Molecular characterization of potato virus X : development of detection probes and identification of the resistance-breaking capacity of strain HB

One of the principal threats to potato production is the high susceptibility of this food crop to diseases, the causal agents including bacteria, fungi, mycoplasmas, nematodes, viruses and viroids. In particular, close to 30 different viruses, and one viroid, are known to infect potato worldwide (Salazar, 1990). Most, if not all, potato viruses are tuberperpetuated through clonal propagation, while only few of them (including the viroid PSTVd), are known to be transmitted also through true potato seeds.Potatoes are usually propagated by planting potato seed pieces or sets, or whole tubers. This method of propagation lends itself to the introduction of pathogens from one area to another and to the overwintering of pathogens in the tubers stored for propagative purposes. In particular, in many tropical and sub-tropical countries, yields of potatoes are very low as a result of several factors, but poor quality of the seeds used is one of the main limiting factors. For that reason, sanitation, systems of seed certification, and quarantine programs invest a great importance in attempt to control the spread of potato viruses. The use of virus-free seed potatoes may control at least initially the spread of the viruses, however, it seems to be difficult, if not impossible, to maintain stocks completely free of viruses, especially those viruses which are easily transmitted. Potato virus X (PVX), for example, has a tendency to build up rapidly in seed stocks if not checked regularly (Manzer et al ., 1975).For inspection and certification, rapid and reliable tests for potato viruses are essential. In the case of some crops, such as potatoes, and in the analysis of seed samples, these procedures should allow daily assays of a large number of samples. To apply the techniques in viral seed health evaluation in developing countries it is important to stress the needs of poorly-equipped laboratories for tests which are cheap, simple and labour-saving yet reliable. Further, the sensitivity should be high enough to reliably detect low levels of the disease agent as well as the various strains that may occur in the field.The current interest in virus and viroid detection techniques and the enormous scope for their application both as research tools and in practical large-scale routine testing led to the development of several procedures and methods now widely used. It is essential to point out the importance of a reliable detection of a specific virus (independently from the strain) for the institutions which include in their mandate the distribution of kits for/and the testing of viruses in different areas in the world.The more specific diagnostic methods fall into two classes: the immunological approaches rely on the use of antibodies, usually prepared against the viral coat protein, the second involves the use of hybridization analysis. In the latter case complementary DNA prepared from the purified viral or viroid nucleic acid, or a recombinant DNA clone of that nucleic acid, is incubated with plant extracts to test for the presence of viral or viroid nucleic acid by the formation of a highly specific cDNA:viral nucleic acid hybrid (Chapter 2).Up to now most nucleic acid probes have been prepared by radioactive labeling which has obvious disadvantages for routine use. However, an increasing number of nonradioactive labeling systems have been developed and the most relevant ones are discussed in Chapter 3. Most of these involve enzymes to give color reactions. A limitation of nonradioactive approaches for large-scale practical application is still the interference by components of plant sap. The sensitivity of detection, however, is often similar to that of the most commonly used serological methods, As well as for detection of viruses in plant samples, nucleic acid hybridization can also be used to assess relationships between viruses.Potato virus X (PVX), is the type member of the genus Potexvirus . PVX is known also as potato latent virus (Miller and Polland, 1977), potato mottle virus, potato virus B, healthy potato virus, and potato mild mosaic virus (Harrison, 1971; Smith, 1972). PVX is of worldwide distribution and it is considered as the most common virus infecting potato (Schultz and Bonde, 1944). Since the potato originated in the Andean region of South America and was later introduced to Europe and other continents, it has been proposed that potato viruses, like PVX, were also spread from this region into other areas of potato production and that the divergences observed among the different virus strains and isolates could reflect adaptation to different Solanum varieties. In the case of PVX several studies have been conducted on variability and distribution of the different strains and on the consequent (in)ability of the techniques currently used to detect all variants. Torrance et al. (1986) found a large degree of serological variation among PVX strains (which could be divided into 4 main serological groups or "serogroups"), they also concluded that serological reactions do not always correlate with the specific classification (Cockerham, 1955) based on resistance genes. Strains of PVX have also been distinguished as PVX Aserotype (found only in South America), and the PVX 0or common serotype, found elsewhere (Fernandez-Northcote and Lizárraga, 1991). It has been shown that differences found in the coat protein gene (which determines serological variability or similarity between strains) do not necessary correspond to similar divergences in other regions of the viral genome (Chapter 4).In Chapter 4 two selected probes, pX61, derived from PVX strain cp (serotype PVX A) and probe pPVX 19, derived from an European strain-group 3 isolate (serotype PVXo were used and their specificity tested on a broad range of PVX isolates and strains. Probe pX61 corresponded to the region between nucleotides 3008 and 4107 of the genomic PVX cp RNA, probe pPVX19 corresponded to the region between nucleotides 2909 and 3845 in the PVX genome, thus comprising a large portion of overlap. Both probes were derived from a highly conserved region in the 3' end of the PVX ORF1, sharing an overall homology of about 78 %. In general, probe pX61 detected strain cp (homologous strain, PVX A ) and HB (also PVX Aserotype) with a strong signal, but poorly isolates belonging to serotype PVXo On the contrary, probe pPVX19 detected the isolates belonging to serotype PVX 0readily, and cp and HB poorly. However, several PVX 0isolates, mostly from Bolivia, reacted with this probe as weakly as PVX Aisolates did and were more readily detected by probe pX61. Despite the use of two probes derived from the highly conserved region in the 3' end of the PVX ORF1, a high specificity in the detection was observed, as demonstrated by the ability of pX61 and pPVX19 to detect isolates belonging to the serotype from which they were prepared or closely related isolates. The results suggest the existence of variability in the nucleotide sequence among the different PVX strains and isolates tested, even belonging to the same serotype and/or serogroup. The differences observed among PVX Aand PVX 0isolates stress the importance of using the appropriate technique for detection, especially for breeding programs for resistance, and quarantine purposes. As a next step towards the development of a universal probe for broad spectrum detection of PVX isolates, a chimaeric recombinant probe was prepared, consisting of sequences from both original probes pX61 and pPVX19. This recombinant probe (pX6119) gave a strong reaction (and therefore reliable detection) with all isolates tested so far (Chapter 5). This result indicates that recombinant probes, containing cDNA fragments from different strains of a virus (as shown here) or from different viruses, may become powerful tools which may find wide application.The following part of my work, presented here in Chapters 6 and 7, was focussed on the characterization of a virulent strain of PVX. Different genes have been found in Solanum sp. conferring resistance to PVX. Hence, Cockerham (1955) classified known PVX strains into four groups on the basis of their interactions with the dominant resistance genes Nx. and Nb, which, in cultivated varieties of potato, determine a hypersensitive response, and with the extreme resistance gene Rx. Group 1 strains are able to infect only susceptible plants. Several PVX strains, classified in groups 2 and 3, are able to overcome genes Nx and Nb, respectively. Finally, group 4 strains overcome both Nx and Nb but fail to infect plants carrying gene Rx. Only one strain, PVX HB , isolated in Bolivia, has been reported to be able to overcome all Nx, Nb and Rx resistance genes (Moreira etal. , 1980). This strain is unique in that it causes typical PVX symptoms in most common indicator species but it does not produce local lesions in inoculated leaves of Gomphrena globosa which is the main indicator host for PVX.PVX HB , is serologically very closely related to PVX cp (Fribourg, 1975), the common Peruvian strain from the resistance group 2. Using polyclonal antibodies in diagnostic tests such as direct ELISA, latex agglutination or electron microscope serology it is not possible to differentiate between these two strains. Indeed, Torrance et al. (1986) could distinguish PVX HB , from other PVX isolates only using selected monoclonal antibodies. However, PVX HB , behaves on potato genotypes differently from this and all other known strains of PVX.In Chapter 6 the genomic RNA of the strain HB was cloned, entirely sequenced, and analyzed. Sequence comparison of the two South American strains (PVX cp and PVX HB ) at the nucleotide level revealed a high rate of homology in all ORFs, ranging between 95 and 98%.The aim of the work was to localize, within the viral genomic sequence, the viral determinant(s) responsible of the ability of strain HB to overcome the resistance provided by the gene Rx, and to understand the mechanism involved. Some information on the involvement of the coat protein in resistance (breaking) and symptoms expression was already available. Santa Cruz et al. (1989) showed with experimental data that the virulence determinants for the Rx gene are located in the 3' portion of the viral genome, outside the replicase gene. Kavanagh etal. (1992) demonstrated that in the case of PVX the coat protein gene plays an important role in viral pathogenicity and, analyzing different hybrid PVX genomes obtained by joining elements of PVX UK3 (a British group 3 non-breaking isolate) and PVX HB , located the resistance-breaking properties of PVX HB , in the coat protein gene. Therefore, as shown in Chapter 6, the nucleotide and amino acid sequences of four PVX strains were compared, but particular attention was given to the coat protein gene. It was supposed that difference in behaviour between HB and the non resistance-breaking strains of PVX could be determined by differences in the secondary structure of a particular region of the coat protein. Coat protein sequence comparisons between HB and three non resistance-breaking strains of PVX, i.e. strains cp (Orman et al., 1990), X3 (Huisman et al., 1988), and S (Skryabin et al., 1988) indicated the presence of eight amino acid residues "unique" for PVX HB . Computer-directed mutational analysis and secondary structure predictions of the viral coat protein allowed the analysis of the possible effect of single amino acids changes in the predicted secondary structure of the coat protein. Only two positions (K 121 and A 226 ) were found to determine variation in the protein structure and were therefore considered as being potentially involved in the resistance-breaking capacity of strain HB. When in the PVX HB coat protein sequence the lysine (K, basic) at position 121 is substituted with a (cp-specific) threonine (T, polar) the alpha-helix at this position is converted into a beta-sheet structure as found in the coat protein structure of PVX cp . The substitution at position 226 (alanine to valine) also determines a slight alteration of the prediction of the coat protein structure, although the modification is located in a highly hydrophobic region and might therefore be masked. Both amino acid substitutions at positions 121 and 226 lead to a reversion of the predicted HB-type coat protein structure to the cp-type coat protein structure.In the meantime, Goulden et al. (1993) analyzed a series of hybrid and mutant isolates of PVX HB and PVX CP4 , a group 4 isolate (Jones, 1985) and concluded that elicitation of extreme resistance expressed in the var. Cara (Rx) is affected by amino acids 121 and 127 of the viral coat protein. PVX HB and hybrid or mutant isolates with lysine and arginine at positions 121 and 127 were able to overcome resistance expressed on var. Cara, whereas those with threonine and arginine were not. In that work, and in accordance with the prediction in Chapter 6, it was also pointed out that position 121 is the major determinantin the resistance-breaking activity, while the importance of residue 127 is proposed to be linked to stability of the mutants.Having been demonstrated that the coat protein, and specifically, only a single amino acid position (121) in the coat protein, affects extreme resistance in Rx genotypes, the interest was devoted to the analysis of a wider pool of Rx genes available. Indeed, wild tuber-bearing Solanum species possess great variability for many traits, having been indispensable sources for resistance to pathogens (Ross, 1986), but only a small portion of the valuable genes found in those species has been incorporated in the genome of commercial varieties. Resistance actually used in breeding programs is still often limited to only few proveniences. In addition, in the specific case of resistance to PVX, even the relationship between the different Rx genes incorporated in various potato varieties is essentially unknown.To gain insight in this matter, in Chapter 7 a recombinant clone KH2 (Kavanagh et al., 1992) and two mutant isolates, CP4-KR and HB-TK (Goulden et al., 1993), were analyzed and tested for their ability to overcome the effects of various extreme resistance genes originating from different Solanum species. The experiments confirmed in all cases the involvement of amino acid 121 of the PVX coat protein in resistance breaking. Hence, the extreme resistance found in S. xchaucha, S. x curtilobum, S. x juzepczukii, and S. vernei as carried by Rx genes (Rx cha , Rx cur , Rx juz , and Rx vm , respectively) interacts with the PVX coat protein in the same manner as Rx (USDA 41956), Rx acl , and Rx adg .However, when the S. sucrense accession OCH 11926 (genotype used OCH 11926.4) reported to be immune to PVX HB (C. Chuquillanqui and L.F. Salazar, personal communication) was inoculated with the same strains and mutant clones, it was found that the recombinant clone KH2 and the mutant isolates CP4-KR and HB-TK were able to systemically infect this genotype and induced a lethal necrosis. The finding that both the recombinant clone KH2 and the mutant isolates CP4-KR and HB-TK, but not the parental isolates, are able to systemically infect this accession, opens up a new question on what kind of resistance mechanism is involved. Although at this time is not possible to exactly define the mechanism of action of this gene, a first interpretation of the results suggests that in this case resistance or susceptibility is not determined only by an interaction with a specific determinant in the viral coat protein, but also by a second determinant, located outside the coat protein and 3' non-coding region.Even though the involvement of the coat protein gene, in resistance (breaking) and symptom expression, has been confirmed in the case of various other plant-virus interactions (Culver and Dawson, 1989; Meshi et al ., 1989; Neeleman et al ., 1991; Shintaku et al ., 1992), it could also be supposed that other viral components are the recognition entity or play an important role in the recognition with the plant gene product(s).Indeed, plant species and viruses of plants interact with each other in various ways which range from non-pathogenesis through to symptoms formation and virus multiplication. The interactions controlled by single genes are likely to involve discrete recognition events between appropriate gene products in the host and factors in the virus. Because of the small size of most viral genomes, and of their limited coding capacity, it is unlikely that any viral gene would be involved solely in the determination of virulence/avirulence. Therefore, a determinant is likely to be located in a gene for some other primary function, such as viral replication, assembling, movement, or transmission, but it could be also possible that the viral nucleic acid (and not a viral protein) could be the recognition entity. Fraser (1986) suggested that in different virus groups, virulence against the resistance genes of various hosts maps to various genomic segments, probably specifying a number of different functions. Examples of known cases in which changes in a viral gene, other than the coat protein gene, determine whether the virus exhibits virulence or avirulence against a host resistance gene, are found in the works of Meshi et al. (1988) and Yamafuji et al. (1991) in which it has been proposed that the product of the Tm1 gene of tomato confers resistance to tobacco mosaic virus (TMV) by interacting with the virus-encoded replicase component, and in the work of Padgett and Beachy (1993) in which the induction of the N genemediated hypersensitive response by tomato mosaic virus (ToMV-Ob) is also supposed to be associated to the viral replicase.In the case of the resistance expressed in the S. sucrense accession OCH 11926, further studies are needed to specifically characterize the gene and the mechanism involved. The approach followed in Chapter 6 and the use of additional mutants could lead to insight in the genetic basis of this resistance, and in the complementary genetic system of PVX and could provide clues to understand the mechanism of operation.