Watermelon mosaic virus-Morocco is a distinct potyvirus

Summary The relationship of the Morocco isolate of watermelon mosaic virus (WMV) to WMV 2, soybean mosaic virus (a virus closely related to WMV 2) and the W strain of papaya ringspot virus (PRSV-W), formerly WMV 1, was examined by comparing tryptic peptide profiles using high performance liquid chromatography. The profiles indicated that the coat protein sequence of WMV-Morocco differed substantially from those of the other potyviruses. This conclusion was supported by sequence data from five tryptic peptides from the coat protein of WMV-Morocco, which showed only 61–68% identity to equivalent sequences in PRSV-W, WMV 2 and zucchini yellow mosaic, another potyvirus infecting cucurbits. Based on the above data, and on known correlations between coat protein sequence similarities and potyvirus relationship, it is concluded that WMV-Morocco should be regarded as a distinct potyvirus.     

Coat protein of Potyviruses

Summary The amino acid sequence of the coat protein of watermelon mosaic virus 2 (WMV 2) was determined by a combination of peptide and nucleic acid sequencing. The coat protein of WMV 2 contained 281 amino acid residues including a single cysteine at position 132 and a blocked amino terminus. Comparison with the coat protein sequences of 20 strains of ten distinct potyviruses showed sequence homologies ranging from 43% to 69% except for the N strain of soybean mosaic virus (SMV-N), where the sequence homology with WMV 2 was 83%. This degree of homology and the location of sequence differences between WMV 2 and SMV-N is much closer to that observed between strains of the same virus than that found between distinct potyviruses. These data suggest that WMV 2 and SMV-N may be strains of the same virus.     

Coat protein of Potyviruses

Three strains of passionfruit woodiness virus, Tip Blight (PWV-TB), Severe (PWV-S) and Mild (PWV-M), were compared on the basis of their biological, serological and coat protein structural properties. Each of the strains could be distinguished on the basis of their reactions on selected test plant species but no differences were observed in the serological properties of the three PWV strains. Molecular weight estimates on SDS-PAGE suggest the PWV coat protein contains 275 amino acid residues and sequence data for 269 of these residues is presented. The amino terminal peptide is blocked and has not been sequenced. The coat proteins of PWV-TB and PWV-S, which cause severe symptoms in infected plants, showed only three sequence differences compared to the eleven or twelve sequence changes between their coat proteins and that of the mild strain. The high sequence homology (96-99%) between the three PWV strains is similar to that previously reported for two strains of tobacco etch virus and three strains of sugarcane mosaic virus. Sequence homology between the three strains of PWV and published sequences for the coat proteins of eight distinct potyviruses ranged from 43-71% (average 57%).     

Coat protein of potyviruses

Summary The amino acid sequence of the coat protein of the Johnsongrass (JG) strain of sugarcane mosaic virus (SCMV) has been determined by protein sequencing techniques. The protein contains 303 amino acid residues corresponding to a molecular weight of 33,510 and when compared to the coat proteins of other potyviruses that have been characterized (263–267 residues) is found to have additional residues at its N-terminus. The N-terminus is acetylated as shown by fast atom bombardment mass spectrometry. Partial amino acid sequences of the coat proteins of the other three Australian SCMV strains, sugarcane (SC), Queensland blue couch grass (BC) and sabi grass (Sabi) have also been obtained. The sequence data and the comparative tryptic peptide HPLC profiles showed that the JG coat protein was substantially different from those of the other three SCMV strains, the sequence homology being around 66 per cent. This is in marked contrast to the high sequence homology between SC, BC and Sabi strains (95–100 per cent) but similar to that (51–62 per cent) found between coat proteins of distinct members of the potyvirus group. On the basis of these structural findings and other information on major differences in serological, biological and biochemical properties we believe that the present JG strain should not be considered a strain of SCMV but should be regarded as an independent member of the potyvirus group. The name Johnsongrass mosaic virus is proposed for this new member.     

Molecular characterization and coat protein serology of watermelon leaf mottle virus (Potyvirus)

Summary.  A cDNA library was generated from purified RNA of watermelon leaf mottle virus (WLMV) (Genus Potyvirus). Two overlapping clones totaling 2,316 nucleotides at the 3′terminus of the virus were identified by immunoscreening with coat protein antiserum. The sequence analyses of the clones indicated an open reading frame (ORF) of 2,050 nucleotides which encoded part of the replicase and the coat protein, a 243-nucleotide non-coding region (3′UTR), and 23 adenine residues of the poly (A) tail. The taxonomic status of WLMV was determined by comparisons of the sequence of the cloned coat protein gene and 3′UTR with potyvirus sequences obtained from GenBank. The nucleotide sequence identities of WLMV compared with 17 other potyviruses ranged from 55.6 to 63.5% for the coat protein, and from 37.2 to 48.3% for the 3′UTR. Phylogenetic analyses of the coat protein region and the 3′UTR indicated that WLMV did not cluster with other potyviruses in a clade with high bootstrap support. The coat protein gene was expressed in Escherichia coli and a polyclonal antiserum was prepared to the expressed coat protein. In immunodiffusion tests, WLMV was found to be serologically distinct from papaya ringspot virus type W, watermelon mosaic virus 2, zucchini yellow mosaic virus, and Moroccan watermelon mosaic virus. In Western blots and ELISA, serological cross-reactivity with other cucurbit potyviruses was observed. Serological and sequence comparisons indicated that watermelon leaf mottle virus is a distinct member of the Potyvirus genus. Accepted September 23, 1999     

Possible members of the potyvirus group transmitted by mites or whiteflies share epitopes with aphid-transmitted definitive members of the group

Summary There are at least ten viruses identified in the literature that resemble definitive potyviruses in having flexuous filamentous particles and inducing the formation of pinwheel cytoplasmic inclusions in infected cells but that are transmitted by eriophyid mites, whiteflies or soil fungi and not by aphids, the vectors of the definitive potyviruses. The taxonomic status of these viruses is uncertain at present. Using a broadly cross-reactive antiserum raised against the dissociated coat protein core (residues 68–285) of a definitive potyvirus (Johnsongrass mosaic virus), we have shown that wheat streak mosaic virus which is transmitted by mite and sweet potato mild mottle virus which is transmitted by whitefly have coat proteins that share epitopes with definitive potyviruses. This finding further supports their classification as definitive members of the potyvirus group. The cross-reactive antiserum used here had been shown previously to react with coat proteins of fifteen different definitive potyviruses. The antiserum did not react with coat proteins of potexviruses and tobamoviruses.     

Indigenous and introduced potyviruses of legumes and <Emphasis Type="Italic">Passiflora</Emphasis> spp. from Australia: biological properties and comparison of coat protein nucleotide sequences

Five Australian potyviruses, passion fruit woodiness virus (PWV), passiflora mosaic virus (PaMV), passiflora virus Y, clitoria chlorosis virus (ClCV) and hardenbergia mosaic virus (HarMV), and two introduced potyviruses, bean common mosaic virus (BCMV) and cowpea aphid-borne mosaic virus (CAbMV), were detected in nine wild or cultivated Passiflora and legume species growing in tropical, subtropical or Mediterranean climatic regions of Western Australia. When ClCV (1), PaMV (1), PaVY (8) and PWV (5) isolates were inoculated to 15 plant species, PWV and two PaVY P. foetida isolates infected P. edulis and P. caerulea readily but legumes only occasionally. Another PaVY P. foetida isolate resembled five PaVY legume isolates in infecting legumes readily but not infecting P. edulis. PaMV resembled PaVY legume isolates in legumes but also infected P. edulis. ClCV did not infect P. edulis or P. caerulea and behaved differently from PaVY legume isolates and PaMV when inoculated to two legume species. When complete coat protein (CP) nucleotide (nt) sequences of 33 new isolates were compared with 41 others, PWV (8), HarMV (4), PaMV (1) and ClCV (1) were within a large group of Australian isolates, while PaVY (14), CAbMV (1) and BCMV (3) isolates were in three other groups. Variation among PWV and PaVY isolates was sufficient for division into four clades each (I-IV). A variable block of 56 amino acid residues at the N-terminal region of the CPs of PaMV and ClCV distinguished them from PWV. Comparison of PWV, PaMV and ClCV CP sequences showed that nt identities were both above and below the 76-77% potyvirus species threshold level. This research gives insights into invasion of new hosts by potyviruses at the natural vegetation and cultivated area interface, and illustrates the potential of indigenous viruses to emerge to infect introduced plants.     

Towards a system for the identification and classification of potyviruses: I. Serology and amino acid composition of six distinct viruses

Antigenic relationships of six distinct potyviruses were studied by immunodiffusion tests using highly purified sonicated virus preparations and anti-intact virus sera devoid of detectable antibodies to host-plant antigens. Three variants of bean yellow mosaic virus (BYMV) including BYMV sensu stricto and two variants of pea mosaic virus (PMV and SPMV) were shown to be antigenically very similar and also relatively closely related to lettuce mosaic virus (LMV). Distant antigenic relationships were detected between the BYMV variants and bean common mosaic virus (BCMV); between BCMV and passionfruit woodiness virus (PWV); and between PWV and potato virus Y (PVY). No antigenic relationships were detected between any of these viruses and sugarcane mosaic virus (SCMV). Antibodies in anti-viral sera were very poor in recognizing coat proteins dissociated with LiCl from homologous viruses and failed altogether to recognise those dissociated with pyrrolidine. Attempts to prepare antisera in mice against isolated viral coat proteins dissociated with either LiCl or pyrrolidine were unsuccessful due to poor immunogenicity of the preparations.Electrophoretic mobilities of the viral coat proteins relative to marker proteins in the presence of sodium dodecyl sulphate suggest that the protein subunits of all the viruses studied have molecular weights of about 33,000. However, the coat proteins were prone to partial degradation. The amino acid compositions of the antigenically closely related viruses were very similar, but similarities of those distantly related were no greater than those of the apparently unrelated viruses.The problems in the use of serological and amino acid composition data obtainable with currently available techniques for the classification of potyviruses are discussed.     

Characterisation and epitope analysis of monoclonal antibodies to virions of clover yellow vein and Johnsongrass mosaic potyviruses

Summary Mouse monoclonal antibodies (MAbs) against the Australian B strain of clover yellow vein (C1YVV-B) and the JG strain of Johnsongrass mosaic (JGMV) potyviruses were produced, characterised and the epitopes with which they reacted were deduced. Using intact particles of C1YVV a total of ten MAbs were obtained which reacted strongly with C1YVV-B in an enzyme-linked immunosorbent assay and Western blots. Four of these MAbs (1, 2, 4, and 13) were found to be ClYVV-specific, as they reacted with all five C1YVV strains from Australia and the U.S.A. but not with 11 strains of bean yellow mosaic (BYMV), pea mosaic (PMV), and white lupin mosaic (WLMV) viruses which, together with C1YVV, form the BYMV subgroup of potyviruses. These MAbs failed to react with eight other potyvirus species, including six which infect legumes like the viruses in the BYMV subgroup. The C1YVV MAb 10 was found to be BYMV subgroup-specific. It reacted strongly with 15 of the 16 strains of viruses in the subgroup and gave no reaction with eight other potyviruses. The other five C1YVV MAbs reacted with varying degrees of specificity with the BYMV subgroup viruses and also with other potyviruses. Eight of the C1YVV MAbs (1, 2, 4, 5, 13, 17, 21, and 22) reacted with the intact coat proteins only and not with the truncated (minus amino terminus) coat protein of C1YVV suggesting that the epitopes for these MAbs are located in the surface-exposed, amino-terminal region of the C1YVV coat protein. Comparison of published coat protein sequences of BYMV and C1YVV isolates indicated that the epitopes for the four ClYVV-specific MAbs may be in the amino-terminal region spanning amino acid residues 18 to 30, whereas those for the other four MAbs may be located in the first 17 amino-terminal amino acid residue region. The epitopes that reacted with BYMV subgroup-specific MAb 10 and MAb 30 which reacted with 20 of the 24 potyvirus isolates, are probably located in the core region of C1YVV coat protein as these MAbs reacted with the intact as well as truncated coat protein of C1YVV. Analysis, in Western blot immunoassay, of 17 MAbs raised against virions of JGMV revealed that only two MAbs (1–25 and 4–30) were JGMV-specific, whereas others displayed varying degrees of specificity to different potyviruses. When these MAbs were screened against the intact and truncated (minus 67 amino-terminal amino acid residues) coat proteins of JGMV, the two JGMV-specific MAbs reacted only with the intact coat protein, whereas the other MAbs reacted with the intact as well as with truncated coat proteins, in Western blots. These results suggest that the epitopes for the two JGMV-specific MAbs are located in the surface-exposed amino-terminal 67 amino acid residue region and those for the cross-reactive MAbs are contained in the conserved core region of the JGMV coat protein. Screening of potyvirus MAbs against intact and truncated coat proteins thus appears to be a simple procedure to select virus-specific MAbs to potyviruses.     

Genome sequence of vanilla distortion mosaic virus infecting Coriandrum sativum

The 9573-nucleotide genome of a potyvirus was sequenced from a Coriandrum sativum plant from India with viral symptoms. On analysis, this virus was shown to have greater than 85 % nucleotide sequence identity to vanilla distortion mosaic virus (VDMV). Analysis of the putative coat protein sequence confirmed that this virus was in fact VDMV, with greater than 91 % amino acid sequence identity. The genome appears to encode a 3083-amino-acid polyprotein potentially cleaved into the 10 mature proteins expected in potyviruses. Phylogenetic analysis confirmed that VDMV is a distinct but ungrouped member of the genus Potyvirus.     

Determination of the taxonomic position and characterization of yam mosaic virus isolates based on sequence data of the 5′-terminal part of the coat protein cistron

Summary The sequences of the N-terminal part of the coat protein cistron from six isolates of yam mosaic virus (YMV-TOG, YMV-COT, YMV-12, YMV-CAR, YMV-BU1 and YMV-BU2) were determined. The analysis of the deduced amino acid sequences revealed the presence of consensus motifs characteristic of the potyvirus genus supporting the classification of YMV as a potyvirus member. The alignment of the N-terminal part of the coat protein of YMV-TOG, YMV-COT, YMV-12 and YMV-CAR showed that they were identical in size (152 aa) while YMV-BU1 and YMV-BU2 were shorter (140 aa) due to a deletion of 12 aa. These amino acid sequences exhibit an overall sequence identity ranging from 70.4% to 97.4% while the identity level with the other potyviruses sequenced in the considered region is below 50%, confirming that YMV is a distinct member of the potyvirus genus. The detailed analysis of the amino acid sequence alignment and of the identity levels observed between the N-terminal part of the coat protein of the six YMV isolates lead us to suggest that they have to be considered as distantly related strains of YMV rather than closely related but distinct viruses.The sequence data reported in this paper have been submitted to the EMBL database under the following accession numbers: YMV-TOG: Z48172, YMV-COT: Z48155, YMV-12: Z48173, YMV-CAR: Z48174, YMV-BU1: Z48156, YMV-BU2: Z48157.     

Phylogenetic analysis of the Potyviridae with emphasis on legume-infecting potyviruses

Summary.  The 3′-terminal nucleotide sequences of thirteen authenticated strains of bean common mosaic virus (BCMV) and one strain of bean common mosaic necrosis virus (BCMNV) were obtained. The regions sequenced included the coat protein coding sequence and 3′-end non-coding region. These data, combined with sequence information from other legume-infecting potyviruses and the Potyviridae were used for phylogenetic analysis. Evidence is provided for delineation of BCMNV as distinct from BCMV and the inclusion of azuki mosaic, dendrobium mosaic, blackeye cowpea mosaic, and peanut stripe viruses as strains of BCMV. This relationship defines the members of the BCMV and BCMNV subgroups. These data also provide a basis upon which to define virus strains, in combination with biological data. Other aspects and implications of legume-infecting potyvirus phylogenetics are discussed. Received December 24, 1996 Accepted June 3, 1997     

The complete nucleotide sequences of the coat protein cistron and the 3′ non-coding region of a newly-identified potyvirus infecting sweetpotato,as compared to those of sweetpotato feathery mottle virus

Summary Complementary DNA representing 728 nucleotides of the 3 end of the genomic RNA of sweetpotato virus G (SPV-G), a newly-identified potyvirus infecting sweetpotato, was cloned and sequenced. This sequence was combined with that previously determined for the 5 terminal part of the coat protein cistron of the virus. The whole sequence contained a single open reading frame (ORF) of 1065 nucleotides, with the capacity to encode a coat protein of 355 amino acids, significantly larger than that of other potyviruses. The ORF was followed by an untranslated region of 222 nucleotides and a poly (A) tail. The coat protein of SPV-G was only distantly related to that of known potyviruses, with the exception of sweetpotato feathery mottle virus (SPFMV). Indeed, sequence identity in the C-terminal three quarters of the coat protein (more than 80%) and in the 3 untranslated region (more than 70%) indicate that SPV-G should be considered as closely related to, though distinct from SPFMV. This subset relationship is similar to that previously reported for members of the bean yellow mosaic virus subgroup or the bean common mosaic virus subgroup.     

The Complete Nucleotide Sequence of a Pakistani Isolate of Watermelon Mosaic Virus Provides Further Insights into the Taxonomic Status in the Bean Common Mosaic Virus Subgroup

Watermelon mosaic virus (WMV) is a potyvirus with a worldwide distribution, but is mostly found in temperate and Mediterranean regions. The complete nucleotide (nt) sequence of a Pakistani isolate of WMV (WMV-Pk) was determined and compared with French isolate (WMV-Fr) and other closely related potyviruses. WMV-Pk showed overall identities of 94.4% (nt) and 96% (amino acid; aa) with the WMV-Fr. However, variability was observed in the 5′ UTR and P1 region. Although sequence identities over most of the genome were well above 90% at both the nt and aa levels, reaching 99.6% (aa) in the CP and 100% (aa) in the 6K1 and 6K2, thereby suggesting that WMV-Pk and WMV-Fr are identical strains, but the sequence identities in the P1 region were only 80.6% (aa) and 82.8% (nt), while that in the 5′ UTR was 82%. These differences may be due to different mutation phenomena of a common ancestor virus or mutations caused by different selection pressures in two different agro-ecological zones. The sequence of WMV-Pk is very close to that of Soybean mosaic virus (SMV) over most of the genome, except for the N-terminal region, which is subject to recombination between SMV and Peanut stripe virus (PSV)/Bean common mosaic virus (BCMV), as revealed by Simplot and phylogenetic analyses of N- and C-terminal P1, HC-Pro, and 5′ UTR regions of the genome.     

<Emphasis Type="Italic">Algerian watermelon mosaic virus</Emphasis> (AWMV): a new potyvirus species in the PRSV cluster

A potyvirus was isolated from a naturally infected squash plant in Algeria in 1986. Biological and serological data have revealed that the virus, initially described as H4, is related to other cucurbit-infecting potyviruses, particularly Moroccan watermelon mosaic virus (MWMV) and Papaya ringspot virus (PRSV). To establish unequivocally the taxonomic status of H4, its full-length genome sequence was established. H4 shared identities of 70% and 65% at the amino acid level with MWMV and PRSV, respectively, indicating that H4 is a distinct species of the PRSV cluster. The name Algerian watermelon mosaic virus (AWMV) is proposed for this new potyvirus species.     

A virus related to <Emphasis Type="Italic">Soybean mosaic virus</Emphasis> from <Emphasis Type="Italic">Pinellia ternata</Emphasis> in China and its comparison with local soybean SMV isolates

Summary. A potyvirus isolated from Pinellia ternata in China was characterised and shown to be related to Soybean mosaic virus (SMV). The virus was pathogenic on P. ternata and some soybean cultivars, whereas the local soybean SMV isolate HH5 did not infect P. ternata. Western blot experiments demonstrated a serological relationship between the virus from Pinellia, SMV and Watermelon mosaic virus (WMV). The complete nucleotide sequences of the Pinellia virus (isolate P-1, 9735nt) and of the Chinese soybean SMV isolates HH5 (9585nt) and HZ (9588nt) were determined. A 1733nt sequence at the 3-terminus of a second isolate from Pinellia (isolate P-2) was also determined. The predicted polyprotein of isolate P-1 has 83% amino acid (aa) identity with those of published SMV sequences. In many parts of the genome, aa identity was about 90% but it was much lower in the P1 protein region (24–29%), where it more closely resembled Dasheen mosaic virus (62%). The partial sequence of isolate P-2 had 91%nt identity to P-1 and both isolates resembled a recent sequence in the public databases (AF469171) wrongly named Zantedeschia mosaic virus. The two complete SMV soybean sequences had 93–95%nt identity with those of the previously sequenced isolates and >97% amino acid identity. Phylogenetic analysis and comparisons of coat proteins suggest that the Pinellia, WMV and SMV potyviruses should probably be treated as strains of the same species.     

Identification and sequence analysis of potyviruses infecting crops in Vietnam

Summary Fifty-two virus isolates from 13 distinct potyvirus species infecting crops in Vietnam were identified and the 3′ region of each genome was sequenced. The viruses were: bean common mosaic virus (BCMV), potato virus Y (PVY), sugarcane mosaic virus (SCMV), sorghum mosaic virus (SrMV), chilli veinal mottle virus (ChiVMV), zucchini yellow mosaic virus (ZYMV), leek yellow stripe virus (LYMV), shallot yellow stripe virus (SYSV), onion yellow dwarf virus (OYDV), turnip mosaic virus (TuMV), dasheen mosaic virus (DsMV), sweet potato feathery mottle virus (SPFMV) and a novel potyvirus infecting chilli, tentatively named chilli ringspot virus (ChiRSV). With the exception of BCMV and PVY, this is first report of these viruses in Vietnam. Further, rabbit bell (Crotalaria anagyroides) and typhonia (Typhonium trilobatum) were identified as new natural hosts of the peanut stunt virus (PStV) strain of BCMV and of DsMV, respectively. Sequence and phylogenetic analyses of the entire CP-coding region revealed considerable variability in BCMV, SCMV, PVY, ZYMV and DsMV. Correspondence: A/Prof. Rob M. Harding, Tropical Crops and Biocommodities Domain, Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane 4001, Australia     

Complete genome sequence of Habenaria mosaic virus, a new potyvirus infecting a terrestrial orchid (Habenaria radiata) in Japan

The complete genomic sequence of Habenaria mosaic virus (HaMV), which infects terrestrial orchids (Habenaria radiata), has been determined. The genome is composed of 9,499 nucleotides excluding the 3′-terminal poly(A) tail, encoding a large polyprotein of 3,054 amino acids with the genomic features typical of a potyvirus. Putative proteolytic cleavage sites were identified by sequence comparison to those of known potyviruses. The HaMV polyprotein showed 58 % amino acid sequence identity to that encoded by the most closely related potyvirus, tobacco vein banding mosaic virus. Phylogenetic analysis of the polyprotein amino acid sequence and its coding sequences confirmed that HaMV formed a cluster with the chilli veinal mottle virus group, most of which infect solanaceous plants. These results suggest that HaMV is a distinct member of the genus Potyvirus.     

A new potyvirus from Thunberg fritillary (<Emphasis Type="Italic">Fritillaria thunbergii</Emphasis> Miq.) in Zhejiang,China

Summary. A potyvirus causing mosaic symptoms in Thunberg fritillary (Fritillaria thunbergii) was found at two sites in Zhejiang province, China. The virus was readily mechanically transmitted to its original host but not to any of 17 other widely used plant virus indicators. A polyclonal antiserum raised to purified virus particles reacted with its homologous virus but not with a range of other viruses (including 16 potyvirus species). In electron microscopy, virus particles and inclusion bodies typical of a potyvirus were seen. The complete nucleotide sequence of an isolate from Ningbo was determined. It was 9723 nt long and sequence analyses predicted the standard potyvirus organisation. The partial sequence (1664 nts at the 3′-terminus) of an isolate from Panan was also determined; the two sequences had 96.9% nt identity. In sequence comparisons and phylogenetic analyses with completely sequenced potyviruses, the new virus was most closely related to Lily mottle virus (53.0% aa identity) and Leek yellow stripe virus. The most closely related incomplete sequence in the international databases was for Lycoris mild mottle virus (72.8% nt identity in their coat proteins). These results suggest that the virus studied is a new species in the genus Potyvirus, which we have tentatively named Thunberg fritillary mosaic virus.     

Nucleotide sequence of the 3'-terminal region of potato virus A RNA.

The sequence of the 3'-terminal region of the genome of the potato virus A (PVA) was obtained from two independent cDNA clones. This sequence is 1383 nucleotides long and contains an open reading frame of 1178 nucleotides, ending with the translation termination codon TAA and followed by untranslated region of 205 nucleotides. Since the N-terminal amino acid of the coat protein of PVA was blocked, the position of the putative coat protein cleavage site has been deduced by searching for consensus sequences and by the analogy to other potyviruses. The resulting coat protein is 269 amino acids long and has a calculated MW of 30257. Two independently sequenced cDNA clones show sequence heterogeneity at four nucleotide positions: C422/A422, G432/A432, G446/A446 and T706/C706. Three first nucleotide differences are located at the PVA coat protein N-terminal region and led to the change of the amino acid. The coat protein of PVA displayed significant (73-78%) sequence homology to the coat proteins of six other potyviruses: papaya ringspot virus (PRV), pepper mottle virus (PeMV), plum pox virus (PPV), potato virus Y (PVY), sugarcane mosaic virus (SCMV) and tobacco vein mottling virus (TVMV). Even higher sequence homology (82%) was detected with a coat protein of a seventh potyvirus, tobacco etch virus (TEV). Major differences among the coat protein of PVA and of other potyviruses are located at the N-terminal region of the coat protein.