capacity and compact genome organization, utilizes a bifunctional mRNA with the first AUG codon in a favorable context but with a short upstream leader and relatively small unstructured region between the first and second AUG codons to achieve high (and possibly coordinated) expression of both 5' proximal and internally located cistrons. 4.5 Concluding Remarks Cucumber necrosis virus represents a very useful and convenient model system for studying the regulation of gene expression in (+) strand RNA plant viruses. The ability to generate highly infectious transcripts from cloned CNV cDNA combined with its small genome size and potential to reach very high titers in infected plants creates a desirable situation in which to examine the production and function of its encoded proteins. CNV utilizes a number of strategies for the expression of its genome, including the generation of subgenomic mRNAs, possible readthrough suppression for production of its replicase, and leaky ribosomal scanning for accession of the downstream p20 ORF of the 0.9 kb subgenomic mRNA. Examination of sequences comprising the core promoter of the 0.9 kb subgenomic mRNA represents the first analysis of a subgenomic promoter from a member of supergroup II of (+) strand RNA viruses and provides insight into the regions that regulate subgenomic promoter activity in related viruses. The importance of codon context and leader length in the translational regulation of CNV p20 and p21 production from the bifunctional 0.9 kb subgenomic mRNA also appears applicable to other systems, particularly those of viral origin where constraints placed upon their genome size likely necessitate compact coding arrangements and versatile expression strategies. In addition to examining certain aspects of the regulation and translation of CNV coding regions, this study also provides some insight into the functions of CNV proteins and their importance in the CNV life cycle. Such fundamental information concerning the organization and expression of plant viral genomes, as well as the functions of the proteins they encode, is essential for an understanding of viral pathogenesis and for the eventual development of effective strategies of vims control. References Abad-Zapatero, C, Abdel-Meguid, S.S., Johnson, J.E., Leslie, A.G.W., and Rayment, I., Rossmann, M.G., Suck, D and Tsukihara, T. (1980). Structure of southern bean mosaic virus at 2.8 A resolution. Nature 286, 33-39. Agranovsky, A.A., Lesemann, D.E., Maiss, E., Hull, R. and Atabekov, J.G. (1995). "Rattlesnake" structure of a filamentous plant RNA virus built of two capsid proteins. Proc. Natl. Acad. Sci. USA, 92, 2470-2473. Ahlquist, P., Bujarski, J.J., Kaesberg, P. and Hall, T.C. (1984a). Localization of the replicase recognition site within brome mosaic virus by hybrid arrested RNA synthesis. Plant Mol. Biol. 3,37-44. Ahlquist, P., French, P., Janda, M. and Loesch-Fries, L.S. (1984b). Multicomponent RNA plant virus infection derived from cloned viral cDNA. Proc. Natl. Acad. Sci. USA 81, 7066-7070. Allison, R.F., Janda, M., and Ahlquist, P. (1989). Sequence of cowpea chlorotic mottle virus RNAs 2 and 3 and evidence of a recombination event during bromovirus evolution. Virology 172, 321-330. Andino, R., Rieckhof, G.E. and Baltimore, D. (1990). A functional ribonucleoprotein complex forms around the 5' end of poliovirus RNA. Cell 63, 369-380. Aoki, S and Takebe, I. (1969). Infection of tobacco mesophyll protoplasts by tobacco mosaic virus ribonucleic acid. Virology 39, 439-448. Atabekov, J.G.and Taliansky, M.E. (1990). Expression of a plant virus-coded transport function by different viral genomes. Adv. Virus Res. 38, 201-248. Atkins, J.F., Weiss, R.B., and Gesteland, R.F. (1990). Ribosome gymnastics - degree of difficulty 9.5, style 10.0. Cell 62,413-423. Atreya, P.L., Atreya, CD. and Pirone, T.P. (1991). Amino acid substitutions in the coat protein result in loss of insect transmissibility of a plant virus. Proc. Natl. Acad. Sci. USA 88, 7887- 7891. Bachmair, A., Finley, D. and Varshavsky, A. (1986). In vivo half-life of a protein is a function of its amino-terminal residue. Science 234, 179-186. Bailey, J.M. and Davidson, N. (1976). Methylmercury as a reversible denaturing agent for agarose gel electrophoresis. Anal. Biochem. 70, 75-85. Bairn, S.G. and Sherman, F. (1988). mRNA structures influencing translation in the yeast Saccharomyces cerevisiae. Mol. Cell Biol. 8, 1591-1601. Balmori, E., Gilmer, D., Richards, K., Guilley, H., and Jonard, G. (1993). Mapping the promoter for subgenomic RNA synthesis on beet necrotic yellow vein virus RNA3. Biochimie 75,517-521. Barrera, I., Schuppli, D., Sogo, J.M. and Weber, H. (1993). Different mechanisms of recognition of Qp plus and minus strand RNAs by Qp replicase. J. Mol. Biol. 232, 512-521. Bastin, M. and Hall, T.C. (1976). Interaction of elongation factor 1 with aminoacylated brome mosaic virus and tRNAs. J. Virol. 20, 117-122. Bawden, F.C, and Pirie, N.W. (1937). The isolation and some properties of liquid crystalline substances from solanaceous plants infected with three strains of tobacco mosaic virus. Proc. R. Soc. 128, 274-320. Beachy, R.N., Zaitlin, M., Bruening, G. and Israel, H.W. (1976). A genetic map for the cowpea strain of TMV. Virology 73, 498-507. Beier, H., Barciszewska, M., Krupp, G., Mitnacht, R. and Gross, H.J. (1984a). UAG readthrough during TMV RNA translation: isolation and sequence of two tNRAsTyr with suppressor activity from tobacco plants. EMBO J. 3, 351-356. Beier, H., Barciszewska, M. and Sickinger, H.-D. (1984b). The molecular basis for the differential translation of TMV RNA in tobacco protoplasts and wheat germ extracts. EMBO J. 3, 1091-1096. Beijerinck, M.W. (1898). Ueber ein contagium vivum fluidum als ursache der fleckenkrankheit der tabaksblatter. Verh. K. Akad. Wet. 6, 3-21 (Dutch). Translation: Concerning a contagium vivum fluidum as a cause of the spot-disease of tobacco leaves, translated by J.Johnson, Phytopath. Class. (1942) 7, 33-52. Benfey, P.N. and Chua, N.-H. (1990). The cauliflower mosaic virus 35S promoter: Combinatorial regulation of transcription in plants. Science 250, 959-966. Berna, A., Gafny R., Wolf, S., Lucas, W.J., Holt, C.A., and Beachy, R.N. (1991). The TMV movement protein: role of the C-terminal 73 amino acids in subcellular localization and function. Virology 182, 682-689. Blumenthal, T. and Carmichael, G.G. (1979). RNA Replication: Function and structure of Qp replicase. Annu. Rev. Biochem. 48, 525-48. Boeck, R., and Kolakovsky, D. (1994). Positions +5 and +6 can be major determinants of the efficiency of non-AUG initiation codons for protein synthesis. EMBO J. 13, 3608-3617. Boccard, F. and Baulcombe, D. (1993). Mutational analysis of ds-acting sequences and gene function in RNA3 of cucumber mosaic virus. Virology 193, 563-578. Bonneville, J.M., Sanfagon, H., Fiitterer, J., and Hohn, T. (1989). Posttranscriptional trans• activation in cauliflower mosaic vims. Cell 59, 1135-43. Boyko, V.P. and Karasev, A.V. (1992). Tombusvirus genome may encode the sixth small protein near its 3' terminus. Virus Genes 6, 143-148. Bradford, M. (1976). A rapid and sensitive method for the quantification of microgram quantities of protein using the principles of protein-dye binding. Anal. Biochem. 72, 248-254. Brault, V. and Miller, W.A. (1992). Translational frameshifting mediated by a viral sequence in plant cells. Proc. Natl. Acad. Sci. USA 89, 2262-2266. Brierley, I., Boursnell, M.E.G., Binns, M.M., Bilimoria, B., Blok, V.C., Brown, T.D.K. and Inglis, S.C. (1987). An efficient ribosomal frameshifting signal in the polymerase-encoding region of the coronavirus IBV. EMBO J. 6, 3779-378 Bruening, G., Beachy, T.N., Scalla, R. and Zaitlin, M. (1976). In vitro and in vivo translation of the ribonucleic acids of a cowpea strain of tobacco mosaic virus. Virology 71,498-517. Bujarski, J.J., Dreher, T.W. and Hall, T.C. (1985). Deletions in the 3' terminal tRNA-like structure of brome mosaic virus RNA differentially affect aminoacylation and replication in vitro. Proc. Natl. Acad. Sci. USA 82, 5636-5640. Bujarski, J.J., Ahlquist, P., Hall, T.C., Dreher, T.W. and Kaesberg, P. (1986). Modulation of replication, aminoacylation and adenylation in vitro and infectivity in vivo of BMV RNAs containing deletions within the multifunctional 3' end. EMBO J. 5, 1769-1774. Burgyan, J., Russo, M., and Gallitelli, D. (1986). Translation of cymbidium ringspot virus RNA in cowpea protoplasts and rabbit reticulocyte lysates. J. Gen. Virol. 67, 1149-1160. Burgyan, J., Grieco, F., and Russo, M. (1989). A defective interfering RNA molecule in cymbidium ringspot virus infections. J. Gen. Virol. 70, 235-239. Calder, V.L. and Palukaitis, P. (1992). Nucleotide sequence analysis of the movement genes of resistance breaking strains of tomato mosaic virus. J. Gen. Virol. 73, 165-168. Campbell, R.N., Lecoq, H, Wipf-Scheibel, C. and Sim, S.T. (1991). Transmission of cucumber leaf spot virus by Olpidium radicale. J. Gen. Virol. 72, 3115-3119. Campbell, R.N., Sim, S.T. and Lecoq, H. (1995). Virus transmission by host-specific strains of Olpidium bornovanus and Olpidium brassicae. Eur. J. Plant Path. 101, 273-282. Carroll, R. and Derse, D. (1993). Translation of equine infectious anemia virus bicistronic tat- rev mRNA requires leaky ribosome scanning of the tat CTG initiation codon. J. Virol. 67, 1433-1440. Cavener, D.R. and Ray, S.C. (1991). Eukaryotic start and stop translation sites. Nucleic Acids Res. 19, 3185-3192. Chang, Y.C., Borja, M., Scholthof, H.B., Jackson, A.O. and Morris, T.J. (1995). Host effects and sequences essential for accumulation of defective interfering RNAs of cucumber necrosis and tomato bushy stunt tombusviruses. Virology 210,41-53. Chen, Z., Stauffacher, C, Li, Y., Schmidt, T., Bomu, W., Kamer, G., Shanks, M., Lomonossoff, G., and Johnson, J.E. (1989). Protein-RNA interactions in an icosahedral virus at 3.0 A resolution. Science 245, 154-159. Chenik, M., Chebli, K., and Blondel, D. (1995). Translation initiation at alternate in-frame AUG codons in the rabies virus phosphoprotein mRNA is mediated by a ribosomal leaky scanning mechanism. J. Virol. 69, 707-712. Cigan, A.M., Pabich, E.K. and Donahue, T.F. (1988). Mutational analysis of the HIS4 translational initiator region in Saccharomyces cerevisiae. Mol. Cell Biol. 8, 2964-2975. Citovsky, V., Knorr, D., Schuster, G. and Zambryski, P. (1990). The p30 movement protein of tobacco mosaic virus is a single-strand nucleic acid binding protein. Cell 60, 637-647. Citovsky,V. and Zambryski, P. (1991). How do plant virus nucleic acids move through intercellular connections? BioEssays 13, 373-379. Citovsky, V., Wong, M.L., Shaw, A.L., Venkataram Prasad, B.V. and Zambryski, P. (1992). Visualization and characterization of tobacco mosaic vims movement protein binding to single- stranded nucleic acids. Plant Cell 4, 397-411. Citovsky, V. and Zambryski, P. (1993). Transport of nucleic acids through membrane channels: snaking through small holes. Annu. Rev. Microbiol. 47, 167-197. Comelissen, G.J.C., Linthorst, H.J.M., Brederode, F.T., and Bol, J.F. (1986). Analysis of the genome structure of tobacco rattle vims strain PSG. Nucleic Acids Res. 14, 2157-2169. Dalmay, T., Rubino, L. and Russo M. (1992). Replication and movement of a coat protein mutant of cymbidium ringspot tombusvirus. Mol. Plant Microbe Int. 5, 379-383. Dalmay, T., Rubino, L., Burgyan, J., Kollar, A. and Russo, M. (1993). Functional analysis of cymbidium ringspot vims genome. Virology 194, 697-704. Dalmay, T., Szittya, G. and Burgyan, J. (1995). Generation of defective interfering RNA dimers of cymbidium ringspot tombusvirus. Virology 207, 510-517. David, C, Gargouri-Bouzid, R. and Haenni, A-L. (1992). RNA replication of plant viruses containing an RNA genome. Prog. Nucleic Acid Res. Mol. Biol. .42, 157-227. Dawson, W.O. and Hilf, M.E. (1992). Host-range determinants of plant viruses. Annu. Rev. Plant Physiol. Mol. Biol. 43, 527-555. Deom, CM., Oliver, M.J. and Beachy, R.N. (1987). The 30-kilodalton gene product of tobacco mosaic vims potentiates vims movement. Science 237, 389-237. Deom, CM., Lapidot, M. and Beachy, R.N. (1992). Plant vims movement proteins. Cell 69, 221-224. Dias, H.F. (1970). Transmission of cucumber necrosis virus by Olpidium cucurbitacearum Barr & Dias. Virology 40, 828-839. Dinesh-Kumar, S.P., Brault, V. and Miller, W.A. (1992). Precise mapping and in vitro translation of a trifunctional subgenomic RNA of barley yellow dwarf vims. Virology 187, 711- 722. Dinesh-Kumar, S.P. and Miller W.A. (1993). Control of start codon choice on a plant viral RNA encoding overlapping genes. Plant Cell 5, 697-692. Ding, S., Keese, P. and Gibbs, A. (1990). The nucleotide sequence of the genomic RNA of kennedya yellow mosaic tymovirus-Jervis Bay isolate: relationships with potex- and carlaviruses. /. Gen. Virol. 71, 925-931. Ding, S.-W., Anderson, B.J., Haase, H.R., and Symons, R.H. (1994). New overlapping gene encoded by the cucumber mosaic vims genome. Virology 198, 593-601. Dreher, T.W., Bujarski, J.J. and Hall, T.C (1984). Mutant viral RNAs synthesized in vitro show altered aminoacylation and replicase template activities. Nature 311, 171-175. Dreher, T.W. and Hall, T.C. (1988a). Mutational analysis of the sequence and structural requirements in brome mosaic vims RNA for minus strand promoter activity. J. Mol. Biol. 210, 31-40. Dreher, T.W. and Hall, T.C. (1988b). Mutational analysis of the tRNA mimicry of brome mosaic vims RNA: sequence and structural requirements for aminoacylation and 3' adenylation. J. Mol Biol 210,41-55. Dreher, T.W., Rao, A.L.N, and Hall, T.C. (1989). Replication in vivo of mutant brome mosaic vims RNAs defective in aminoacylation. J. Mol. Biol. 206, 425-438. Duggal, R., Lahser, F.C. and Hall, T.C. (1994). cw-Acting sequences in the replication of plant viruses with plus-sense RNA genomes. Annu. Rev. Phytopathol. 32, 287-309. Durbin, R.K. and Stollar, V. (1985). Sindbis vims mutants able to replicate in methionine - deprived Aedes albopictus cells. Virology 144, 529-533. Durham, A.C.H., Hendry, D.A., and Von Wechmar, M.B. (1977). Does calcium ion binding control plant vims disassembly. Virology 77, 524-533. Edwards, J.B.D.M., Delort, J. and Mallet, J. (1991). Oligodeoxyribonucleotide ligation to single stranded cDNAs: a new tool for cloning 5' ends of mRNAs and for constructing cDNA libraries by in vitro amplification. Nucleic Acids Res. 19, 5227-5232. Envy C, Schoumacher, F., Jung, C, Gagey, M.-J., Godefroy-Colburn, T. Stussi-Garaud, C. and Berna, A. (1992). An N-proximal sequence of alfalfa mosaic virus movement protein is necessary for association with cell walls in transgenic plants. J. Gen. Virol. 73, 2115-2119. Feng, Y., Gunter, L.E., Organ, E.L. and Cavener, D.R. (1991). Translation initiation in Drosophila melanogaster is reduced by mutations upstream of the AUG initiator codon. Mol. Cell Biol. 11,2149-2153. Finnen, R.L. and Rochon, D.M. (1993). Sequence and structure of defective interfering RNAs associated with cucumber necrosis vims infections. /. Gen. Virol. 74, 1715-1720. Finnen, R.L. and Rochon, D.M. (1995) Characterization and biological activity of DI RNA dimers formed during cucumber necrosis vims coinfections. Virology 207, 282-2867. Fouillot, N., Tlouzeau, S., Rossignol, J.M., and Jean-Jean, O. (1993). Translation of the hepatitis B virus P gene by ribosomal scanning as an alternative to internal initiation. J. Virol. 67,4886-4895. Fraenkel-Conrat, H. and Williams, R.C. (1955). Reconstitution of active tobacco mosaic vims from its inactive protein and nucleic acid components. Proc. Natl. Acad. Sci. USA. 41, 690-698. Fraenkel-Conrat, H. (1956). The role of the nucleic acid in the reconstitution of active tobacco mosaic vims. J. Am. Chem. Soc. 78, 882-883. Francki, R.I.B., Fauquet, CM., Knudson, D.L. and Brown. F., Eds., (1991). Classification and nomenclature of viruses. Fifth report of the International Committee on Taxonomy of Viruses. Archives of Virology (Suppl. 2), 1. French R., and Ahlquist, P. (1987). Intercistronic as well as terminal sequences are required for efficient amplification of brome mosaic vims. J. Virol. 61, 1457-1465. French, R., and Ahlquist, P. (1988). Characterization and engineering of sequences controlling in vivo synthesis of brome mosaic vims subgenomic RNA. J. Virol. 62, 2411-2420. Furuki, I. (1981). Epidemiological studies on melon necrotic spot. Shizuoka Agricultural Experimental Station Technical Bulletin 14. 121 Futterer, J., Kiss-Laszlo, Z. and Hohn, T. (1993). Nonlinear ribosome migration on cauliflower mosaic virus 35S RNA. Cell 73, 789-802. Gafney, R., Lapidot, M., Berna, A., Holt, C.A., Deom, CM. and Beachy, R.N. (1992). Effects of terminal deletion mutations on function of the movement protein of tobacco mosaic virus. Virology 187,499-507. Gallie, D.R., Sleat, D.E., Watts, J.W., Turner, P.C, Wilson, T.M.A. (1987a). The 5'-leader sequence of tobacco mosaic virus RNA enhances the expression of foreign gene transcripts in vitro and in vivo. Nucleic Acids Res. 15, 3257-3273. Gallie, D.R., Sleat, D.E., Watts, J.W., Turner, P.C, Wilson, T.M.A. (1987b). A comparison of eukaryotic viral 5'-leader sequences as enhancers of mRNA expression in vivo. Nucleic Acids Res. 15, 8693-8711. Gallie, D. R. (1993). Posttranscriptional regulation of gene expression of plants. Anna. Rev. Plant Physiol. Plant Mol. Biol. 44, 77-105. Gallitelli, D., Hull, R. and Koenig, R. (1985). Relationships among viruses in the tombusvirus group: nucleic acid hybridization studies. J. Gen. Virol. 66, 1523-1531. Garcia, A., Van Duin, J. and Pleij, C.W.A. (1993). Differential response to frameshift signals in eukaryotic and prokaryotic translational systems. Nucleic Acids Res. 21, 401-406. Gargouri, R., Joshi, R.L., Bob, J.F., Astier-Manifacier, S., and Haenni A-L. (1989). Mechanism of synthesis of turnip yellow mosaic virus coat protein subgenomic RNA in vivo. Virology 171, 386-393. Gesteland, R.F., Weiss, R.B., and Atkins, J.F. (1992). Recoding: Reprogramming genetic decoding. Science 257, 1640-1641. Gierer, A. and Schramm, G. (1956). Infectivity of ribonucleic acid from tobacco mosaic virus. Nature 177, 702-703. Giesman-Cookmeyer, D. and Lommel, S.A. (1993). Alanine scanning mutagenesis of a plant virus movement protein identifies three functional domains. Plant Cell 5, 973-982. Gilmer, D, Allmang, C, Ehresmann, C, Guilley, H, Richards, K., Jonard, G, et al. (1993). The secondary structure of the 5'-noncoding region of beet necrotic yellow vein virus RNA3: evidence for a role in viral RNA replication. Nucleic Acids Res. 21, 1389-1395. Goldbach, R., LeGall, O. and Wellink, J. (1991). Alphavirus-like viruses in plants. Semin.Virol. 2,19-25. Goldbach, R., Wellink, J., van Kammen, A. Kastee., D. and van Lent, J. (1994). Adaptation of positive-strand RNA viruses to plants. Arch. Virol. [Suppl] 9, 87-97. Gordon, K., Futterer, J. and Hohn, T. (1992). Efficient initiation of translation at non-AUG triplets in plant cells. Plant J. 2, 809-813. Goulden, M.G., Lomonossoff, G.P., Davies, J.W., and Wood, K.R. (1990). The complete nucleotide sequence of PEBV RNA2 reveals the presence of a novel open reading frame and provides insights into the structure of tobraviral subgenomic promoters. Nucleic Acids Res. 18, 4507-4512. Grieco, F., Burgyan, J., and Russo, M. (1989a). Nucleotide sequence of the 3'-terminal region of cymbidium ringspot virus RNA. J. Gen. Virol. 70, 2533-2538. Grieco, F., Burgyan, J., and Russo, M. (1989b). The nucleotide sequence of cymbidium ringspot virus RNA. Nucleic Acids Res. 17, 6383. Grieco, P. and Gallitelli, D. (1990). Nucleotide sequence of the 3' terminal region of artichoke mottled crinkle tombusvirus RNA. Nucleic Acids Res. 18, 1300. Grunert, S. and Jackson, RJ. (1994). The immediate downstream codon strongly influences the efficiency of utilization of eukaryotic translation initiation codons. EMBO J. 13, 3618- 3630. Guerineau, F., Lucy, A. and Mullineaux, P. (1992). Effect of two consensus sequences preceeding the translation initiator codon on gene expression in plant protoplasts. Plant Mol. Biol. 18, 815-818. Guilley, H., Carrrington, J.C., Balazs, E., Jonard, G., Richards, K., and Morris, TJ. (1985). Nucleotide sequence and genome organization of carnation mottle virus RNA. Nucleic Acids Res. 13, 6663-6677. Habili, N. and Symons, R.H. (1989). Evolutionary relationship between luteoviruses and other RNA plant viruses based on sequence motifs in their putative RNA polymerases and nucleic acid helicases. Nucleic Acids Res. 17, 9543-9555 Hacker, D.L., Petty, I.T.D., Wei, N. and Morris, TJ. (1992). Turnip crinkle virus genes required for RNA replication and virus movement. Virology 186, 1-8. Hahn, C.S., Strauss, E.G. and Strauss, J.H. (1985). Sequence analysis of three Sindbis virus mutants temperature-sensitive in the capsid protein autoprotease. Proc. Natl. Acad. Sci. USA 82,4648-4652. Hall, T.C., Shih, D.S. and Kaesberg, P. (1972). Enzyme-mediated binding of tyrosine to brome mosaic virus ribonucleic acid. Biochem. J. 129, 969-976. Harrison, B.D. (1987). Plant virus transmission by vectors: Mechanisms and consequences. In "Molecular basis of virus disease", pp. 319-344. Cambridge University Press, Cambridge, England. Harrison, S.C, Olson, A.J., Schutt, CE., Windier, F.K. and Bricogne, G. (1978). Tomato bushy stunt virus at 2.9 A resolution. Nature 276, 368-373. Harrison, S.C. (1983). Virus structure: High-resolution perspectives. Adv. Virus Res. 28, 175- 239. Haseloff, J., Goelet, P., Zimmern, D., Ahlquist, P., Dasgupta, R. and Kaesberg, P. (1984). Striking similarities in amino acid sequence among nonstructural proteins encoded by RNA viruses that have dissimilar genome organization. Proc. Natl. Acad. Sci. USA 81,4358-4362. Hayes, RJ., Brunt, A.A., and Buck, K.W. (1988). Gene mapping and expression of tomato bushy stunt virus. J. Gen. Virol. 69, 3047-3057. Hearne, P.Q., Knorr, D.A., Hillmann, B.I. and Morris, TJ. (1990). The complete genome structure and synthesis of infectious RNA from clones of tomato bushy stunt virus. Virology 177, 141-151. Hershey, A.D. and Chase, M. (1952). Independent functions of viral protein and nucleic acid in growth of bacteriophage. J. Gen. Physiol. 36, 39-56. Hertz, J.M. and Huang, H.V. (1992). Utilization of heterologous alphavirus junction sequences as promoters by Sindbis virus. /. Virol. 66, 857-864. Herzog, E., Guilley, H. and Fritsch, C. (1995). Translation of the second gene of peanut clump vims RNA 2 occurs by leaky scanning in vitro .Virology 208, 215-225. Hillman, B.L, Carrington, J.C, and Morris, T.J. (1987). A defective interfering RNA that contains a mosaic of a plant vims genome. Cell 51, 427-433. Hillman, B.L, Hearne, P., Rochon, D., and Morris, T.J. (1989). Organization of tomato bushy stunt vims genome: characterization of the coat protein gene and the 3' terminus. Virology 169, 42-50. Hogle, J.M., Maeda, A., and Harrison, S.C. (1986). J. Mol. Biol. 191, 625-638. Holland, J.J., Spindler, K., Horodyski, F., Grabau, E., Nichol, S. and VandePol, S. (1982). Rapid evolution of RNA genomes. Science 215, 1577-1585. Hopper, P., Harrison, S.C. and Sauer, R.T. (1984). Structure of tomato bushy stunt virus. V. Coat protein sequence determination and its structural implications. J. Mol. Biol. 177, 701-713. Horvath, CM., Williams, M.A. and Lamb, R.A. (1990). Eukaryotic coupled translation of tandem cistrons: identification of the influenza B vims BM2 polypeptide. EMBO J. 9, 2639- 2647. Hsaio, K. (1991). A fast and simple procedure for sequencing double stranded DNA with Sequenase. Nucleic Acids Res. 19, 2878. Hughes, S.S. (1977). The Vims: A History of the Concept, Heinemann Educational Books, London. Hull, R. (1978). The stabilization of the particles of turnip rosette vims. III. Divalent cations. Virology 89,418-422. Ishikawa, M., Meshi, T., Ohno, T. and Okada, Y. (1991). Specific cessation of minus strand RNA accumulation at an early stage of tobacco mosaic virus infection. J. Virol. 65, 861-868. Ivanovski, D.I. (1892). On two diseases of tobacco. Sel. Khoz. Lesov. 169, 108-121 (Russian). Translation: In S.S. Hughes The origins and development of the concept of the virus in the late ninetheenth century, doctoral dissertation in the History and Philosophy of Science, University of London, 1972. Jacks, T. and Varmus, H.E. (1985). Expression of the Rous sarcoma vims pol gene by ribosomal frameshifting. Science 230, 1237-1242. Jacks, T., Madhani, H.D., Masiarz, F.R. and Varmus, H.E. (1988). Signals for ribosomal frameshifting in the Rous sarcoma vims gag-pol region. Cell 55,447-458. Jackson, R.J., Hunt, S.L., Gibbs, CL. and Kaminski. (1994). Internal initiation of translation of picornavirus RNAs. Mol. Biol. Rep. 19, 147-159. Jaspars, E.M. (1985). Interaction of alfalfa mosaic virus nucleic acid and protein. In "Molecular Plant Virology" (J.W. Davies, Ed.), Vol I, pp. 155-221. CRC Press, Boca Raton, FL. 124 Jefferson, R.A., Burgess, S.M. and Hirsh, D. (1986). p-glucuronidase from Escherichia coli as a gene-fusion marker. Proc. Natl. Acad. Sci. USA 83, 8447-8451. Jobling, S.A. and Gehrke, L. (1987). Enhanced translation of chimaeric messenger RNAs containing a plant viral untranslated leader sequence. Nature 325, 622-625. Johnston, J.C, and Rochon, D.M. (1990). Translation of cucumber necrosis vims RNA in vitro. J. Gen. Virol. 71, 2233-2241. Jones, J.D.G., Dean, C, Gidoni, D., Gilbert, D., Bond-Nutter, D., Lee R., Bedbrook, J. and Dunsmuir, P. (1988). Expression of bacterial chitinase protein in tobacco leaves using two photosynthetic gene promoters. Mol. Gen. Genet. 212, 536-542. Joshi, CD. (1987). An inspection of the domain between putative TATA box and translation start site in 79 plant genes. Nucleic Acids Res. 15, 6643-6653. Kaesberg, P. (1987). Organization of tripartite plant vims genomes: The genome of brome mosaic vims. In "The Molecular Biology of the Positive Strand RNA Viruses" pp 219-236. Academic Press, London. Kajitani, M. and Ishiama, A. (1991). Identification and sequence determination of the host factor gene for bacteriophage Qp. Nucleic Acids Res. 19, 1063-1066. Kao, CC, Quadt, R., Hershberger, R.P. and Ahlquist, P. (1992). Brome mosaic vims RNA replication proteins la and 2a form a complex in vitro. J. Virol. 66, 6322-6329. Keese, P., Mackenzie, A. and Gibbs, A. (1989). Nucleotide sequence of an Australian isolate of turnip yellow mosaic tymovirus. Virology 172, 536-546. Keese, P.K. and Gibbs, A. (1992). Origins of genes: "Big bang" or continuous creation? Proc. Natl. Acad. Sci. USA 89, 9489-9493. Kelly, L., Gerlach, W.L., and Waterhouse, P.M. (1994). Characterization of the subgenomic RNAs of an Australian isolate of barley yellow dwarf luteovims. Virology 202, 565-573. Kim, K.H. and Lommel, S.A. (1994). Identification and analysis of the site of -1 frameshifting in red clover necrotic mosaic vims. Virology 200, 574-582. Knorr, D.A.,Mullin, R.H., Heame, P.Q., and T.J. Morris. (1991). De novo generation of defective interfering RNAs of tomato bushy stunt vims by high multiplicity passage. Virology 181, 193-202. Koenig, R. and Gibbs, A. (1986). Serological relationships among tombusviruses. J. Gen. Virol. 67, 75-82. Kohl, R.J. and Hall, T.C. (1974). Aminoacylation of RNA from several viruses: Amino acid specificity and differential activity of plant, yeast and bacterial synthetases. J. Gen. Virol. 25, 257-261. Kollar, A. and Burgyan, J. (1994). Evidence that ORF 1 and 2 are the only vims-encoded replicase genes of cymbidium ringspot tombusvirus. Virology 201, 169-172. Koonin, E.V. and Gorbalenya, A.E. (1989). Evolution of RNA genomes: Does the high mutation rate necessitate high rate of evolution of viral proteins? J. Mol. Evol. 28, 524-527. Koonin, E.V., Mushegian A.R., Ryabov E.V. and Dolja, V.V. (1991). Diverse groups of plant DNA and RNA viruses share related movement proteins that may posses chaperone-like activity. J. Gen. Virol. 72, 2895-2903. Koonin, E.V. and Dolja, V.V. (1993). Evolution and taxonomy of positive-strand RNA viruses: implications of comparative analysis of amino acid sequences. Crit. Rev. Biochem. Mol. Biol. 28, 375-430. Kozak, M. (1984). Point mutations close to the AUG initiator codon affect the efficiency of translation of rat preproinsulin in vivo. Nature 308, 241-246. Kozak, M. (1986). Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44, 283-292. Kozak, M. (1987). An analysis of 5'-noncoding sequences from 699 vertebrate messenger RNAs. Nucleic Acids Res. 15,8125-8132. Kozak, M. (1989a). Context effects and inefficient initiation at non-AUG codons in eucaryotic cell-free translation systems. Mol. Cell Biol. 9, 5073-5080. Kozak, M. (1989b). Circumstances and mechanisms of inhibition of translation by secondary structure in eucaryotic mRNAs. Mol. Cell Biol. 9, 5134-5142. Kozak, M. (1990). Downstream secondary structure facilitates recognition of initiator codons by eukaryotic ribosomes. Proc. Nad. Acad. Sci. USA 87, 8301-8305. Kozak, M. (1991a). An analysis of vertebrate mRNA sequences: intimations of translational control. J. Cell Biol. 115, 887-903. Kozak, M. (1991b). Structural features in eukaryotic mRNAs that modulate the initiation of translation. J. Biol. Chem. 266, 19867-19870. Kozak, M. (1991c). A short leader sequence impairs the fidelity of initiation by eukaryotic ribosomes. Gene Expression 1, 111-115. Kozak, M. (1995). Adherence to the first-AUG rule when a second AUG codon follows closely upon the first. Proc. Natl. Acad. Sci. USA. 92, 2662-2666. Kuhlemeier, C. (1992). Transcriptional and post-transcriptional regulation of gene expression in plants. Plant Mol. Biol. 19, 1-14. Kujawa, A.B., Drugeon, G., Hulanicka, D., and Haenni, A.-L. (1993). Structural requirements for efficient translational framshifting in the synthesis of the putative viral RNA-dependent RNA polymerase of potato leafroll virus. Nucleic Acids Res. 21, 2165-2171. Kunkel, T.A., Roberts, J.D., and Zakour, R.A. (1987). Rapid and efficient site-specific mutagenesis without phenotype selection. In "Methods in Enzymology" (R. Wu, Ed.), Vol. 154, pp. 367-382. Academic Press, New York. Laakso, M.M. and Heaton, L.A. (1993). Asp —> Asn substitutions in the putative calcium- binding site of the turnip crinkle virus coat protein affect virus movement in plants. Virology 197, 774-777. Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lai, M.M., Baric, R.S., Brayton, P.R., and Stohlman, S.A. (1984). Characterization of leader RNA sequences on the virion and mRNAs of mouse hepatitis virus, a cytoplasmic RNA virus. Proc. Natl. Acad. Sci. USA 81, 3626-3630. Lai, M.M. (1990). Coronavirus - organization, replication and expression of genome. Annu. Rev. Microbiol. 44, 303-333. Lai, M.M.C. (1992). RNA recombination in animal and plant viruses. Microbiol Rev. 56, 61-79. Lain, S., Riechmann, J.L., and Garcia, J.A. (1990). RNA helicase: A novel activity associated with a protein encoded by a positive strand RNA virus. Nucleic Acids Res. 18, 7003-7006. Lain, S., Martin, M.T., Riechmann, J.L., and Garcia, J.A. (1991). Novel catalytic activity associated with positive-strand RNA virus infection: nucleic acid-stimulated ATPase activity of the plum pox potyvirus helicase-like protein. J. Virol. 65, 1-6. Lazzarini, R.A., Keene, J.D. and Schubert, M. (1981). The origins of defective interfering particles of negative-strand RNA viruses. Cell 26, 145-154. Lehto, K. and Dawson, W.O. (1990). Changing the start codon context of the 30K gene of tobacco mosaic virus from "weak" to "strong" does not increase expression. Virology 174, 169-176. Lehto, K., Grantham, G.L., and Dawson, W.O. (1990). Insertion of sequences containing the coat protein subgenomic RNA promoter. Virology 174, 145-157. Leisner, S.M. and Howell, S.H. (1993). Long distance movement of viruses in plants. Trends in Microbiol. 1,314-317. Levis, R., Schlesinger, S., and Huang, H.V. (1990). Promoter for Sindbis virus RNA-dependent subgenomic RNA transcription. J. Virol. 64, 1726-33. Li, X.H., Heaton, L.A., Morris, T.J., and Simon A.E. (1989). Turnip crinkle virus defective interfering RNAs intensify viral symptoms and are generated de novo. Proc. Natl. Acad. Sci. USA 86, 9173-9177. Li, X.H., and Simon, A.E. (1991). In vivo accumulation of a turnip crinkle virus defective interfering RNA is affected by alterations in size and sequence. J. Virol. 65,4582-4590. Lin, C. and Lo, S.J. (1992). Evidence for involvement of a ribosomal leaky scanning mechenism in the translation of the hepatitis B virus pol gene from the viral pregenome RNA.Virology 188, 342-352. Liu, D.X. and Inglis, S.C. (1992). Internal entry of ribosomes on a tricistronic mRNA encoded by infectious bronchitis virus. J. Virol. 66, 6143-6154. Lommel, S.A., Kendall, T.L., Xiang, Z., and Nutter, R.C. (1991). Identification of the maize chlorotic mottle virus capsid protein cistron and characterisation of its subgenomic messenger RNA. Virology 181, 382-385. Lomonossof, G.P. and Wilson, T.M.A. (1985). Structure and in vitro assembly of tobacco mosaic virus. In "Molecular Plant Virology" (J.W. Davies. Ed.), Vol. I, pp. 43-83. CRC Press, Boca Raton, Florida. Long, D.M. and Uhlenbeck, O.C. (1993). Self-cleaving catalytic RNA. FASEB J., 7, 25-30. Lutcke, H.A., Chow, K.C. Mickel, F.S., Moss, K.A., Kern, H.F. and Scheele, GA. (1987). Selection of AUG initiation codons differs in plants and animals. EMBO J. 6,43-48. Macejak, D.G. and Sarnow, P. (1991). Internal initiation of translation mediated by the 5' leader of a cellular mRNA. Nature 353, 90-94. Marsh, L.E. and Hall, T.C. (1987). Evidence implicating a tRNA heritage for the promoters of (+) strand RNA synthesis in brome mosaic vims and related viruses. The Evolution of Catalytic Function. Cold Spring Harbor Symp. Quant. Biol. 52, 331-341. Marsh, L.E., Dreher. T.W., and Hall, T.C. (1988). Mutational analysis of the core promoter and modulator sequences of the BMV RNA3 subgenomic promoter. Nucleic Acids Res. 16, 981- 995. Marsh, L.E., Pogue, G.P., and Hall., T.C. (1989). Similarities among plant vims (+) and (-) RNA termini imply a common ancestry with promoters of eukaryotic tRNAs. Virology 172, 415-427. Marsh, L.E., Huntley, CC, Pogue, G.P., Connell, J.P. and Hall, T.C. (1991). Regulation of (+):(-)-strand asymmetry in replication of brome mosaic vims RNA. Virology 182, 76-83. Martelli, G.P., Gallitelli, D., and Russo, M. (1988). Tombusvimses. In "The Plant Viruses" (R. Koenig, Ed.), Vol. 3, pp. 13-72. Plenum, New York. Martelli, G.P., Russo, M., and Gallitelli, D. (1989). Tombusvirus group. AAB Descr. Plant Vimses No. 352. Martin, R.R, Keese, P.K., Young, M.J., Waterhouse, P.M. and Gerlach, W.L. (1990). Evolution and molecular biology of luteoviruses. Annu. Rev. Phytopathol. 28, 341-363. Matthews, R.E.F. (1991). Plant Virology. Third Edition. Academic Press. San Diego Mayo, M.A. (1987). A comparison of the translation strategies used by bipartite genome, RNA plant vimses. In "The Molecular Biology of the Postive Strand RNA Vimses "pg. 177-205. Academic Press, London. McBratney, S., Chen. C.Y. and Sarnow, P. (1993). Internal initiation of translation. Curr. Opinion Cell Biol. 5, 961-965. McElroy, D., Blowers, A.D., Jenes, B. and Wu, R. (1991). Construction of expression vectors based on the rice actin 1 (Act 7)5' region for use in monocot transformation. Mol. Gen. Genet. 231, 150-160. McKeen, CD. (1959). Cucumber necrosis vims. Can. J. Bot. 37, 913-925. McLean, M.A., Hamilton, R.L, and Rochon, D.M. (1993). Symptomatology and movement of a cucumber necrosis vims mutant lacking the coat protein protruding domain. Virology 193, 932-939. McLean, M.A., Campbell, R.N., Hamilton, R.L, and Rochon, D.M. (1994). Involvement of the cucumber necrosis vims coat protein in the specificity of fungus transmission by Olpidium bornovanus. Virology 204, 840-842. Mehdi, H., Ono, E. and Gupta, K.C. (1990). Initiation of translation at CUG, GUG and ACG codons in mammalian cells. Gene 91, 173-178. Melcher, U. (1990). Similarities between putative transport proteins of plant viruses. J. Gen. Virol. 71, 1009-1018. Meyer, F., Weber, H. and Weissman, C. (1981). Interactions of Qp replicase with Qp RNA. /. Mol. Biol. 153, 631-660. Mi, S., Durbin, R., Huang, H.V., Rice CM. and Stollar, V. (1989). Association of the Sindbis virus RNA methyltransferase activity with the nonstructural protein nsPl. Virology 170, 385- 391. Mi S. and Stollar, V. (1991). Expression of Sindbis virus nsPl methyltransferase activity in Escherichia coli. Virology 184, 423-427. Miller, W.A., Dreher, T., and Hall, T. (1985). Synthesis of brome mosaic virus subgenomic RNA in vitro by internal initiation on (-) sense genomic RNA. Nature 313, 68-70. Morch, M.C and Haenni, A.L. (1987). Organization of plant virus genomes that comprise a single RNA molecule. In "The Molecular Biology of the Positive Strand RNA Viruses" pp. 153-175. Academic Press, London. Morch, M.D., Joshi, R.L., Denial, T.M. and Haenni A.L. (1987). A new "sense" RNA approach to block viral RNA replication in vitro. Nucleic Acids Res. 15, 4123-4130. Morris, T.J., and Carrington, J.C. (1988). Carnation mottle virus and viruses with similar properties. In "The Plant Viruses" (R. Koenig, Ed.), Vol. 3, pp. 73-112. Plenum Publishing Corporation, New York. Munemitsu, S.M. and Samuel, CE. (1988). Biosynthesis of reovirus-specified polypeptides: Effect of point mutation of the sequences flanking the 5' proximal AUG initiation codons of the reovirus S1 and S4 genes on the efficiency of mRNA translation. Virology 163, 643-646. Murray, E.E., Lotzer, J. and Eberle, M. (1989). Codon usage in plants. Nucleic Acids Res. 17, 447-493. Mushegian, A.R. and Koonin, E.V. (1993). Cell-to-cell movement of plant viruses. Arch. Virol. 133, 239-257. Naleway, J.J. (1992). Histochemical, spectrophotometric, and fluorometric GUS substrates. In "GUS protocols: Using the GUS gene as a reporter of gene expression" (S.R. Gallagher, Ed.), pp 61-75. Academic Press, San Diego. Negrutiu, I., Shillito, R., Potrykus, I., Biasini, G. and Sala, F. (1987). Hybrid genes in the analysis of transformation conditions I. Setting up a simple method for direct gene transfer in plant protoplasts. Plant Mol. Biol. 8, 363-373. Nesters, H.G.M. and Strauss, J.H. (1990). Defined mutations in the 5' nontranslated sequence of Sindbis virus RNA. /. Virol. 64, 4162-4168. Nutter, R.C, Scheets, K., Panganiban, L.C. and Lommel, S.A. (1989). The complete nucleotide sequence of the maize chlorotic mottle virus genome. Nucleic Acids Res. 17, 3163-3176. Oberg, B. and Philipson, L. (1972). Binding of histidine to tobacco mosaic virus RNA. Biochem. Biophys. Res. Commun. 48, 927-932. Osman, T.A.M., Hayes, R.J., and Buck, K.W. (1992). Cooperative binding of the red clover necrotic mosaic movement protein to single stranded nucleic acids. J. Gen. Virol. 73, 223-227. 129 Pacha, R.F., Allison, R.F., and Ahlquist, P. (1990). cis-Acting sequences required for in vivo amplification of genomic RNA3 are organized differently in related bromovimses. Virology 174,436-443. Pacha, R.F. and Ahlquist, P. (1992). Substantial portions of the 5' and intercistronic noncoding regions of cowpea chlorotic mottle virus RNA3 are dispensible for systemic infection but influence viral competitiveness and infection pathology. Virology 187, 298-307. Paigen, K. (1989). Mammalian beta-glucuronidase: Genetics, molecular biology, and cell biology. Prog. Nucleic Acid Res. Mol. Biol. 37, 155-205. Palmenberg, A.C. (1990). Proteolytic processing of picornaviral polyprotein. Annu. Rev. Microbiol. 44, 603-623. Pelletier, J. and Sonenberg, N. (1988). Internal initiation of eukaryotic mRNA directed by a sequence derived from poliovims RNA. Nature (London) 334, 320-325. Pinck, M., Chan, S.-K., Genevaux, M., Hirth, L. and Duranton, H. (1972). Valine specific tRNA-like stuctures in RNAs of two vimses of turnip yellow mosaic vims group. Biochimie 54, 1093-1094. Pogue, G.P. Marsh, L.E., and Hall, T.C. (1990). Point mutations in the ICR2 motif of brome mosaic vims RNAs debilitate (+) strand replication. Virology 178, 152-160. Pogue, G.P. and Hall T.C. (1992). The requirement for a 5' stem-loop structure in brome mosaic vims replication supports a new model for viral (+) strand RNA initiation. J. Virol. 66, 674-684. Pogue, G.P., Marsh, L.E., Connell, J.P. and Hall, T.C. (1992). Requirement for ICR-like sequences in the replication of brome mosaic vims genomic RNA. Virology 188, 742-753. Pogue, G.P., Huntley, CC. and Hall, T.C. (1994). Common replication strategies emerging from the study of diverse groups of positive-strand RNA vimses. Arch. Virol. [Suppl] 9, 181- 194. Pollard, J.W., Walker, J.M. (1990). Plant cell and tissue culture. Humana Press, New Jersey. Prody, G.A., Bakos, J.T., Buzayan, J.M., Schneider, LR. and Bruening, G. (1986). Autolytic processing of dimeric plant vims satellite RNA. Science 231, 1577-1580. Pmfer, D., Tacke, E., Schmitz, J., Kull, B., Kaufmann, A., and Rohde, W. (1992). Ribosomal frameshifting in plants: a novel signal directs the -1 frameshift in the expression of the putative viral replicase of potato leafroll luteovirus. EMBO J. 11, 111 1-1117. Quadt, R., Kao, CC, Browning, K., Hershberger, R. and Ahlquist, P. (1993). Characterization of a host protein associated with brome mosaic vims RNA dependent RNA polymerase. Proc. Natl. Acad. Sci. USA 90, 1498-1502. Quadt, R., Ishikawa, M., Janda, M. and Ahlquist, P. (1995). Formation of brome mosaic vims RNA dependent RNA polymerase in yeast requires coexpression of viral proteins and viral RNA. Proc. Natl. Acad. Sci. USA 92, 4892-4896. Raju, R. and Huang, H.V. (1991). Analysis of Sindbis vims promoter recognition in vt'vo,using novel vectors with two subgenomic mRNA promoters. J. Virol. 65, 2501-2510. Rao, A.L.N., Dreher, T.W., Marsh, L.E. and Hall, T.C. (1989). Telomeric function of the tRNA-like structure of brome mosaic virus RNA. Proc. Natl. Acad. Sci. USA 84, 7383-7387. Rao, A.L.N, and Hall, T.C. (1991). Interference in trans with brome mosaic virus replication by RNA-2 bearing aminoacylation-deficient mutants. Virology 180, 16-22. Riviere, C.J., Pot, J., Tremaine, J.H., and Rochon, D.M. (1989). Coat protein of the carmovirus melon necrotic spot is more similar to those of tombusviruses than to carmoviruses. J. Gen Virol.10, 3033-3042. Riviere, C.J. and Rochon, D.M. (1990). Nucleotide sequence and genomic organization of melon necrotic spot virus. 7. Gen. Virol. 71, 1887-1896. Rochon, D.M., and Tremaine, J.H. (1988). Cucumber necrosis virus is a member of the tombusvirus group. J. Gen. Virol. 69, 395-400. Rochon, D.M., and Tremaine, J.H. (1989). Complete nucleotide sequence of the cucumber necrosis virus genome. Virology 169, 251-259. Rochon, D.M. (1991). Rapid de novo generation of defective interfering RNA from cucumber necrosis virus mutants which do not express the 20 kDa non-structural protein. Proc. Natl. Acad. Sci. USA 88, 11153-11157. Rochon, D.M., and Johnston, J.C. (1991). Infectious transcripts from cloned cucumber necrosis virus cDNA: evidence for a bifunctional subgenomic mRNA. Virology 181, 656-665. Rochon, D.M., Johnston, J.C. and Riviere, C.J. (1991). Molecular analysis of the cucumber necrosis virus genome. Can. J. Plant Path. 13, 142-154. Rohde, W., Gramstst, A., Schmitz, J., Tacke, E., and Prufer, D. (1994). Plant viruses as model systems for the study of non-canonical translation mechanisms in higher plants. /. Gen. Virol. 75, 2141-2149. Rohll, J.C, Holness, C.L., Lomonossoff, G.P. and Maule, A.J. (1993). 3'-terminal nucleotide sequences important for the accumulation of cowpea mosaic virus M-RNA. Virology 193, 672- 679. Rossmann, M.G. and Johnston, J.E. (1989). Icosahedral RNA virus structure. Annu. Rev. Biochem. 58, 533-573. Russo, M., Burgyan, J., Carrington, J.C, Hillman, B.I., and Morris, T.J. (1988). Complementary DNA cloning and characterization of cymbidium ringspot virus RNA. J. Gen. Virol. 69,401-406. Russo, M., Burgyan, J. and Martelli, G.P. (1994). Molecular biology of Tombusviridae. Adv. Virus Res. 44, 381-428. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). "Molecular Cloning: A Laboratory Manual," 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Sanfacon, H, Brodmann, P. and Hohn, T. (1991). A dissection of the cauliflower mosaic virus polyadenylation signal. Genes Dev. 5, 141-149. Sanger, F., Nicklen, S and Coulson, A.R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467. Sankar, S. and Porter, A.G. (1992). Point mutations which drastically affect the polymerization activity of encephalomyocarditis virus RNA-dependent RNA polymerase correspond to the active site of Escherichia coli DNA polymerase I. /. Biol. Chem. 267, 10168-10176. Sawicki, D.L., Sawicki, S.G., Keranen, S. and Kaariainen, L. (1981). A Sindbis virus coded function for minus-strand synthesis. J. Virol. 39, 348-358. Sawicki, D.L. and Sawicki, S.G. (1987). Alphavirus plus and minus strand RNA synthesis. In "Positive-strand RNA viruses" (M. Briton and R. Ruckert, Eds.), pp. 251-259. Alan R. Liss, Inc., New York. Sawicki, D.L. and Sawicki, S.G. (1993). A second nonstructural protein functions in the regulation of alphavirus negative-strand RNA synthesis. J. Virol. 67, 3605-3610. Scholthof, H.B., Morris, T.J. and Jackson, A.O. (1993). The capsid protein gene of tomato bushy stunt virus is dispensable for systemic movement and can be replaced for localized expression of foreign genes. Mol. Plant-Microbe Interact. 6, 309-322. Scholthof, H.B., Scholthof, K.-B.G. and Jackson A.O. (1995a). Identification of tomato bushy stunt virus host-specific symptom determinants by expression of individual genes from a potato virus X vector. Plant Cell 7, 1157-1172. Scholthof, K.-B.G., Scholthof, H.B. and Jackson A.O. (1995b). The tomato bushy stunt vims replicase proteins are coordinately expressed and membrane associated. Virology 208, 365-369. Schwartz, S., Felber, B.K. and Pavlakis, G.N. (1992). Mechanism of translation of monocistronic and multicistronic human immunodeficiency virus type 1 mRNAs. Mol. Cell. Biol. 12, 207-219. Sedman, S, and Mertz, J. (1988). Mechanisms of synthesis of virion proteins from the functionally bigenie late mRNAs of simian vims 40. J. Virol. 62, 954-961. Sedman, S.A., Gelembiuk, G.W. and Mertz, J.E. (1990). Translation initiation at a downstream AUG occurs with increased efficiency when the upstream AUG is located very close to the 5' cap. J. Virol. 64, 453-457. Shih, D.S. and Kaesberg, P. (1973). Translation of brome mosaic viral ribonucleic acid in a cell-free system derived from wheat embryo. Proc. Natl. Acad. Sci. USA 70, 1799-1803. Sit, T.L, Johnston, J.C, terBorg, M.G., Frison, E., McLean, M.A. and Rochon, D. (1995). Mutational analysis of the cucumber necrosis virus coat protein gene. Virology 206, 38-48. Skuzeski, J.M., Nichols L.M. and Gesteland, R.F. (1990). Analysis of leaky viral translation termination codons in vivo by transient expression of improved p-glucuronidase vectors. Plant Mol. Biol. 15, 65-79. Skuzeski, J.M., Nichols L.M., Gesteland, R.F. and Atkins, J.F. (1991). The signal for a leaky UAG stop codon in several plant vimses includes the two downstream codons. J. Mol. Biol. 218, 365-373. Slusher, L.B., Gillman, E.C, Martin, N.C. and Hopper, A.K. (1991). mRNA leader length and initiation codon context determine alternative AUG selection for the yeast gene MOD5. Proc. Natl. Acad. Sci. USA 88, 9789-9793. Smirnyagina, E., Hsu, Y-H., Chua, N., and Ahlquist, P. (1994). Second-site mutations in the brome mosaic virus RNA3 intercistronic region partially suppress a defect in coat protein mRNA transcription. Virology 198,427-436. Spaan, W., Delius, H, Skinner, M., Armstrong, J., Rotteir, P., Smeekens, S., van der Zeijst, B.A.M., and Siddell, S.G. (1983). Coronavirus mRNA synthesis involves fusion of non• contiguous sequences. EMBO J. 2, 1839-1844. Stanley, W.M. (1935). Isolation of a crystalline protein possessing the properties of tobacco- mosaic virus. Science 81, 644-645. Stauffacher, C.V., Usha, R., Harrington, M., Schmidt, T., Hosur, M.V., Johnson, J.E. (1987). In Crystallography in Molecular Biology (D. Moras, J. Drenth, B. Strandberg, D. Suck, K. Wilson, Eds), pp. 293-308. Plenum, New York. Stobbs, L.W., G.W. Cross, and M.S. Manoch. (1982). Specificity and methods of transmission of cucumber necrosis virus by Olpidium radicale zoospores. Can. J. Plant Pathol. 4, 134-142. Strauss, J.H. and Strauss, E.G. (1994). The alphaviruses: gene expression, replication and evolution. Microbiol. Rev. 58, 491-562. Tacke, E., Prufer, D., Salamini, F. and Rohde, W. (1990). Characterization of a potato leafroll luteovirus subgenomic RNA: differential expression by internal translation initiation and UAG suppression. J. Gen. Virol. 71, 2265-2272. Takamatsu, N, Watanabe, Y., Meshi, T. and Okada, Y. (1990). Mutational analysis of the pseudoknot region in the 3' noncoding region of tobacco mosaic virus RNA. J. Virol. 64, 3686- 3693. Tan, T.M.C., Gloss, B., Bernard, H-U. and Ting, R.C.Y. (1994). Mechanism of translation of the bicistronic mRNA encoding human papillomavirus type 16 E6-E7 genes. /. Gen. Virol. 75, 2663-2670. Tavazza, M., Lucioli, A., Ancora, G. and Benvenuto, E. (1989). cDNA cloning of artichoke mottle crinkle virus RNA and localization and sequencing of the coat protein gene. Plant Mol. Biol. 13, 685-692. Tavazza, M., Lucioli, A., Calogero, A., Pay, A., and Tavazza, R. (1994). Nucleotide sequence, genomic organization and synthesis of infectious transcripts from a full-length clone of artichoke mottle crinkle virus. /. Gen. Virol. 75, 1515-1524. Taylor, J.L., Jones, J.D.G., Sandler, S, Mueller, G.M., Bedbrook, J. and Dunsmuir, P. (1987). Optimizing the expression of chimeric genes in plant cells. Mol. Gen. Genet. 210, 572-577. Ten Dam, E.B., Pleij, C.W.A., and Bosch, L. (1990). RNA pseudoknots: translational frameshifting and readthrough on viral RNAs. Virus Genes 4, 121-136. Thomas, A.A., ter Haar, E., Wellink, J. and Voorma, H.O. (1991). Cowpea mosaic virus middle component RNA contains a sequence that allows internal binding of ribosomes and that requires eukaryotic initiation factor 4F for optimal translation. J. Virol. 65, 2953-2959. Tsai, C.-H. and Dreher, T.W. (1991). Turnip yellow mosaic virus RNAs with anticodon loop substitutions that result in decreased valylation fail to replicate efficiently. J. Virol. 65, 3060- 3067. Tsai, C.-H. and Dreher, T.W. (1992). Second-site suppressor mutations assist in studying the function of the 3' noncoding region of turnip yellow mosaic virus RNA. J. Virol. 66, 5190- 5199. Tsunasawa, S., Stewart, J.W. and Sherman, F. (1985). Amino-terminal processing of mutant forms of yeast iso-1-cytochrome c. J. Biol. Chem. 260, 5382-5391. van der Kuyl, A.C, Langereis, K., Houwing, C.J., Jaspars, E.M.J., and Bol, J.F. (1990). cis- Acting elements involved in replication of alfalfa mosaic vims RNAs in vitro. Virology 176, 346-354. van der Kuyl, A.C, Neeleman, L. and Bol, J.F. (1991). Deletion analysis of cis- and trans• acting elements involved in replication of alfalfa mosaic vims RNA 3 in vivo. Virology 183, 687-694. van der Vossen, E.A.G., Neeleman, L. and Bol, J.F. (1993). Role of the 5' leader sequence of alfalfa mosaic virus RNA3 in replication and translation of the viral RNA. Nucleic Acids Res. 21, 1361-1367. van der Vossen, E.A.G., Notenboom, T. and Bol, J.F. (1995). Characterization of sequences controlling the synthesis of alfalfa mosaic vims subgenomic RNA in vivo. Virology 212, 663- 672. Vankan, P., Edoh, D. and Filipowicz, W. (1988). Structure and expression of the U5 snRNA gene of Arabidopsis thaliana. Conserved upstream sequence elements in plant U-RNA genes. Nucleic Acids Res. 16, 10425-10440. van Lent, J., Storms, M., van der Meer, R., Wellink, J. and Goldbach R. (1991). Tubular structures involved in movement of cowpea mosaic vims are also formed in infected cowpea protoplasts. J. Gen. Virol. 72, 2615-2623. Vrati, S., Mann, D.A., and Reed, K.C. (1987). Alkaline northern blots: transfer of RNA from agarose gels to Zeta-Probe membrane in dilute NaOH. Molecular Biology Reports 1, 1-4. Wang, C, Sarnow, P. and Siddiqui, A. (1994). A conserved helical element is essential for internal initiation of translation of hepatitis C virus RNA. /. Virol. 68, 7301-7307. Weiland, J.J. and Dreher, T.W. (1989). Infectious TYMV RNA from cloned cDNA: effects in vitro and in vivo of point substitutions in the initiation codons of two extensively overlapping ORFs. Nucleic Acids Res. 17, 4675-4687. Weiner, A.M. and Maizels, N. (1987). tRNA-like structure tag the 3' termini of genomic RNA molecules for replication: implications for the origin of protein synthesis. Proc. Natl. Acad. Sci. USA 84, 7383-7387. White, K.A. and Morris, T.J. (1994). Enhanced competitiveness of tomato bushy stunt virus defective interfering RNAs by segment duplication or nucleotide insertion. J. Virol. 68, 6092- 6096. Widholm, J.M. (1972). The use of fluorescein diacetate and phenosafranine for determining viability of cultured plant cells. Stain Technol. 47, 189-194. Wieczorek, A. and Sanfacon, H. (1993). Characterization and subcellular localization of tomato ringspot nepovims putative movement protein. Virology 194, 737-742. Wieczorek, A. and Sanfacon, H. (1995). An improved method for the generation and transfection of protoplasts from Cucumis sativus cotyledons. Plant Cell Rep. 14, 603-610. Wilson, M.A. (1985). Nucleocapsid disassembly and early gene expression by positive-strand RNA viruses. J. Gen. Virol. 66, 1201-1207. Wolf, S., Deom, CM., Beachy, R.N. and Lucas, W. (1989). Movement protein of tobacco mosaic modifies plasmodesmatal size exclusion limit. Science 246, 377-379. Wu, G. J. and Bruening, G. (1971). Two proteins from cowpea mosaic virus. Virology 46, 596- 612. Wu, S., Rinhart, C.A., and Kaesberg, P. (1987). Sequence and organization of southern bean mosaic virus genomic RNA. Virology 161, 73-80. Xiong, Z., Kim, K.H., Kendall, T.L. and Lommel, S.A. (1993). Synthesis of the putative red clover necrotic mosaic virus RNA polymerase by ribosomal frameshifting in vitro. Virology 193,213-221. Yot, P., Pinck, M., Haenni, A., Duranton, H.M. and Chapeville, F. (1970). Valine specific tRNA-like structure in turnip yellow mosaic virus. Proc. Natl. Acad. Sci. USA 67, 1345-1352. Zerfass, K. and Beier, H. (1992). Pseudouridine in the anticodon G\|/A of plant cytoplasmic tRNATyr is required for UAG and UAA suppression in the TMV-specific context. Nucleic Acids Res. 20,5911-5918. Zimmern, D. (1988). Evolution of RNA viruses. In "RNA Genetics" (J.J. Holland, E. Domingo and P. Ahlquist, Eds.), pp. 211-240. CRC Press, Boca Raton, FL. Zuker, M. (1989). Computer prediction of RNA structure. In "Methods in Enzymology" (J.E. Dahlberg and J.N. Abelson, Eds.), Vol. 180, pp. 262-288. Academic Press, San Diego, California. 135 Chapters Appendix 5.1 The p-glucuronidase (GUS) enzyme system The gene encoding p-glucuronidase, gusA , was originally isolated from E. coli (Jefferson et al, 1986) but its analog has been found in virtually all mammalian tissues (for review, see Paigen, 1989). Because of the absence of appreciable background activity in higher plants, the GUS enzyme system has been widely adopted for use in plant molecular biology (Jefferson et ai, 1986). There are a number of conveniences associated with the use of this system including the variety of substrates suitable for histochemical , spectrophotometric and fluorometric analyses. The substrate chosen for use in the present study is the spectrophotometric substrate, p-nitrophenyl glucuronide (pNPG) because the high activity obtained in the present work did not require the use of a more sensitive fluorometric assay and because a spectrophotometric assay was easily adapted to a microtitre plate based approach using an available microtitre plate reader. 5.1.1 /?-nitrophenyl p-D-glucuronide (pNPG) substrate All of the p-glucuronide substrates available for detection of GUS activity contain the sugar D-glucopyranosiduronic acid attached by glycosidic linkage to a hydroxyl group of a chromogenic, fluorogenic, or other detectable molecule (Naleway, 1992). In the case of pNPG, the glucuronide is attached to a phenolic hydroxyl and detection of activity is a result of a shift in the absorption maximum of the phenol upon cleavage of the glycosidic bond (see below). The p-nitrophenol released is measured spectrophotometrically at 402-410 nm, and absorbance intensity at these wavelengths relates directly to the specific activity (Naleway, 1992). In the present study, since it was necessary only to determine GUS activity from a given construct 136 relative to WT, the data are represented not in terms of specific, but rather relative, GUS activity. COOH HO COOH HO [GUS] HO •N03 + H20 NO3 p-nitrophenyl /J-D-glucuronide D-gl ucur onic a ci d p - ni tr ophenol 5.1.2 Quantitative analysis of GUS activity The spectrophotometric assay for detecting GUS activity is quantitative and linear over extended periods of time. In addition, the GUS enzyme is quite stable and is capable of tolerating large amino-terminal additions making it suitable for analyzing translational fusions (Naleway, 1992) such as those in the present study. In this study, nucleotide substitutions were made in a short (18 nt) 5' extension corresponding to a multicloning site in the original pAGUS- 1 plasmid (Skuzeski et al, 1990; see below). Since the alterations did not occur within the GUS coding region, it is unlikely that any of the changes made would affect the kinetic parameters of the enzyme. For this reason, it was assumed appropriate to use a standard spectrophotometric assay with a 1 mM substrate concentration and adapt this for use in a microtitre plate-based assay . 5.1.3 Determination of relative GUS activity from transfected protoplasts The GUS activity from protoplasts transfected with pCGUS and pBGUS constructs in the present study was obtained using a kinetic spectrophotometric assay as described in Materials and Methods. The following data represents original measurements taken to determine GUS activity in the present study. Table 5.1, Fig. 5.1 and Table 5.2 represent analysis of a single experiment to determine GUS activity directed by pCGUS constructs culminating with the graph depicted in Fig. 3.19. Table 5.3 represents final data used to construct graphs depicting the results of three independent experiments involving pCGUS constructs in Fig. 5.2. Similarly, Tables 5.4 and 5.5 represent the original data used to determine GUS activity directed by pBGUS constructs in Fig. 3.21 and Table 5.6 and Fig. 5.3 includes final data from two independent pBGUS experiments. 5.1.4 The pAGUS-1 expression vector The pAGUS-1 vector, kindly provided by J. Skuzeski and R.F. Gesteland (University of Utah School of Medicine, Salt Lake City), consists of the coding region for GUS flanked by a reiterated CaMV 35S promoter and the nopaline synthetase (NOS) termination signal in pUC19. The diagram below, modified from Skuzeski et al. (1990), indicates the nucleotide sequence of the region altered in pAGUS-1 from that of the commercially available pBI221(Clontech). Restriction enzyme recognition sites were introduced into a region which includes the CaMV 35S promoter transcription start site (included in the BamHl site), the ATG initiation codon for GUS (included in the Ncol site) and a short amino terminal extension of the GUS coding region (includes the Hindlll and Apal sites). The boxed arginine codon corresponds to codon three of the WT GUS coding region (Skuzeski et al., 1990). CaMV 35S GAGGATCC GTCGACCATGGTAAGCTT AGCGGGCCC CGTC CGT BamHl Sail Ncol Hindlll Apal Table 5.1 Spectrophotometric measurement of p-nitrophenol absorbance in protoplast samples transfected with pCGUS contructsa Time0 pCGUS construct0 Replicate mock wt 1 8 0 0.000 0.002 0.000 0.002 0.002 0.000 0.000 0.001 0.000 0.000 15 0.001 0.060 0.059 0.049 0.082 0.022 0.053 0.073 0.028 0.067 30 0.002 0.107 0.110 0.091 0.145 0.035 0.098 0.133 0.052 0.122 60 0.004 0.219 0.216 0.187 0.297 0.078 0.196 0.272 0.100 0.242 90 0.004 0.346 0.334 0.298 0.449 0.124 0.309 0.430 0.168 0.357 150 0.005 0.615 0.590 0.509 0.796 0.229 0.518 0.751 0.297 0.640 Replicate H 0 0.000 0.003 0.000 0.003 0.003 0.000 0.000 0.001 0.000 0.000 15 0.001 0.049 0.052 0.059 0.063 0.015 0.056 0.052 0.023 0.043 30 0.003 0.085 0.097 0.101 0.117 0.028 0.104 0.096 0.042 0.078 60 0.002 0.172 0.197 0.204 0.232 0.057 0.208 0.196 0.082 0.156 90 0.003 0.258 0.317 0.310 0.344 0.095 0.319 0.321 0.137 0.250 150 0.004 0.430 0.566 0.495 0.628 0.163 0.581 0.493 0.242 0.438 Replicate in 0 0.000 0.002 0.000 0.001 0.000 0.001 0.000 0.001 0.000 0.000 15 0.001 0.043 0.045 0.067 0.077 0.018 0.057 0.067 0.033 0.052 30 0.001 0.076 0.083 0.118 0.136 0.033 0.105 0.118 0.064 0.097 60 0.003 0.164 0.169 0.247 0.278 0.072 0.207 0.253 0.120 0.188 90 0.003 0.233 0.269 0.372 0.437 0.108 0.343 0.385 0.181 0.307 150 0.004 0.367 0.475 0.641 0.742 0.205 0.538 0.606 0.318 0.524 a pCGUS constructs were separately transfected into N. plumbaginofolia protoplasts and incubated for 24 hr afterwhich time the protoplasts were collected, lysed in GUS extraction buffer and the protein concentration measured using a Bradford assay (see Materials and Methods). b incubation time (in minutes) at 37 °C after the addition of ImM />nitrophenol glucuronide spectrophotometric substrate to 5 pg soluble protein (protoplast extract) afterwhich the reaction was arrested with the addition of 0.5 M 2-amino-2-methylpropanediol and held at 4 °C. c values for each pCGUS contruct representing /?-nitrophenol absorbance at 415nm from separately transfected protoplast samples. 139 0 1 2 30 1 2 30 1 2 3 0 1 2 30 1 2 30 1 2 3 0 1 2 30 1 2 30 1 2 3 time (hours) time (hours) time (hours) Fig. 5.1 Time course of GUS activity as determined by p-nitrophenol absorbance. The absorbance of p-nitrophenol was measured at 415 nm for each pCGUS replicate which was separately transfected into the same batch of protoplasts (see Table 5.1). The slope of the relationship between absorbance and time, determined by simple regression analysis, represents the GUS activity for each pCGUS replicate (see Table 5.2). Each graph includes data for three replicates (shown by color) of the pCGUS construct indicated (by key) and also contains a reference line representing the average slope value for pCGUS-wt (shown in black). Table 5.2 GUS activity computed from kinetic spectrophotometric measurement of p-nitrophenol absorbance in Table 5.1a pCGUS Construct Replicate mock wt 1 2 3 4 5 6 7 8 I 0.244 0.234 0.204 0.315 0.091 0.207 0.299 0.118 0.253 II 0.170 0.226 0.197 0.247 0.065 0.230 0.200 0.096 0.174 III 0.147 0.189 0.255 0.296 0.081 0.217 0.244 0.126 0.209 Average*5 0.187 0.216 0.219 0.286 0.079 0.218 0.248 0.113 0.212 S.D. 0.051 0.024 0.032 0.035 0.013 0.011 0.050 0.015 0.039 Average0 0.00 1.00 1.16 1.17 1.53 0.42 1.17 1.32 0.60 1.13 S.D. 0.00 0.27 0.13 0.17 0.19 0.07 0.06 0.27 0.08 0.21 a data from Table 5.1 was used to plot the relationship of p-nitrophenol glucuronide absorbance vs. time (Fig. 5.1); the slope of the relationship was obtained by simple regression analysis and represents the GUS activity for each of three replicates pCGUS constructs wt through 8. b indicates the mean slope values for three replicates (from Fig. 5.1) and represents the average GUS activity for each pCGUS construct; S.D. is the standard deviation of the three slope values. c indicates transformation of the original average values where wt is arbitrarily assigned the value of 1.00 and all other values are given relative values; S.D. is the transformed standard deviation. Table 5.3 GUS activity computed from three independent experiments Experiment pCGUS Construct Average slope2 mock wt 1 2 3 4 5 6 7 8 I (J300) 0.000 0.187 0.216 0.219 0.286 0.079 0.218 0.248 0.113 0.212 n (A283) 0.001 0.451 0.437 0.449 0.533 0.151 0.482 0.545 0.232 0.391 HI (A282) 0.000 0.334 0.322 0.336 0.447 0.110 0.301 0.441 0.158 0.372 Stand. dev.b I (J300) 0.000 0.051 0.024 0.032 0.035 0.013 0.011 0.050 0.015 0.039 H (A283) 0.000 0.023 0.002 0.034 0.039 0.001 0.040 0.006 0.010 0.083 HI (A282) 0.000 0.034 0.028 0.009 0.051 0.010 0.122 0.035 0.032 0.033 Trans, slope0 I (J300) 0.00 1.00 1.16 1.17 1.53 0.42 1.17 1.33 0.60 1.13 n (A283) 0.00 1.00 0.97 1.00 1.18 0.34 1.07 1.21 0.51 0.87 m (A282) 0.00 1.00 0.96 1.00 1.34 0.33 0.90 1.32 0.47 1.11 Trans stand dev I (J300) 0.00 0.27 0.13 0.17 0.19 0.07 0.06 0.27 0.08 0.21 II (A283) 0.00 0.05 0.00 0.07 0.09 0.00 0.09 0.01 0.02 0.18 HI (A282) 0.00 0.10 0.08 0.03 0.15 0.03 0.36 0.10 0.10 0.10 Combined 0.00 1.00 1.03 1.06 1.35 0.36 1.05 1.29 0.53 1.04 a indicates the slope computed from three replicates in each experiment ; the slope represents GUS activity b indicates standard deviation c transformation of the original values such that wt is equal to 1.00 and all other values are made relative 141 B *+-» < o oo < 00 0 O > > 1 Pi Pi mock wt 1 D pCGUS constructs (I) < 13 01 mock wt 1 2 3 4 5 6 7 mock wt 1 2 3 4 5 6 7 pCGUS constructs (III) pCGUS constructs (combined) Fig. 5.2 Relative GUS activity directed by pCGUS construct series in three independent experiments. N. plumbaginofolia protoplasts were transfected with 20 ug of each pCGUS construct containing nucleotide substitutions surrounding the CNV p21 initiation codon which starts the synthesis of GUS. GUS activities for each construct were measured using a kinetic spectrophotometric assay. The GUS activity directed by pCGUS-wt in each experiment was arbitrarily assigned the value of 1 and the activities for the remaining constructs made relative to 1. Each of graphs A, B, and C represent the values obtained from independent experiments in which two or three replicates of each pCGUS contruct was transfected into the same batch of protoplasts. Graph D represents the combined values for the three separate experiments. The AUG contexts for each construct (indicated on the y axis) are as follows: wt - UUCAUGGA, 1 - UUCAUGGA. 2 - AUCAUGGA. 3 - AUCAUGUC, 4 - UUCAUGUA. 5 - UUCAUGGC, 6 - AUCAUGGC. 7 - UUCAUGUC, 8 - AUCAUGUA (p21 AUG codon underlined). Table 5.4 Spectrophotometric measurement of p-nitrophenol absorbance in protoplast samples transfected with pBGUS contructsa Timeb pBGUS construct0 Replicate I mock 1-1 1-2 4-1 4-2 5-1 5-2 7-1 7-2 0 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1 0.001 0.279 0.236 0.335 0.355 0.219 0.271 0.351 0.424 2 0.003 0.642 0.571 0.746 0.789 0.516 0.607 0.750 0.905 3 0.002 0.939 0.944 1.108 1.220 0.761 0.865 1.111 1.264 4 0.004 1.279 1.134 1.537 1.624 1.119 1.136 Replicate II 0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1 0.001 0.257 0.297 0.419 0.436 0.199 0.265 0.353 0.409 2 0.007 0.602 0.648 0.890 0.911 0.503 0.600 0.786 0.891 3 0.002 0.878 0.991 1.261 1.272 0.778 0.895 1.202 1.244 4 0.004 1.256 1.378 1.699 1.659 1.112 1.259 a pBGUS constructs were separately transfected into Af. plumbaginofolia protoplasts and incubated for 24 hr afterwhich time the protoplasts were collected, lysed in GUS extraction buffer and the protein concentration measured using a Bradford assay (see Materials and Methods). b incubation time (in hours) at 37 °C after the addition of ImM p-nitrophenol glucuronide spectrophotometric substrate to 5 ug soluble protein (protoplast extract) afterwhich the reaction was arrested with the addition of 0.5 M 2-amino-2-methylpropanediol and held at 4 °C. c values for each pBGUS contruct representing p-nitrophenol absorbance at 415nm from protoplast samples in a microtitre plate using an ELISA Titertek reader. Table 5.5 GUS activity computed from kinetic spectrophotometric measurement of p-nitrophenol absorbance in Table 5.4a pBGUS Construct Replicate mock 1-1 1-2 4-1 4-2 5-1 5-2 7-1 7-2 I 0.001 0.322 0.298 0.385 0.411 0.278 0.286 0.373 0.427 II 0.001 0.313 0.345 0.424 0.415 0:280 0.314 0.404 0.421 Averageb 0.001 0.320 0.409 0.290 0.406 S.D. 0.000 0.020 0.017* 0.017 0.024 Average0 0.00 1.00 1.28 0.91 1.27 S.D. 0.00 0.06 0.05 0.05 0.07 a data from Table 5.4 was used to plot the relationship of p-nitrophenol glucuronide absorbance vs. time; the slope of the relationship was obtained by simple regression analysis and represents the GUS activity for each of three replicates pBGUS constructs 1,4,5 and 7. b indicates the mean slope values for two replicates of two samples of each pBGUS construct and represents the average GUS activity for each; S.D. is the standard deviation of the three slope values c indicates transformation of the original average values where wt is arbitrarily assigned the value of 1.00 and all other values are given relative values; S.D. is the transformed standard deviation. Table 5.6 GUS activity computed from two independent experiments Experiment pBGUS Construct Average slopea mock 1 4 5 7 I 0.001 0.320 0.409 0.290 0.406 II 0.002 0.368 0.471 0.372 0.464 Stand. dev.b I 0.000 0.020 0.017 0.017 0.024 II 0.000 0.043 0.037 0.035 0.041 Trans, slope0 I 0.00 1.00 1.28 0.91 1.27 II 0.00 1.00 1.29 1.01 1.26 Trans stand dev I 0.00 0.06 0.05 0.05 0.07 II 0.00 0.12 0.10 0.10 0.11 Combined 0.00 1.00 1.29 0.96 1.27 a indicates the slope computed from two replicates in each experiment; the slope represents GUS activity b indicates standard deviation c transformation of the original values such that wt is equal to 1.00 and all other values are made relative 144 Fig. 5.3 Relative GUS activity directed by pBGUS constructs in two independent experiments. N. plumbaginofolia protoplasts were transfected with 20 ug of each pBGUS construct. GUS activities for each construct were measured using a kinetic spectrophotometric assay. The GUS activity directed by pBGUS-1 in each experiment was arbitrarily assigned the value of 1 and the activities for the remaining constructs made relative to 1. A. Relative GUS activity for each pBGUS construct from experiment II B. Relative GUS activity for each pBGUS construct from combining experiments I (shown in Fig. 3.21 in Results) and II (shown in A). The CNV p21 AUG codon contexts for each construct (indicated on the y axis) are as follows: 1 - UUCAUGGA. 4 - UUCAUGUA. 5 - UUCAUGGC and 7 - UUCAUGUC (p21 AUG underlined).