Insertional mutagenesis of the cauliflower mosaic virus genome

A series of small insertions has been introduced into the various translational reading frames of the DNA of a "severe" strain of cauliflower mosaic virus (CaMV). A selectable gene (the kanamycin phosphotransferase gene of Tn903), flanked by a series of symmetrically arranged cloning sites taken from M13mp7, was used to prepare the site-specific mutants. In-phase insertions of 12 or 30 bp, which introduced unique SalI sites into reading regions I, III, IV, V and into the amino-proximal portion of region VI, destroyed infectivity. Insertions in the amino-distal portion of region VI, in the large intergenic region, and in region II retained infectivity. The amino-distal insertions in region VI reduced the severity of symptoms in plants. The insertion in region II destroyed aphid transmissibility. Longer DNA segments when inserted into region II or into the amino-distal portion of region VI destroyed infectivity, but similar insertions in the intergenic region were without effect on virus infection or development.     

Mutagenesis of cauliflower mosaic virus

A series of insertion mutants of cauliflower mosaic virus (CaMV) DNA has been constructed in vitro. These insertions consist of a short DNA sequence (10 or 22 bp) containing a restriction endonuclease site (SmaI) not represented on the viral DNA. Viral infectivity was analyzed by inoculating plants with the mutated cloned viral DNA and observing symptoms. Insertions within ORFVII, and in one site within the large intergenic region, did not interfere with viral infectivity, whilst insertions within ORFII and at the end of ORFIV retarded the development of viral symptoms. All other insertion mutants analyzed were lethal. CaMV with a deletion of 105 bp within ORFVII was viable. Such viable mutants can be used to construct additional deletions or to insert foreign DNA into the viral genome.     

Studies on the single-stranded discontinuities of the cauliflower mosaic virus genome.

The Cauliflower Mosaic Virus (CaMV) genome is a double-stranded DNA molecule of about 5 million daltons. Native DNA molecules appear heterogeneous when analysed by gel electrophoresis. We have examined the nature of this apparent heterogeneity. Besides, this genome is shown here to contain three single-stranded breaks, as revealed by different denaturation experiments: heating at 75 degrees C, treatment with NaOH or dimethyl sulfoxide (DMSO). Labelling with terminal transferase proves that the 3' ends at these interruptions all have free hydroxyl groups. Electron microscopy and alkaline gel electrophoresis indicate that these three discontinuities are shared by both strands, and that they are not randomly located. S1 nuclease is active on CaMV DNA and generates three fragments. The comparison between the sizes of these fragments and of the products of denaturation leads us to consider that S1 acts at the level of the interruptions. We have determined that two of them, distant by one third genome unit, are in the same strand; the other is in the opposite strand, distant by one sixth genome unit from the nearest other one. The combined use of restriction enzymes and S1 nuclease has enabled us to locate these three discontinuities on the restriction map of the CaMV genome that we have otherwise established.     

Nucleotide sequence of cauliflower mosaic virus DNA

The complete nucleotide sequence (8024 nucleotides) of the circular double-stranded DNA of cauli-flower mosaic virus has been established. The DNA molecule is known to possess three discrete single-stranded discontinuities, often referred to as “gaps”, two in one strand and one in the other. The sequence data indicate that gap 1, the single discontinuity in the α strand, corresponds to the absence of no more than one or two nucleotides with respect to the complementary β strand. The two discontinuities in the β strand, however, are not authentic gaps since no nucleotides are missing, but are instead regions of sequence overlap: a short sequence (19 residues for gap 2, at least 2 residues for gap 3) at one terminus of each discontinuity, probably the 5′ terminus, is displaced from the double helix by an identical sequence at the other boundary of the discontinuity. Analysis of the distribution of nonsense codons in the DNA sequence is consistent with other evidence that only the α strand is transcribed. The coding region extends around the circular molecule from 4 map units of gap 1, the map origin, to map position 91, and consists of six long open reading frames. Our findings suggest, but do not prove, that the DNA sequence of the open reading frames is colinear with viral protein sequences. The cistron for the viral coat protein, which is probably synthesized in the form of a precursor, has been situated in coding region IV on the basis of its unusual amino acid composition.     

Identification of structural domains within the cauliflower mosaic virus movement protein by scanning deletion mutagenesis and epitope tagging.

Plant viruses encode proteins that mediate their movement from cell to cell through plasmodesmata. Currently, the mechanisms of action of these movement proteins (MPs) can be divided broadly into two types, requiring or not requiring the presence of viral capsid protein. Cauliflower mosaic virus encodes a multifunctional MP (P1) that modifies plasmodesmata through the formation of tubules that contain virus particles. To investigate the structure of P1, 26 small deletions (scanning deletions) were used to characterize regions of P1 essential for full biological activity. These deletions identified an N-terminal region and a region close to but not at the C terminus as domains that could tolerate manipulation, although gross deletions of either domain abolished infection. In sequence comparisons with other caulimovirus MPs, these regions coincided with the areas of least amino acid homology. Epitope tags inserted into either of these regions were stably maintained in systemic infections, and in extracts from infected plants, tagged P1 was detected on immunoblots. We predicted that, from the hypervariability of these regions, they would be located on the surface of the native P1 structure. Immunofluorescence of P1-specific tubules formed on the surface of infected protoplasts confirmed that the N-terminal and C terminus-proximal regions were exposed on the surface of the P1 tubule subunit. These findings establish a structure for P1 that is likely to be applicable to other tubule-forming MPs.     

Nuclei purified from cauliflower mosaic virus-infected turnip leaves contain subgenomic, covalently closed circular cauliflower mosaic virus DNAs

Nuclei isolated from cauliflower mosaic virus (CaMV) infected turnip leaves contain subgenomic CaMV DNA species in addition to the genome length CaMV DNA. These subgenomic CaMV DNA species are present as covalently closed circles (form I), relaxed circles (form II) and linear (form III) molecules. The subgenomic form I DNA species range in size from about 10% of genome length to nearly genome length. These subgenomic DNA species appear in tissue infected with cloned CaMV DNA, indicating that they arise rapidly and have not accumulated in the virus population from serial propagation of CaMV. No specific region of the CaMV genome appears to be preferentially deleted to form the subgenomic CaMV DNA species. At least three distinct subgenomic species appear to accumulate preferentially in nuclei isolated from infected tissue. Two of these abundant subgenomic CaMV DNA species are form I and the other one is form III. Some of the subgenomic CaMV DNA species appear to be minichromosomes.     

Selective allele loss and interference between cauliflower mosaic virus DNAs

Summary Some cauliflower mosaic virus (CaMV) alleles are selectively lost during growth of the virus in mixedly infected turnip plants. Viral DNA from plants co-inoculated with DNA of the cabbage S isolate and infectious cabbage S DNA with an extra EcoRI restriciion site lacked the extra site. The EcoRI allele was also lost in most plants co-inoculated with a non-infectious mutant of cabbage S DNA while little selective allele loss was observed with two other non-infectious mutant DNAs. Plants co-inoculated with DNAs of closely-related isolates (CM4-184 and W) contained both parental viral DNAs and some DNAs with characteristics of both parents. Interference, scored as a reduced frequency of infection or a delay in symptom appearance relative to plants inoculated with wild-type DNA, occurred when plants were inoculated with wild-type and mutant DNAs covalently attached to one another in partial dimer plasmid DNAs. Similarities in the conditions leading to selective allele loss and those leading to interference suggest that both may have been due to active gene conversion between CaMV DNA molecules.     

Pathogenic interactions between variants of cauliflower mosaic virus and Arabidopsis thaliana

Pathogenic interactions between genetic variants of cauliflower mosaic virus (CaMV) and Arabidopsis thaliana were characterized to identify combinations potentially useful in molecular genetic analysis. Infections of a glabrous mutant (gl1) of Arabidopsis ecotype Columbia (Col-0 gl1) by 30 CaMV isolates were assessed by recording symptom character. Thirteen isolates failed to cause symptoms; the remainder induced symptoms that varied between mild and very severe. Some CaMV isolates produced symptoms in Arabidopsis that differed significantly in severity or character from those produced in a standard host Brassica rapa (turnip). A greater variety of symptom types was observed in a single Arabidopsis ecotype infected with a range of CaMV isolates than was found in a range of Arabidopsis ecotypes infected with a single, typical CaMV isolate (Cabb B-JI). One isolate, Bari-1, that was asymptomatic but accumulated virus in Arabidopsis ecotype Col-0 gl1, caused mild symptoms in ecotype Ler gl1. A hybrid virus constructed from CaMV isolates Cabb B-JI and Bari-1 produced symptoms in Arabidopsis variants that were more severe than in either parental isolate. From a screen of EMS-mutagenized Arabidopsis, one mutant (Col-0 dv1) with a pale-green, dark-vein phenotype which had an altered symptom response to CaMV, was isolated. From this, a phenotypically near-normal revertant (Col-0 dv1R) spontaneously arose, but which showed altered responses to CaMV. Infection of Col-0 dv1R by CaMV isolate Bari-1 elicited symptoms unlike the parent Arabidopsis ecotype (Col-0 gl1). Also, Col-0 dv1 and Col-0 dv1R expressed an uncharacteristic necrotic reaction to CaMV.     

Expression of cauliflower mosaic virus gene I in Saccharomyces cerevisiae

Cauliflower mosaic virus (CaMV) gene I encodes a 40-kDa protein, P1, which is thought to be involved in the cell-to-cell movement of the virus. In order to investigate its functioning, P1 was expressed in Saccharomyces cerevisiae transformed by an expression vector containing CaMV gene I. When produced in yeast, P1 was 40 kDa in size and not N-glycosylated.     

Expression of the cauliflower mosaic virus capsid gene in vivo

Antisera against the N-terminal and C-terminal parts of the potential ORF IV product were used to analyse extracts from CaMV-infected turnip leaves by immunoblotting. Polypeptides of 87, 83, 82, 60 and 57 kDa were detected. The origin of these proteins is discussed.     

Infection of evacuolated turnip protoplasts with liposome-packaged cauliflower mosaic virus

The infectivity of reverse phase evaporation (REV) liposome-encapsidated cauliflower mosaic virus (CaMV) to turnip protoplasts was tested. The uptake of neutral or negative liposomes was stimulated by polyethylene glycol (PEG), while high levels of uptake of positive liposomes were obtained both in the absence and presence of PEG. The delivery of the vesicle contents to the protoplasts paralleled the uptake of liposomes. CaMV delivered to turnip protoplasts was degraded during the early period of culture. No increase in the amount of CaMV DNA could be detected on longer periods of culture. In contrast, when protoplasts were evacuolated prior to addition of REV liposomes, an increase in the amount of CaMV DNA was noted after some initial degradation of the input DNA.     

Molecular dissection of the cauliflower mosaic virus translation transactivator.

The cauliflower mosaic virus (CaMV) transactivator (TAV) is a complex protein that appears to be involved in many aspects of the virus life cycle. One of its roles is to control translation from the polycistronic CaMV 35S RNA. Here we report a molecular dissection of TAV in relation to its ability to enhance dicistronic translation in transient expression experiments. We have identified a protein domain that is responsible and sufficient for that activity. This 'MiniTAV domain' consists of only 140 of the 520 amino acids in the full-length sequence. A further domain located outside the MiniTAV, and therefore dispensable for transactivation, is probably involved in interactions with other molecules. This was identified by its ability to compete with wild-type TAV and some of its deletion mutants. We found, furthermore, that the TAV protein binds RNA. Two regions needed for RNA-binding properties were defined outside the MiniTAV domain and RNA binding seems not to be directly involved in the transactivation mechanism.     

Oligonucleotide directed mutagenesis of cauliflower mosaic virus DNA using a repair-resistant nucleoside analogue: identification of an agnogene initiation codon

Mutation of the initiation codon of the dispensible open reading frame, ORF VII, of cauliflower mosaic virus (CaMV) delayed the appearance of disease symptoms, but the mutants reverted with high frequency. This suggests a role of this start codon in viral expression. Oligonucleotide-directed mutagenesis, utilizing a novel, repair-resistant deoxyguanosine analogue, 2'-deoxy-7-deazainosine (dDI), highly improved the yield of mutants.