WO1993021329A1

WO1993021329A1 - Virus resistant plants

Info

Publication number: WO1993021329A1
Application number: PCT/GB1993/000829
Authority: WO
Inventors: David Baulcombe; Marian Longstaff
Original assignee: The Gatsby Charitable Foundation
Priority date: 1992-04-21
Filing date: 1993-04-21
Publication date: 1993-10-28
Also published as: AU3961793A; GB9208545D0

Abstract

The invention relates to the genetic modification of a plant in order to confer resistance to infection by a plant virus. Resistance is conferred by transforming the plant with a DNA construct encoding a protein which is a modified form of the viral component of the replicase enzyme of a plant virus in which sequence coding for the motif Gly-Asp-Asp in the wild type enzyme has been replaced by sequence coding for a motif which renders the protein non-functional as a replicase. Preferably the DNA encodes a protein in which the motif Gly-Asp-Asp has been replaced by the motif Ala-Asp-Asp, Gly-Ala-Asp or Gly-Glu-Asp.

Description

VIRUS RESISTANT PLANTS

This invention relates to the genetic modification of a plant in order to confer resistance to infection by a plant virus.

The concept of pathogen derived resistance was propounded in 1985 by Sanford & Johnson (J. Theor. Biol. , 113, 395-405 (1985)) and predicts that disease susceptible organisms may be transformed to resistance by introduction of a transgene derived from a pathogen. This idea as set out in the paper was based on theoretical considerations making use of data obtained earlier with E^. coli transformed to express genes from a pathogen (Qβ phage) , either in the sense or anti-sense orientation. The transformed E_^_ coli . were consequently resistant to infection by Qβ and a related phage. Parallel results were subsequently obtained with transgenic plants expressing the coat protein gene of various RNA viruses (reviewed in Beachy, Annu. Rev. Phytopathol. , 28., 451-474 (1990)). These plants were resistant to infection by the virus from which the αoat protein gene was obtained, although the resistance associated with anti-sense expression was weaker than with the sense orientation (Hemmenway et al. , EMBO J. , 2, 1273-1280 (1988); Powell et al., Proc. Natl. Acad. Sci. USA, £ 6949-52 (1989)).

Pathogen-derived resistance to virus infection was obtained also with plants transformed to express a defective interfering derivative of a gemini-viral genome (Stanley et al. , Proc. Natl. Acad. Sci. USA jT7, 6291-5 (1990)) or a satellite RNA of cucumber mosaic (Harrison et al., Nature 328. 799-802 (1987)) and tobacco ringspot viruses (Gerlach et a_l. , Nature 328, 802-805 (1987)). In these instances it is thought that the resistance was due, at least in part, to competition between the transgenically expressed sequence and virus inoculated to the transgenic plants.

A further example of pathogen-derived resistance was reported by Golemboski et al. (Proc. Natl. Acad. Sci. USA, 81, 6311-6315 (1990)) who transformed tobacco plants with a construct which included sequence containing all but the three 3'-terminal nucleotides of the 54-kDa gene of tobacco mosaic virus (TMV) strain Ul. The resulting transgenic plants were resistant to challenge by TMV strain Ul and by a related strain TMV mutant YSI/1 but were not resistant to challenge by TMV strains U2 or L. or cucumber mosaic virus. The TMV 54-kDa gene encodes a putative component of the TMV replicase complex and the object of the work was to attribute a function to this protein. The authors did not establish the mechanism by which resistance was achieved although they shewed that there was an effect of transgenic 54kDa protein on accumulation of TMV RNA in infected protoplasts (Carr and Zaitlin, Molecular Plant Microbe Interactions 4., 579-585 (1991)) and hypothesised that the protein probably interfered with TMV replication.

However, transgenic expression of plant viral genes does not always produce resistance: tobacco plants expressing the 30000 Mr (3OK) movement protein of TMV (Deom et al., Science, 237. 389-394 (1987)) and the putative replicase proteins of alfalfa mosaic virus (Taschner et ajL. , Virol., 181. 445-450 (1991)) were fully susceptible to infection. In both instances, expression of the transgene complemented the mutant or absent function in the respective viruses. Transgenic expression of the TMV movement protein complemented mutations in and conferred enhanced susceptibility to tobacco rattle virus (Ziegler-Graff et al. , Virol., 182., 145-155 (1991)).

US-A-4774182 (Syzbalski) relates to the use of a so-called "dominant-lethal" gene for imparting immunity to a biological host against an infectious agent. The

"dominant-lethal" gene is a variant of a gene segment from a normally infectious agent which, in its naturally occurring fully operational form, normally expresses a proteinaceous material which is not a repressor protein and which participates in a protein complex of at least two molecules, the complex being used in the structure, replication, or expression of the infectious agent. The defective variant gene produces a defective protein which will bind with at least part of the remainder of the complex and will inhibit the ability of the infectious agent to develop, the said inhibition being caused by the binding of the defective proteinaceous material to the remainder of the complex.

The experimental .work reported in the patent relates to transformation of E__s. coli with mutated 0 and/or P genes of phage A thereby conferring immunity on the host organism against subsequent infection by phage λ. The mutation is brought about by conventional and essentially random means and suitable mutants are selected on the basis of the immunity that they confer. The precise nature of the mutations involved in any particular case appears not to have been investigated.

Irokuchi and Hirashi a. (J. Virol., £1, 3946 (1987)) noted that a segment Tyr-X-Asp-Asp is well conserved in many putative RNA-dependent RNA or DNA polymerases of plant and animal viruses. The authors obtained a cDNA construct which encoded the intact replicase β-subunit protein gene of bacteriophage Qβ which protein contains the conserved segment Tyr-Gly-Asp-Asp as amino acid residues 356 to 359 and they substituted Ala, Ser, Pro,

Met and Val respectively for the Gly at position 357. It was found that whereas the unmutated β-subunit gene was expressed and could function as a normal replicase subunit in Is. coli. the product of the mutated plasmids lost the replicase activity. However when Ej_ coli carrying the various mutant plasmids was infected with wild type Qβ phage, proliferation of the progeny phages was prevented.

The authors did not establish the basis for the phage resistance which they had induced but suggested that the defective replicase might bind to the parental RNA but did not catalyse RNA polymerisation. The authors also suggested that a similar system for conferring immunity might be generally applicable to plant and animal viruses which have RNA-dependent DNA polymerases. However, no experimental work is reported beyond the infection of E. coli with bacteriophage Qβ.

The motif Gly-Asp-Asp is conserved, generally in the viral subunit of the replicase enzyme, in RNA viruses of all known types, ranging from RNA bacteriophages with positive strand genomes, including positive strand viruses of plants and animals, to double stranded negative stranded and ambisense viruses. A similar motif is also present in retroviruses (see Poch et al.. EMBO J. , 8., 3867-3874 (1989), Habili and Symons, Nucl. Acids Res., 22., 9543-9555 (1989), Kamer and Argos, Nucl. Acids. Res., 12., 7269-7282 (1984)).

It has now been found that resistance to a plant virus can be conferred on a plant by transforming that plant with DNA which is capable of expressing a modified form of the viral component of the replicase enzyme of the virus in question. In particular the DNA encodes a modified or mutant form of the viral component of the replicase enzyme in which at least one sequence motif Gly-Asp-Asp has been replaced by Ala-Asp-Asp or another motif having an equivalent effect on the structure of the protein. In particular the effect of the mutation should be such as to render the protein non-functional as a replicase. Accordingly, the present invention provides a DNA molecule encoding a protein which is a modified form of the viral component of the replicase enzyme of a plant virus in which sequence coding for the motif Gly-Asp-Asp in the wild type enzyme has been replaced by sequence coding for a motif which renders the protein non-functional as a replicase.

Preferably the motif Gly-Asp-Asp is replaced by the motif Ala-Asp-Asp. Alternative mutations include replacement of the motif Gly-Asp-Asp by the motif Gly-Ala- Asp or by Gly-Glu-Asp.

For example, the DNA molecule may be a plasmid vector containing DNA encoding the modified viral component of the replicase enzyme in a form suitable for expression in a plant cell and which vector can be used for the transformation of a plant using standard techniques.

The present invention also provides a plant having incorporated into its genome DNA encoding a protein which is a modified form of the viral component of the replicase enzyme of a plant virus in which sequence coding for the motif Gly-Asp-Asp in the wild type viral component of the enzyme has been replaced by sequence coding for a motif which renders the protein non-functional as a replicaes, preferably the motif Ala-Asp-Asp, said sequence being in a form in which it is capable of expression in the plant or in a part thereof. The fact that the plant has incorporated into its genome DNA encoding the modified viral component of the viral replicase enzyme will generally render the plant resistant against infection by the virus in question. The invention also extends to plant propagation material, including seeds, of such a plant. The invention also provides a method of imparting to a plant resistance against a specific plant virus which method comprises the steps of:

(i) providing a DNA construct which encodes and is capable of expressing in the plant a protein which is a modified form of the viral component of the replicase enzyme of the plant virus in which at least one sequence motif Gly-Asp-Asp in the wild type viral component of the replicase enzyme has been replaced by a motif which renders the protein non-functional as a replicase, preferably the motif Ala-Asp-Asp, Gly-Ala-Asp or Gly-Glu-

Asp, most preferably the motif Ala-Asp-Asp; and

(ii) introducing the construct into the plant by transformation. The method will generally also involve propagating from the transformed plant to provide plants having a phenotype which includes the ability tc express the modified viral component of the viral replicase enzyme.

The present invention is generally applicable to any

RNA-plant virus in which viral replication involves a replicase enzyme which includes the sequence motif Gly-Asp-Asp. Such replicase enzymes often occur as multi sub-unit enzymes of which the virally encoded protein is only a part. In this case the invention is applicable to RNA-plant viruses in which the sequence motif Gly-Asp-Asp occurs in the virally encoded component of the replicase enzyme. Accordingly, as used herein, the term "viral component of the replicase enzyme" can mean the whole enzyme if this is encoded by the virus or the virally encoded component of the replicase enzyme in the case of a multi sub-unit enzyme. The RNA sequences of many RNA-plant viruses are already known and have been produced as cDNA. Alternatively cDNA corresponding to the whole genome of a plant virus or the viral component of the replicase enzyme thereof can be produced by the application of standard cDNA cloning techniques. Once DNA has been obtained encoding the wild type viral component of the replicase enzyme, then appropriate changes can be introduced into the DNA sequence using standard mutation techniques of recombinant DNA technology to produce DNA encoding the modified viral component of the replicase enzyme.

It may be particularly advantageous to render susceptible plants resistant to viruses such as following: cucumber mosaic virus; tomato bushy stunt virus; tomato spotted wilt virus; potato leafroll virus; barley yellow dwarf virus; potato virus Y.

DNA encoding the modified viral component of the replicase enzyme will generally be incorporated into an expression plasmid in which the DNA is operably linked to a suitable promoter which is capable of expressing the DNA in the host plant. Generally the plasmid will also include suitable regulatory and control sequences appropriate for expression of the DNA in the plant in question.

Plants transformed with DNA encoding the modified viral component of the viral replicase enzyme may be produced by standard techniques already known for the genetic manipulation of plants. For example the DNA encoding the modified viral component of the viral replicase enzyme together with a promoter and other regulatory and control sequences may be incorporated into an Aσrobacterium vector and plant material may then be infected by a strain of Agrobacterium carrying this vector. In this way the DNA encoding the modified viral component of the viral replicase enzyme becomes integrated into the genome of the plant tissue so that plants propagated from the tissue also carry this DNA.

It is preferred that the promoter and other regulatory and control sequences are such that the modified viral component of the viral replicase will be constitutively expressed throughout the plant at a sufficient level to render the plant resistant to the virus in question. Expression of the modified viral component of the viral replicase will generally be at such a low level that growth of the plant will not be affected. However, by appropriate choice of promoter it may be possible to confine expression of the modified viral component of the viral replicase to specific parts of the plant. Alternatively it may be possible to use a promoter which directs expression of the DNA encoding the modified viral component of the viral replicase only in specific circumstances, for example in the presence of the virus against which resistance is desired or the presence of specific components of that virus.

Expression of the modified viral component of the viral replicase by a plant will generally render that plant resistant against infection by the virus from which the viral replicase is derived and against infection by closely related viruses although in some cases the plant may also be resistant to infection by a wider range of viruses. It is also possible to transform plants with a construct which is capable of expressing^' modified viral components of viral replicase enzymes derived from more than one plant virus, the modified viral components of the viral replicases having in each case at least one motif Gly-Asp-Asp replaced by a motif which renders the protein non-functional as a replicase, preferably the motif Ala-Asp-Asp. Plants transformed with DNA capable of expressing two or more modified viral components of viral replicase enzymes may be produced by use of the same general techniques as are described above. As well as imparting resistance to infection by the specific viruses from which the modified viral components of viral replicase enzymes are derived, transformation of a plant with DNA encoding two or more modified viral components of viral replicase enzymes may also impart resistance against a broader spectrum of viruses.

It is possible to engineer multigene virus resistance by use of a combination of modified viral components of viral replicase enzyme and coat protein-based transgenes (Beachy et al. , Annu. Rev. Phytopathol. , 2J3., 451-474 (1990)). Such a combination has the potential to provide potent resistance since the replicase gene(ε) target the replication stage of the infection cycle, whereas the coat protein may affect several stages including viral disassembly and systemic movement.

In the context of the present invention, reference to the viral component of the replicase enzyme of a plant virus also includes the viral component of the replicase enzymes of a hybrid plant viruses. Such hybrid viruses may have a viral component of the replicase enzyme which includes domains from two or more different strains of a particular virus. Transgenic plants which have been engineered to express a modified viral component of such a hybrid virus, i.e. a modified protein which includes domains derived from two different strains of a particular virus, may have resistance against a broader range of viruses then plants which express a modified viral component of the replicase enzyme of a single strain of the virus.

The invention is illustrated further by the following experimental section, the first part of which describes the effect of the introduction of the mutation

Gly-Asp-Asp to Ala-Asp-Asp and other mutations affecting the Gly-Asp-Asp motif into a transgenically expressed gene for the 166K protein of potato virus X (PVX; Huisman et al., J. Gen. Virol., 69., 1789-1798 (1988), Skryabin et al., Nucl. Acids Res., .16., 10929-10930 (1988) and Or an et al. , Virus Res., 16., 293-306 (1990)). The 166K protein of potato virus X is the 5'-terminal gene of the positive stranded genome of the virus and encodes a single

Gly-Asp-Asp motif in the C-terminal part of the protein product. The virus infects plants in the Solanaceae including tobacco which was selected as the recipient of the 166K protein based transcripts. The final part of the experimental section relates to constructs of cucumber mosaic virus (CMV) designed to show the effects of mutations of the Gly-Asp-Asp motif encoded in CMV RNA2.

The description of the experimental section refers to the accompanying drawings in which:

Fig. 1 shows construction of 166K protein mutants in the 35S RNA expression cassette of plasmid pRok2; Fig. 2 shows the symptom development on plants expressing modified and wild type forms of the 166K protein;

Fig. 3 shows the type of symptoms shown by the FI progeny of one of the plants (3.1) of Fig. 2; Fig. 4 illustrates the general method of PCR directed mutagenesis;

Fig. 5 illustrates the use of the method of Fig. 4 to construct DNA expressing the mutant 166K protein in the expression cassette of pRok2. Fig. 6 shows a schematic representation of the pVX genome together with constructs involved in the production of hybrid PVX viruses.

Fig 7 shows CMV RNA2 constructs.

MATERIALS AND METHODS

Nomenclature Genes are italicised, so that the gene for the 166K protein is 166K. Sequence co-ordinates are based on the nearest homologous site in the published PVX sequence of Huisman et al., (J. Gen. Virol., .69., 1789-1798 (1988)). The RNA derived from cDNA clones is given the prefix t; transcripts and RNA of pTXS are tTXS, for example.

Viral Strains

Potato virus X (PVX_UK3) was described by Kavanagh et al. , (Virology, in press (1992)). PVX_cp was described by Moreira et al., (Ann. Appl. Biol. , ϋ, 93-103 (1980)) and has been renamed PVX ,, to identify it as a group 2 strain (Cockerham, Heredity, 2_5, 309-348 (1970)) distinct from a mutant derivative (PVX_cp4) Jones, PI. Pathol., 34. 182-189 (1985)) in strain group 4 (Cockerham, 1970). Other strains used were PVX_HB (Moreira et al. , 1980) , PVXg^ (Adams et al., PI. Pathol., H, 435-437 (1984)), Jones 1985) and PVX_DX (Jones, Pi. Pathol., 3JL, 325-331 (1982) and PVX_B (Cockerham, 1970) .

Transcription reaction in vitro

T7 RNA polymerase (Stratagene) was used to synthesize capped RNA transcripts from 1 μg Spel-linearized plasmid DNA in a final volume of 10 μl containing 40 mM tris(hydroxymethyl)-aminomethane-HCl (Tris-HCl) pH 8.0: 25 mM NaCI; 8 mM MgCl₂; 2mM spermidine; 10 mM dithiothreitol; 2 mM of each ATP, UTP, and CTP; 0.2 mM GTP; 0.5 mM cap analogue m⁷G(5')ppp(5')G; 0.8 units μl^"1 RNase inhibitor (Pharmacia); 0.1 μg μl'¹ linearized DNA and 5 units μl^'1 T7 RNA polymerase (GIBCO/BRL) . Reactions were incubated at 37°C for 20 min before addition of 20 mM GTP to a final concentration of 2 mM. After a further 40 min incubation at 37°C the reactions were terminated by phenol/chloroform extraction. The reaction products were further purified by two ethanol precipitations and resuspended in 2.5 μl water.

Inoculation procedure

Each plant was inoculated with 20 μl of sap extracted from N. clevelandli or N. tabac m infected with the indicated strain of virus. Alternatively the inocula were RNA samples from infected plants or in vitro transcription products. The RNA inocula of 20 μl plant"¹ were made up with 2.5 μl RNA to which were added 17.5 μl of 42mM sodium phosphate (pH 7.0) containing bentonite (5mg ml"¹) . The inoculum was rubbed onto leaves dusted with 600 grit carborundum. Infected tobacco plants were grown in glasshouses at 20°C-30°C. The titre of each inoculum was assessed by inoculation to Chenopodium amaranticolor, a local lesion host for PVX and comparison of the lesions produced with lesions produced by purified viral RNA or virus particles.

DNA manipulations

Plasmid DNAs were all propagated in E^. coli strain MC1022 and DH5α or in a (dam-) strain if it would be necessary to digest the DNA with Bell. Restriction enzyme digestions, ligations, transformations of E^ coli and other standard manipulations of DNA were carried out as described, or according to standard procedures of Sambrook et al., Molecular Cloning: A Laboratory^' Manual, Second Edition, Cold Spring Harbor Laboratory Press (1989) .

Plant Transformation

For plant transformation, the binary plasmid vectors

(see below) were mobilised into Aσrobacterium tumefaciens (strain LBA4404) using E__s_ coli HB101 harbouring pRK2013, as described by Bevan (Nucl. Acids Res. , 12., 8711)

(1984) ) . Plant transformation was carried out using the leaf disk transformation method of Horsch et a . , (Science, 227, 1229-1231 (1985)) with Nicotiana tabacu (cv. Sa sun NN) . Transformed shoots and roots were selected by propagation on medium containing kanamycin (100 μg ml^"1) . In the selfed FI progeny, or the progeny of crosses, the segregants with the T-DNA were identified by detection of neomycin phosphotransferase (NPT) activity using the method of McDonnell et l. , (PI. Molec. Biol. Rep., 5., 380-386 (1987) ) .

DNA Constructions

The experiments with transformed plants are based on 4 constructions in which 166K of PVX was placed in the expression cassette of a transformation vector for plants. The vector used was pRok2 in which the expression cassette comprises the 35S RNA promoter from cauliflower mosaic virus (CaMV) and the transcriptional termination sequence of the nopaline synthase gene (nos) from Aσrobacterium tumefaciens. The plasmid pRok2 is essentially the same as pRokl (Baulcombe et al.. Nature, 321. 446-449 (1986)), but with the expression cassette in the opposite orientation relative to the T-DNA border sequences and with additional Smal , Kpnl and SstI sites in the expression cassette. Insertion of 166K was at the SstI site of the pRok2 vector in the sense orientation, so that the plants transformed with these constructions would express the protein product of 166K. In one of the 4 constructions, the 166K gene was as isolated directly from a cDNA clone of PVX RNA and was fully functional. The other 3 constructions were modified in the region of 4068-4072 of the PVX-RNA sequence and are summarised in Fig. l of the accompanying drawings.

Fig. 1 indicates the location of SstI sites used in the construction and location within the constructions of the 166K coding sequence and the utagenised region. Sequence coordinates, in white letters, refer to the PVX genome (Huisman et al., J. Gen. Virol., 69., 1789-1798 (1988)). The sequence in lower case letters is the viral cDNA in the constructions and in upper case letters of the variable protein sequence encoded in the mutagenised region. The 35S RNA promoter of CaMV is shown as "35s pro" and the transcriptional terminator of the nopaline synthase gene A_^. tumefaciens as "nos ter".

The four constructions were introduced into the pRok2 plasmid as SstI fragments (Fig. 1) in which the 5' SstI site was adjacent to the position 15 in the viral genome and was derived from the original cDNA clone, pUKl which lacked the 5' 15 nucleotides of PVX-RNA. The 3' SstI site was introduced into the viral genome, adjacent to position 4475 by i-n vitro mutagenesis using the primer TTC AGA GCT CTA AGG ^' TAA CTT AAC GG which is partly complementary to the viral strand of the cDNA, but which introduces an SstI site 3' of 166K.

The sequence encoding the Gly-Asp-Asp (GDD) motif was mutated using the PCR (polymerase chain reaction) based procedure of Kammen et al. , (Nucl. Acids, Res., 17, 5404, (1989)) to generate variants with the sequence indicated in Fig. 1. The mutant sequence was substituted into the 166K gene between Bell restriction sites at position 3507 and 4203.

The DNA constructions were assembled in several stages, as illustrated in Fig. 4 and Fig. 5 of the accompanying drawings. In summary, the stages in the construction of these genes were:-

i) Mutagenesis

Mutations were introduced into the PVX cDNA using a procedure based on the PCR-mediated amplification of a mutated fragment of DNA, with the mutation site included in one of the PCR primers (Kammen et al. , (1989)). This procedure allows the production of DNA fragments containing mutations despite the lack of useful restriction sites in the target DNA. The procedure- involves 2 steps, as illustrated in general terms in Fig. 7. The first step is an amplification of mutated cDNA using as primers a mutagenic oligonucleotide (prl) and a second primer (pr2) from the complementary DNA strand, within a few hundred nucleotides of the site of mutation. The second step is amplification of the mutated DNA, using the product of the first step as one primer and a third oligonucleotide (pr3) from the opposite strand to pr2 and also within a few hundred nucleotides of the site of mutation (Fig. 4) . The oligonucleotides pr2 and pr3 were selected so that restriction enzyme sites X and Y (Fig. 4) may be used in assembly of the final construct to include the mutated DNA. In the mutation of 166K of PVX, the mutagenic oligonucleotides are indicated in Fig. 1: primers 2 and 3 were the standard M13 forward and reverse primers and the substrate DNA was a fragment of the PVX genome between nucleotides 3453 and 4607 recloned into plasmid pUC19. Sites X and Y were Bell sites at 3507 and 4203.

The mutagenesis of the PVX 166K was carried out on a Sau3A fragment of the viral cDNA (positions 3507-4203), cloned into the standard pUC19 vector at the Bamϋl site in plasmid pPVXMLl (Fig. 5a) . The oligonucleotides (M13 reverse) and (M13 forwards) being complementary to the vector sequence were used as pr2 and pr3 respectively and the Sau3A sites from the viral cDNA were used in the next stage of the construction in which the mutagenised DNA was transferred to the BamHl site of vector plasmid pUC19 to produce plasmids pPVXMLlδ (mutation 1, Fig. l) pPVXMLll

(mutation 2, Fig. l) or pPVXML36 (mutation 3, Fig. l) .

The presence of the respective mutations was verified by analysis of the nucleotide sequence in these plasmids.

ii) Reassembly

The 166K was reassembled by transfer of the Sau3A fragments of pPVXMLlδ, 11 and 36 into an intermediate plasmid pPVXML33. The construction of pPVXML33 is also described in Fig. 5b, and started with pUKl, including a partial cDNA clone of PVX-RNA, extending between positions 15 and 6418 of the viral genome. Part of the viral sequence from pUKl was amplified using the M13 forwards primer (M13f) , referred to above, and primer TTC AGA GCT CTA AGG TAA CTT AAC GG (oli 4) which is partly complementary to the viral strand of the pPVXl cDNA (between positions 4458 and 4475) , but which introduces an SstI site (GAG CTC) 3' of the 166K of the PCR product. Using this SstI site and the SstI site from pUKl, the PCR amplified DNA was transferred to plasmid pUC19 to produce plasmid pMLlS. An Xbal. fragment (from position 2945 to the vector derived Xbal site on the 3' side of the replicase) was then transferred to pUC19, at the Xbal site to produce plasmid pML20. The plasmid pML33 was then produced by Bell digestion and religation of pML20 to delete the Bell fragment (positions 3507-4207) .

The cohesive end of the remaining single Bell site

-in pPVXML33 was compatible with the cohesive ends of Sau3A sites, so that the deleted region of pPVXML33 was restored by insertion of the Sau3A fragments from plasmids pPV ML

18, 11 and 36 to generate corresponding plasmids pPVXML

34, 35 and 39.

It was also necessary to construct a derivative of pPVXML33 in which the reinserted region was from non-mutated cDNA of PVX. The reason for this step was that a spurious mutation was introduced at position 4170 during construction of pML15. The plasmids pPVXML 34, 35 and 39 with deliberate mutations around nucleotides 4064-4075 did not contain that mutation, as the DNA for position 4170 was derived from pPVXMLl rather than pPVXML15. The derivative of pPVXML33 without either spurious or deliberate mutations (pPVXML41) was constructed by addition of the Bell fragment (positions 3507-4207) from pPVXl into Bell digested pPVXML33.

The final stage in the reassembly involved transfer of a SstI (adjacent to position 15) - Nael (position 3081) fragment from pUKl together with Nael - Xmal (adjacent to position 4475) fragment from the plasmids pPVXML 34, 35, 39 or 41 into vector plasmid pUC19 digested with SstI and Xjπal to generate plasmids pPVXML 43, 37, 42, 44, respectively.

iii) Transfer to Binary plasmid pRok2

The 166K of pPVXML 43, 37, 42 and 44 could be transferred into the binary vector plasmids as an SstI fragment. The SstI sites are adjacent to positions 15 and 4475 in the PVX cDNA and the recipient plasmid was binary plasmid pRok2 which contains an SstI site between the 35SRNA promoter and the nos terminator (Fig. 1) .

Other constructions used in the experimental work

pTXS - a full length cDNA clone of PVX_UK3 as described by Kavanagh et al. , (Virology in press (1992)) from which infectious RNA may be produced using T7 RNA polymerase.

pTXS(166Kl) - a derivative of pTXS with the Bell fragment between 3507 and 4207 replaced with the same fragment of pPVXML34 (mutant l) . IS pTXS(166K2) - a derivative of pTXS with the Bell fragment between 3507 and 4207 replaced with the same fragment of pPVXML35 (mutant 2) .

pTXS(166K3) - a derivative of pTXS with the Bell fragment between 3507 and 4207 replaced with the same fragment of pPVXML39 (mutant 3) .

pTHS - a full length cDNA clone of PVX_HB, as described by Kavanagh et al. , (Virology, in press 1992) from which infectious RNA may- be produced using T7 RNA polymerase.

pKHK2 - a hybrid cDNA in which the region of pTXS between 709 and 3211 was replaced with the homologous region of pTHS.

pHKH2 - a hybrid cDNA in which the region of pTXS between 709 and 3211 was replaced with the homologous region of pTHS.

The cDNA content of plasmids pTXS, pTHS, pKHK2 and pHKH2 is illustrated diagramatically in Fig. 6 which also shows a schematic representation of the PVX genome with the viral genes represented by the parts labelled 166K, 25K, 12K 8K and coat. In Fig. 6, Spel represents a restriction enzyme site used for linearisation of the plasmid DNA prior to in vitro transcription, T7 represents the location of the promoter for T7 RNA polymerase, Ncol represents a restriction enzyme site used in the construction of hybrid cDNAs and numbers in brackets show the sequence co-ordinates of the restriction enzyme sites.

RESULTS

Mutations in the GDP box affect infectivitv of viral RNA A series of 4 constructions were assembled in which various forms of 166K were introduced into plasmid pTXS, from which infectious PVX RNA may be generated by transcription in vitro, to create the series of plasmids pTXS(166Kl, 2 or 3) in which 166K was modified around position 4064, as indicated in Fig. 1 for the binary plasmids pPVXML47, 38 and 45.

Inoculation of tTXS(166Kl), tTXS(166K2) or tTXS(166K3) to tobacco plants failed to establish infections whereas inoculation of tTXS initiated systemic infection of the plants within 4 days. These data confirmed the essential nature of the GDD motif in the 166K protein of PVX.

Expression of mutant 166K in transgenic tobacco

Each of the versions of 166K was transferred to a binary Ti plasmid vector for plant transformation. Several lines of transgenic tobacco were recovered with each of the 166K constructs and are referred to as lines N, 1, 2, 3 depending on which binary Ti plasmid was used for the transformation (Fig. l) . Thus, line N. 1 is transformant 1 of line N; N.12 is transformant 12, etc. By RNA gel blotting using a probe specific for 166K of PVX it was verified that each of the lines was expressing the transgene as a 4.5kb RNA, as predicted from the construction of the transgene, although there was several fold variation between lines in the level of expression.

Expression of the transgene at the protein level was monitored by protein-gel blotting, using antisera prepared against the 166K protein. The protein was not detected in unfractionated extracts of the transgenic plants, but, if the extracts were fractionated by centrifugation into 30000g pellet and supernatant fractions, sufficient enrichment was obtained in the pellet fractions to allow detection of the 166K protein from all of the transgenic plants. The amount of 166 protein varied widely between different lines. In most instances, the lines producing most of the RNA were also expressing most protein. The accumulation of the 166K protein in the 30000g pellet was also observed in extracts of non-transformed plants infected with PVX, but at a several fold higher level than in the transgenic plants.

The effect of mutant 166K expression on susceptibility to PVX

In preliminary experiments to screen for an effect of transgenic 166K on virus susceptibility, fourteen of the primary transformed plants were propagated as stem cuttings and the rooted explants were inoculated with 5μg of total RNA isolated from PVX_UK3 infected N. clevelandii . With the exception of line 3.3 (mutation 3, Fig. 1, transformed line 3) all of the plants screened were susceptible to infection. Plants of the 3.3 line failed to accumulate virus or show symptoms in 4 separate experiments with a total of 25 plants. The primary transformed plants were also inoculated with a more dilute inoculum (the equivalent of 0.5 μg ml^"1 PVX RNA) when again, most of the lines were susceptible to infection. The exceptions included line 3.3, as before, but in addition, line 3.1. With line 3.1, the two replicates failed to show any symptoms of infection, whereas with lines N12 and N3, one plant was symptom free and the second plant showed unusual symptoms of isolated chlorotic spots on a few of the systemically infected leaves which were quite distinct from the normal mosaic of PVX (see below) . The further experiments were carried out on the self-fertilized FI progeny of these transformed plants.

A second set of transformed plants was obtained as described above and tested with 5μg of total RNA isolated from PVX_UK3 infected N . clevelandii . Of the 39 lines tested, 18 displayed evidence of resistance in that less than 100% of the inoculated plants developed symptoms on the systemically infected leaves. The partial or complete resistance was found in 4/14 lines tested with mutation 1 (Fig. 1) , in 7/13 lines tested with mutation 2 (Fig. 1) and in 7/12 lines tested with mutation 3 (Fig. 1) .

The FI seedlings (selfed) of line 3.3 were first screened for presence of the transgene, using an assay for neomycin phosphotransferase (NPT II) encoded by the selectable marker gene carried on the T-DNA next to the 166K transgene. • They were then inoculated with 5 μg of total RNA isolated from PVX_UK3 infected N . clevelandii . Plants were inspected for symptoms on the inoculated and systemically infected leaves. In each instance, the plants without NPTII (NPT^") activity developed chlorotic lesions on the inoculated leaf between 3 and 6 days post inoculation and the systemic mosaic symptoms developed between 4 and 6 days. Infection developed at the same rate on non transformed plants, showing that these NPT^" plants were as susceptible as non-transformed plants to infection by PVX_UK3 RNA. The NPT⁺ plants developed chlorotic lesions on the inoculated leaves at the same rate as the non-expressing plants. Only l of the 80 NPT⁺ plants of the 3.3 group showed chlorotic lesions on the inoculated leaf.

The FI progeny of all of the other transformed lines were screened in tests similar to those used to detect virus resistance in the FI progeny of line 3.3, with an inoculum of lμg/plant purified viral RNA and with at least 20 plants from a line of transformed plants used in each test. There was complete or partial resistance in 7/19 lines expressing mutant 1, 5/14 lines expressing mutant 2 and 8/15 lines expressing mutant 3 (Fig. 1) . A total of six lines were completely resistant to infection by PVX in these tests (1 with mutation 1, 2 with mutation 2 and 3 with mutation 3; Fig. 1) (including line 3.3).

In Fig. 2. the plants were FI (selfed) progeny of the primary transformants, grouped either as expressors (E) or non-expressors of the transgene (NE) , based on detection of NPT activity. They were inspected for symptom development on the systemic leaves and the results are expressed as a proportion of the symptomatic plants in each group. The plants were recorded as symptomatic, even when the mosaic was mild. The inoculum was 5 μg of total RNA isolated from PVX_UK3 infected N. clevelandii .

There was no delay in development of PVX symptoms of the infected plants in the 2.1 and 3.2 series, irrespective of the NPT phenotype (Fig. 2a and d) .

The NPT^"1" progeny of 2.5, 3.1, 3.3 and 3.8 plants all showed resistance on the systemic leaves (Fig. 2b, c, e and f) . The effect on the 2.5 plants (Fig. 2b) was a delay in symptom development in 30% of the NPT⁺ plants. With the 3.3 and 3.1 plants there was a more pronounced resistance: none of the 3.3, and only 60% of the plants in the 3.1 lines showed symptoms at 20 days post inoculation (Fig. 2c and e)

In Fig. 3, top row illustrates symptoms of the infected progeny of line 3.1 compared to symptoms of an infected, non-transformed plant (NT#4) or a non-infected plant. The plant 3.1 NE was not expressing the transgene; 3.1#27 and 3.1#22 were both expressing the transgene. The plant 3.1#22 ES completely symptom free and was also free of viral RNA, as detected by RNA gel blotting. The plant 3.1 No. 27 showed ameliorated symptoms of PVX infection in which the usual mosaic was replaced by isolated chlorotic lesions. These ameliorated symptoms were found in 50% of the symptomatic 3.1 plants and approximately 30% of the symptomatic plants in the 3.3 and 2.5 series.

In the 3.3 and the 3.1 plants, the amelioration of

PVX_UK3 symptoms was associated with reduced levels of viral RNA: with the 3.3 plants, PVX_UK3 RNA was not detected in either the inoculated or systemic leaves using RNA-gel blotting. The procedure was sufficiently sensitive to detect 0.1% or less of the PVX_UK3 RNA levels in the leaves of fully susceptible plants. In the 3.1 plants, viral RNA accumulation was detected in inoculated leaves of all plants, and the systemically infected leaves of plants showing symptoms of PVX infection. These results indicate that in the 3.1 plants the resistance mediated by the mutant 166K is less extreme than in the 3.3 line.

In the experiments described above, the PVX_UK3 inoculum was introduced in the form of RNA. There was also resistance against sap inocula in which the infectious agent was virus particles. In an experiment with a sap inoculum containing at least 100 μg ml"¹ PVX particles the 3.3 plants expressing the transgene all failed to accumulate PVX and to show symptoms. The 3.3 progeny in this experiment which were not expressing the transgene accumulated PVX and showed symptoms indistinguishably from non-transformed plants. The complete absence of infection in the NPT^"*^" plants demonstrates the extreme effectiveness of the resistance mechanism in the 3.3 line of plants.

Specificity of virus resistance in transgenic plants expressing mutant 166K

There was no evidence for generalised resistance in the plants transformed with the mutant 166K: inoculation with tobacco mosaic virus elicited the same number, and type of lesions on the 3.3 plants as on non-transformed plants. Similarly inoculation of cucumber mosaic virus (strain Y) led to systemic chlorosis and stunting, to the same extent on the non-transformed and 3.3 plants. The 3.3 plants were also challenged with PVX_cp2 which is 78.4% similar to PVX_UK3 at the nucleotide sequence level (Or an et al-, Virus Res., Iβ _r 293-306 (1990)) but the NPT⁺ plants were as susceptible as the non-transformed progeny both in terms of symptom production and PVX RNA accumulation. In further tests, the 3.3 plants were also susceptible to PVX_B and PVX_EX, both of strain group 2 (Cockerham, 1970) and to PVX_HB of strain group 4 (Cockerham, 1970) . Only one other natural strain of PVX has been tested which was unable to infect the NPT⁺ progeny of the 3.3 plant: this strain was PVX_DX of strain group 3, which is 98% similar to PVX_UK3 within the coat protein gene. The sequence of PVX_DX in 166K is not known. Although more strains of PVX remain to be tested, it is likely that the resistance mediated by the expression of the mutant 3 form of 166K from PVX_UK3 is specific to the homologous or near identical strains.

Analysis of hybrid strains of PVX

It is likely that the specificity of the resistance in the 3. plants is due to interactions between the mutant 166K protein product of the transgene and the 166K protein of the inoculated virus. The evidence for these interactions is based on analysis of hybrid strains of PVX, constructed at the cDNA level with components from PVX_UK3 (pTXS) and PVX_HB (pTHS) . Transcripts prepared in vitro from these cDNA clones infected the 3.3 plants in the same way as the parental viral strains: tTXS infected only the NPT^" progeny of line 3.3 whereas tTHS infected both the NPT* and the NPT" progeny. The hybrid viral constructs, pKHK2 and pHKH2 (Fig. 6) both failed to infect the NPT^"1" progeny of line 3.3 but infected NPT" or non-transformed plants quite readily. In the hybrid virus HKH2, a region within the 166K of PVX_UK3 (nucleotides 709-3211) was substituted into the genome of PVX_HB Fig. 6. The failure of this hybrid virus to infect the 3. plants indicates that the region 709-3211 is implicated in the specificity of the resistance effect probably as a result of an interaction between the products of the transgene and the inoculated virus. In principle, this interaction could involve RNA molecules, but a protein mediated interaction is more plausible. The ability of PVX_HB to overcome the resistance effect could be accounted for if the interaction occurs only between closely related strains of virus.

The hybrid virus KHK2, is the reciprocal of HKH2 (Fig. 6) and similarly failed to infect NPT^÷ progeny of

3.3 plants This result shows that the substitution of the region 709-3211 from PVX_HB is not sufficient to ablate the proposed interaction and indicates that the sequences involved extend beyond 709-3211. In addition, the results indicate that an interaction involving sequences either within or outside of, the 709-3211 domain is sufficient to allow the transgenic expression of mutant 166K to confer resistance. It should be possible, therefore to extend the resistance in the transgenic plants by expression of a mutant replicase of a hybrid strain, such as KHK2 or

HKH2 in which there are apparently domains capable of interaction with both PVX_UK3 and PVX_HB.

CMV Constructs

Figures 7a and 7b show details of CMV constructs suitable for demonstrating the effects of mutations at the GDD motif encoded in CMV RNA2. Figure 7a shows a cDNA of CMV RNA2 with deletion of a Sail fragment at the Sail site (1029). This plasmid is referred to as the acceptor plasmid. Figure 7b shows a cDNA clone of the deleted Sail fragment indicating the position encoding the GDD motif (897-905) . This Sa l fragment was mutagenised (c and g) at position 898 so that the GDD motif encoded the sequence ADD. The nucleotide sequence of the cDNA was confirmed and the full length cDNA of the mutant reconstituted by transfer of the Sa l fragment in the correct orientation into the acceptor plasmid. Other mutations to the GDD motif, including GDD to GAD and GDD to GED can be made in a similar manner.

The gene for the protein with the GDD motif or a mutation thereof can be transferred to a binary expression plasmid for plant transformation in the manner described above.

Claims

CLAIMS :

1. A DNA molecule encoding a protein which is a modified form of the viral component of the replicase enzyme of a plant virus in which sequence coding for the motif Gly-Asp-Asp in the wild type enzyme has been replaced by sequence coding for a motif which renders the protein non-functional as a replicase.

2. A DNA molecule as claimed in Claim 1 encoding a protein in which the motif Gly-Asp-Asp has been replaced by the motif Ala-Asp-Asp.

3. A DNA molecule as claimed in Claim 1 encoding a protein in which the motif Gly-Asp-Asp has been replaced by the motif Gly-Ala-Asp or Gly-Glu-Asp.

4. A DNA molecule as claimed in any of Claims 1 to 3 in the form of a plasmid vector containing DNA encoding the modified viral component of the replicase enzyme in a form suitable for expression in a plant cell.

5. A DNA molecule as claimed in any of Claims 1 to 4 wherein the plant virus is cucumber mosaic virus, tomato bushy stunt virus, tomato spotted wilt virus, potato leafroll virus, barley yellow dwarf virus or potato virus

Y.

6. A DNA molecule as claimed in any of Claims 1 to 5 wherein the plant virus is a hybrid virus.

7. A method of imparting to a plant resistance against a specific plant virus which method comprises the steps of: (i) providing a DNA construct which encodes and is capable of expressing in the plant a protein which is a modified form of the viral component of the replicase enzyme of the plant virus in which at least one sequence motif Gly-Asp-Asp in the wild type viral component of the replicase enzyme has been replaced by a motif which renders the protein non-functional as a replicase; and (ii) introducing the construct into the plant by transformation.

8. A method as claimed in Claim 7 in which the DNA construct encodes a protein in which the motif Gly-Asp-Asp has been replaced by the motif Ala-Asp-Asp.

9. A method as claimed in claim 7 in which the DNA construct encodes a protein in which the motif Gly-Asp-Asp has been replaced by the motif Gly-Ala-Asp or Gly-Glu-Asp.

10. A method as claimed in any of Claims 7 to 9 which includes propagating from the transformed plant to provide plants having a phenotype which includes the ability to express the modified viral component of the viral replicase enzyme.

11. A method as claimed in any of Claims 7 to 10 wherein the plant virus is cucumber mosaic virus, tomato bushy stunt virus, tomato spotted wilt virus, potato leafroll virus, barley yellow dwarf virus or potato virus Y.

12. A method as claimed in .any of Claims 7 to 11 wherein the plant virus is a hybrid virus.

13. A plant having incorporated into its genome DNA encoding a protein which is a modified form of the viral component of the replicase enzyme of a plant virus in which sequence coding for the motif Gly-Asp-Asp in the wild type enzyme has been replaced by sequence coding for a motif which renders the protein non-functional as a replicase, said sequence being in a form in which it is capable of expression in the plant or in a part thereof.

14. A plant as claimed in Claim 13 in which the DNA encodes a protein in which the motif Gly-Asp-Asp has been replaced by the motif Ala-Asp-Asp.

15. A plant as claimed in Claim 13 in which the DNA encodes a protein in which the motif Gly-Asp-Asp has been replaced by the motif Gly-Ala-Asp or Gly-Glu-Asp.

16. A plant as claimed in any of Claims 13 to 15 wherein the plant virus is cucumber mosaic virus, tomato bushy stunt virus, tomato spotted wilt virus, potato leafroll virus, barley yellow dwarf virus or potato virus Y.

17. A plant as claimed in any of Claims 13 to 16 wherein the plant virus is a hybrid virus.

18. Plant propagation material derived from a plant as claimed in any of Claims 13 to 17.