Comparisons of the complete genomes of Asian, African and European isolates of a recent foot-and-mouth disease virus type O pandemic strain (PanAsia)

Transcription

1 Journal of General Virology (2003), 84, DOI /vir Correspondence P. W. Mason Comparisons of the complete genomes of Asian, African and European isolates of a recent foot-and-mouth disease virus type O pandemic strain (PanAsia) P. W. Mason, 1 3 J. M. Pacheco, 1 Q.-Z. Zhao 1,2 and N. J. Knowles 3 1 US Department of Agriculture, Agricultural Research Service, Plum Island Animal Disease Center, Greenport, NY 11944, USA 2 Lanzhou Veterinary Research Institute, Lanzhou, Gansu, PR China 3 Institute for Animal Health, Pirbright Laboratory, Ash Road, Pirbright, Woking, Surrey GU24 0NF, UK Received 28 June 2002 Accepted 4 February 2003 During the last 12 years, a strain of foot-and-mouth disease (FMD) virus serotype O, named PanAsia, has spread from India throughout Southern Asia and the Middle East. During 2000, this strain caused outbreaks in the Republic of Korea, Japan, Russia (Primorsky Territory), Mongolia and South Africa (KwaZulu-Natal Province), areas which last experienced FMD outbreaks in 1934, 1908, 1964, 1974 and 1957, respectively. In February 2001, the PanAsia strain spread to the United Kingdom where, in just over 7 months, it caused outbreaks on 2030 farms. From the UK, it quickly spread to the Republic of Ireland, France and the Netherlands. Previous studies that utilized RT-PCR to sequence the VP1-coding region of the RNA genomes of approximately 30 PanAsia isolates demonstrated that the UK virus was most closely related to the virus from South Africa (99?7 % nucleotide identity). To determine if there was an obvious genetic reason for the apparently high level of fitness of this new strain, and to further analyse the relationships between the PanAsia viruses and other FMDVs, complete genomes were amplified using long-range PCR techniques and the PCR products were sequenced, revealing the sequences for the entire genomes of five PanAsia isolates as well as an animal-passaged derivative of one of them. These genomes were compared to two other PanAsia genomes. These analyses revealed that all portions of the genomes of these isolates are highly conserved and provided confirmation of the close relationship between the viruses responsible for the South Africa and UK outbreaks, but failed to identify any genetic characteristic that could account for the unprecedented spread of this strain. INTRODUCTION Despite considerable knowledge of how foot-and-mouth disease (FMD) spreads, the availability of efficacious vaccines and a battery of diagnostic tests, FMD remains one of the most important pathogens of domestic livestock. The recent epizootic in the UK, the control of which resulted in the destruction of millions of animals from over 2000 infected premises, attests to the economic, social and political devastation that FMD can cause. Over the last few years, there has been an explosion in the use of molecular epidemiology to study the spread of FMD and 3Present address: Department of Pathology, Sealy Center for Vaccine Development, University of Texas Medical Branch, 3.206B MMN Pavilion, 301 University Boulevard, Galveston, TX , USA. The GenBank/EMBL/DDBJ accession numbers of the sequences reported in this paper are AJ539136, AJ539137, AJ539138, AJ539139, AJ and AJ other infectious diseases of plants, animals and man. In the case of foot-and-mouth disease virus (FMDV), its rapid rate of evolution, which results from a polymerase without proof-reading activity and the ability of the genome to accommodate considerable amounts of mutations, has made this pathogen particularly amenable to tracking outbreaks through comparisons of the nucleic acid sequences of the viral genome (Samuel & Knowles, 2001). Shortly after the outbreak in the UK, analyses of a highly variable region of the virus genome encoding the capsid protein VP1 (also called 1D) revealed that the virus responsible for the UK outbreak was related to a virus that spread from India in the early 1990s to the Far East (Knowles et al., 2001b). These VP1 sequence data suggested that the closest relative of the UK outbreak virus was an isolate from an outbreak in the Republic of South Africa, which appeared to have resulted from the introduction of an Asian virus into the port of Durban in the province of KwaZulu-Natal (Knowles et al., 2001b) G 2003

2 P. W. Mason and others 1584 Journal of General Virology 84

3 Comparisons of PanAsia FMDV genomes Previous studies with a distantly related type O virus that caused an epizootic in pigs in Taiwan demonstrated that defined genetic changes outside the VP1-regions used for genotype and epidemiological analyses were responsible for its altered phenotype (Beard & Mason, 2000; Knowles et al., 2001a). Moreover, other authors have demonstrated that changes in different regions of the FMDV genome have correlated with altered viral properties (Escarmis et al., 1992; Martinez-Salas et al., 1993; Escarmis et al., 1995; Feigelstock et al., 1996; Núñez el al., 2001). Thus, we undertook analyses of the sequences of full-length genomes of PanAsia isolates to try to determine if specific changes could be responsible for its rapid spread throughout Asia, Africa and Europe. While these studies did not identify any obvious changes associated with the high fitness of these viruses in nature, they did confirm the close genetic relationships among the PanAsia viruses. METHODS Viruses. Convention for naming viruses generally follows the World Reference Laboratory for FMD nomenclature (serotype/three-letter country code/accession #/year). For the six viruses sequenced in this work, passages (p) of viruses prior to sequencing were performed in pigs (PIG), cattle (BOV), primary bovine thyroid cells (BTY), IB- RS-2 porcine kidney cell line (RS), baby hamster kidney cell line 21, clone 13 (BHK) and secondary porcine kidney cells (PK); this is indicated in parentheses after the strain name. O/TAW/2/99 (BHKp3, BTYp1, BHKp2) is a bovine probang isolate from Kinmen Island Prefecture, Taiwan. O/TAW/2/99bov is an experimental bovinederived derivative of O/TAW/2/99 passaged an additional two times in pigs prior to its inoculation into a heifer (see Results). O/Tibet/ CHA/99 (BOVp1, BHKp3) is a bovine isolate from Tibet (reported to OIE on 31 May 1999; OIE Disease Information 12, no. 21, June 4, 1999); the sequence data reported here are similar to those recently deposited in GenBank for this isolate (accession no. AF506822). O/SKR/2000 (BHKp1) is a bovine isolate from early May 2000, from Chungju county of Kyunggi province, Republic of Korea. SAR/19/ 2000 (PKp2, RSp1, BHKp1) is a bovine isolate from KwaZulu-Natal, Republic of South Africa. O/UKG/35/2001 (PIGp2) is a porcine isolate from Cumbria, UK. O/JPN/2000 designates the Japanese isolate, whose sequence data were provided by T. Kanno [accession nos AB (L-fragment) and AB (S-fragment) (Kanno et al., 2002)]. O/SKR/2000gb is used to designate the data from a 2000 Republic of Korea isolate deposited in GenBank (accession no. AF377945). RT-PCR and sequencing. RNA isolated from the indicated source using TRIzol (Life Technologies) was reverse-transcribed with Superscript II polymerase (Life Technologies) and amplified by PCR with Herculase polymerase (Stratagene) and the indicated primers (see Table 1 of online supplementary data, available at sgmjournals.org), using a modification of the long-distance PCR methodology of Tellier et al. (1996). Following amplification, the cdna fragments were purified from acrylamide gels by elution (S-fragment amplicons) or agarose gels with Qiagen Resin and sequenced using selected primers (see Table 1 of online supplementary data, available at and asymmetric amplification with Big Dye terminators (ABI) followed by resolution on an ABI 3700 sequencer. Genome scanning analyses. A program written by one of the authors (N. J. K.) was used to compare the complete genome sequence of O/UKG/35/2001 to those of 15 FMDVs representing five of the seven serotypes. The percentage nucleotide identities were calculated for a sliding window of 300 nucleotides, stepping at 15 nucleotide intervals. The resultant values were plotted on a graph. Phylogenetic analyses. Distance matrices were calculated with the program CLUSTAL X (Thompson et al., 1997). The phylogenetic tree was constructed using a neighbour-joining algorithm (Saitou & Nei, 1987) implemented in the program CLUSTAL X and drawn using the program TREEVIEW 1.6 (Page, 1996). Confidence limits were calculated by the bootstrap re-sampling method (1000 replicates) (Efron et al., 1996) as implemented within CLUSTAL X. RESULTS AND DISCUSSION Sequencing strategy and sequence data acquisition Typically, RT-PCR analyses are compromised by the fact that the sequence data at the ends of the products are fixed by the primer sequences, and therefore the precise genome sequences are replaced with the primer sequences, which may be divergent. In the case of 39-polyadenylated RNAs (such as the FMDV genome), reverse transcription and amplification with oligo(t)-containing primers ensures that the only genetic information lost is the number of adenosine (A) residues in the poly(a) tract (which is, in any case, variable between genomes within an isolate). In the case of the 59 end of the genome, the problem of loss of sequence data upon amplification can be overcome by using terminal transferase to add a homopolymer to the 39 end of singlestranded cdna produced by reverse transcription, and then amplifying through the 59 end of the cdna, utilizing a homopolymer complementary to the added bases and an antisense primer (Frohman et al., 1988). In the case of FMDV, sequencing of its genome by RT-PCR methodology is further complicated by the presence of a long polypyrimidine tract (of variable length) near the 59 end of the genome, which contains predominantly cytosine (C) residues. Using oligonucleotide primers that border this region, it is possible to reverse-transcribe cdnas and amplify them using PCR, although the poly(c) tracts are truncated by successive rounds of PCR. Using these methods, we developed full-length, authentic sequence data for a 1999 Taiwan isolate of the PanAsia topotype (O/TAW/2/99). This sequence reflects the authentic genome in all positions except the number of A residues in the poly(a) tract and the number of C residues in the poly(c) tract (displayed in Fig. 1 as 10 bases), as well as the possible small number of U and A Fig. 1. Sequence data from the UTRs of eight PanAsia isolates (see Methods for sources of viruses). The number of C residues (as well as other possible residues) in the poly(c) tract (displayed here as 10) and the number of A residues in the poly(a) tract (displayed here as 10) were not determined (see text for these and other details). Standard nomenclature is used for reporting base mixtures. (a) 59 UTR; (b) 39 UTR

4 P. W. Mason and others residues that could be present in this poly(c) tract. Several nucleotides (positions 454, 460 and 461) are reported as C/U mixtures (see Fig. 1; denoted by Y), consistent with the expected quasispecies nature of the genome. All other positions could be assigned to a single nucleotide, but among these there were undoubtedly mixtures at multiple positions. Similar methods (with the exception of RACE) were utilized to determine the sequences of five other viruses (O/TAW/ 2/99bov; O/Tibet/CHA/99; O/SKR/2000; O/SAR/19/2000; O/UKG/35/2001). Since terminal transferase tailing was not employed, the first 19 bases of the S-fragment for these five sequences are dictated by the primer, although it should be noted that these bases are highly conserved among all seven serotypes of FMDV (Harris, 1980), so the 19 bases reported in Fig. 1(a) are highly likely to reflect the authentic genome sequences. In addition, primer-determined sequences account for the last base of the S-fragment for O/TAW/2/99bov, O/Tibet/CHA/99, O/SKR/2000 and O/SAR/19/2000, and the last 10 bases of the S-fragment of O/UKG/35/2001. In addition, for all viruses except O/TAW/2/99 and O/Tibet/ CHA/99 the first 22 bases of the L-fragment of the genome are from primer-encoded sequences, although, once again, due to conservation among viruses, these are also likely to be identical to the authentic genome given the high degree of conservation observed elsewhere. Among these five genomes, unambiguous assignments (with the caveats listed above) could be made for all but one base in the O/SAR/19/ 2000 sequence (an A/G mix at position 2182) and one base in the O/SKR/2000 sequence (a C/G mix at position 4009). Interestingly, our O/SKR/2000 sequence data showed numerous differences to a recently deposited sequence (GenBank accession no. AF377945) of the L-fragment of O/SKR/2000 (2?76 % different). 59 and 39 untranslated regions (UTR) The 59 and 39 UTRs of the FMDV genome contain a number of distinct elements that have been identified on the basis of their predicted secondary structure and in some cases, biological functions. Fig. 2(a c) shows the predicted secondary structures of these elements for O/TAW/2/99. These same panels show the position of differences with the other genomes (derived from data shown in Fig. 1a, b). As expected, the vast majority of the sequence differences are in regions between the elements, or do not affect the predicted secondary structure. Of interest is the finding that the 39 UTR of the Republic of Korea sequence deposited in GenBank has five deletions that were not present in the genomes of the seven other PanAsia isolates (Fig. 1b; see also Concluding comments). Polyprotein As with the above analyses of the UTRs, the analyses of the predicted polyprotein sequences of the eight isolates revealed a high level of identity (Fig. 3). In all of the coding regions, the conservation at the protein level is greater than at the nucleotide level, indicating that silent codon changes were greatly preferred (data not shown). Of interest is the fact that there is not a single deletion or insertion within the polyprotein, although the sequence of O/SKR/2000gb contains a deletion closely followed by an insertion in the coding region for VP2 (1B), which is most likely explained by a sequencing error (producing two new codons, shown as CY at positions 66 to 67 of VP2; see Fig. 3). Complete genome analyses The complete genome sequences of the eight PanAsia isolates were aligned with those of other FMD viruses using CLUSTAL X. A program written by one of us (N. J. K.) was then used to compare O/UKG/35/2001 to each of the others, employing a scanning window ; the percentage nucleotide sequence identity measured over 300 bases between pairs of viruses is plotted on the y-axis as the window of comparison is moved in 15 nucleotide steps. The divisions on the ordinate represent the passage by the comparison-window of 300 nucleotides, starting with positions 1 to 300 and terminating at approximately 7947 to Fig. 4 shows all the type O viruses compared to O/UKG/35/2001. All the PanAsia virus sequences are very closely related across the whole genome, with O/SAR/19/2000 being the most closely related and O/SKR/2000gb the most different. The other two type O viruses are clearly distinct. A similar comparison of O/UKG/35/2001 to viruses of other serotypes revealed clear differences among all the coding regions, with the greatest differences detected in the region encoding the outer capsid polypeptides (see Fig. 1 of online supplementary data, available at Interestingly, complete genome analyses revealed that even minor changes in the genome of FMDV can dramatically alter virulence in animals, presumably by altering functional changes through mechanisms that are not readily apparent. Specifically, we found that there were only a small number of differences in sequences of a virus recovered from a probang sample (O/TAW/2/99) that was avirulent in cattle and poorly infectious in pigs (J. M. Pacheco & P. W. Mason, unpublished data) and a derivative, obtained by two passages in pigs (once by inoculation and the second time by contact) and a single passage in a cow, that presented a severe disease (O/TAW/2/99bov; J. M. Pacheco & P. W. Mason, unpublished data). These differences consisted of eight nucleotides (five amino acids) in the polyprotein-coding region (Fig. 3), one nucleotide in the pseudoknot (PKN) region, one nucleotide in the internal ribosome entry site (IRES) region of the 59 UTR (Figs 1 and 2b) and no changes detected in the S-fragment or 39 UTR (Figs 2a, c and 3). Using the complete genome sequence data, the relationships between the eight PanAsia viruses were determined using phylogenetic algorithms (Fig. 5). In general, the year of isolation correlates well with genetic relationship, and these analyses reveal that O/UKG/35/2001 is most closely related to O/SAR/19/2000, consistent with earlier analyses with data from the VP1-coding region of the genome (Knowles et al., 2001b). However, phylogenetic analysis of the VP1-coding 1586 Journal of General Virology 84

5 Comparisons of PanAsia FMDV genomes Fig. 2. Two-dimensional representation of the structures of a portion of the UTRs of O/TAW/2/99, showing the position of differences detected with the UTR sequences of the other eight PanAsia viruses shown in Fig. 1. Filled triangles show sites of differences detected among seven of the strains reported; open triangles show sites where O/SKR/2000gb differs from the other seven isolates evaluated; (a) S-fragment [predicted using MFOLD version 3.1 (Mathews et al., 1999; Zuker et al., 1999)]; (b) pseudoknots (PKn), drawn on the predicted structures of Clarke et al. (1987); cis-acting replicative element (cre), as defined by Mason et al. (2002); and IRES region (domains 2 to 5 of the IRES region derived from Pilipenko et al. (1989); (c) 39 UTR (predicted using MFOLD version 3.1; Mathews et al., 1999; Zuker et al., 1999). Numbering of bases is from the complete genome sequence of O/TAW/2/99 (see Fig. 1). Arrowheads indicate regions of difference (see also Fig. 1). region alone from the same set of viruses shown in Fig. 5 revealed that O/Yunlin/TAW/97 fell slightly closer to the PanAsia group (see Fig. 2 of online supplementary data, available at It is not clear why this was the case; however, the bootstrap value of the branch leading to the O 1 /Kaufbeuren-O 1 /Campos and

6 P. W. Mason and others Fig. 3. For legend see page Journal of General Virology 84

7 Comparisons of PanAsia FMDV genomes Fig. 3. For legend see page

8 P. W. Mason and others Fig. 3. For legend see page Journal of General Virology 84

9 Comparisons of PanAsia FMDV genomes Fig. 3. Alignment of the complete polyproteins of the seven PanAsia isolates and the bovine-passaged derivative of O/TAW/ 2/99. Cleavage sites between the mature polypeptides are indicated above the polyprotein. Fig. 4. Genome-scanning comparisons of genomic sequence data O/UKG/35/2001 vs FMDV type O isolates. The origins of the sequence data not cited in Methods are as follows: O 1 /Kaufbeuren/FRG/66 (M35873, X00871); O/Yunlin/TAW/97 (AF308157). The S-fragment of O/SKR/2000gb was not available for the analysis. PanAsia groups in the complete-genome-tree was quite low (423) in contrast to that leading to the O/Yunlin/TAW/97 and PanAsia groups in the VP1 tree (916) (see Fig. 2 of online supplementary data, available at Although powerful, these analyses can only be used to demonstrate that the UK virus is the closest relative of the South African virus. They cannot be used to imply the mechanism of introduction into the UK. Thus, these data are consistent with either a common source of the outbreaks in these two regions or indicate that South Africa is the source of the strain that caused the outbreak in the UK. Concluding comments Taken together, analyses of the complete genome sequence data reveal a remarkable conservation among the PanAsia virus isolates, which appear to be much more stable than other type O viruses circulating in Asia during the same time period (Knowles et al., 2001a; Samuel & Knowles, 2001). Not only are the PanAsia genomes devoid of specific

10 P. W. Mason and others Comparisons of the sequence data of O/SKR/2000, an isolate from early May 2000 in the Kyunggi province of the Republic of Korea, revealed a large number of differences (2?76 %) to a recently deposited sequence of the L-fragment of another O/SKR/2000 sequence (GenBank accession no. AF377945). This large number of differences strongly suggests that multiple genetic lineages of FMDV were responsible for the outbreaks reported in the Republic of Korea in 2000, a particularly intriguing hypothesis given the reported re-introduction of FMD in ACKNOWLEDGEMENTS Fig. 5. Neighbour-joining tree (outgroup-rooted on SAT2/KEN/ 3/57) showing relationships among the PanAsia viruses and other FMDV strains for which complete (or nearly complete) genome sequence data are available. Relationships were based on a comparison of the complete genome (see Methods). Database accession numbers for the sequences not cited in Methods or the legend to Fig. 4 are as follows: A 12 /UK/119/32 (L11360; M10975); A 22 /Azerbaijan/USSR/65 (X74811, X74812); O 1 /Campos/BRA/58 (AJ320488); C 1 /Santa Pau/ SPA/70 (AJ133357); C 3 /ARG/85 (AJ007347); Asia1/IND/63/ 72 (X83209, Y17973, AF227965, Y09949, AF207524, AF088224, AF207525, AF207526, AF088223, AF207520); SAT2/KEN/3/57 (AJ251473). The S-fragments of O/SKR/ 2000gb and SAT2/KEN/3/57 were not available. Similarly, the 39 UTR was not available for Asia1/IND/63/72. Confidence limits shown indicate the number of times (out of 1000) that bootstrap replicates produced this branch. hot-spots for change, the scanning analyses reveal that there has not been any recent intra- or intertypic recombination of the PanAsia viruses. These findings, together with the information cited above on how slight changes in the polyprotein correlate with profound differences in virulence, demonstrate that with our current level of understanding of FMDV genetics we are unable to identify changes in sequences that are diagnostic for the new properties. Thus, we have failed to identify specific genetic changes that can help explain why the PanAsia viruses have been so effective in their spread across Asia and appear to have, in some areas, replaced enzootic strains of FMDV type O (N. J. Knowles & A. R. Samuel, unpublished observations). If the reasons for the fitness of the PanAsia virus can be ascribed to some marker in laboratory or animal tests, then reverse genetics technology might be suitable for divining the cause of the apparently high fitness of the PanAsia viruses in nature. We thank Toru Kanno, National Institute for Animal Health, Department of Exotic Disease, Kodaira, Tokyo, Japan for providing pre-publication sequence data on O/JPN/2000. We thank Nigel Ferris, WRL for FMD, Pirbright, UK for supplying O/UKG/35/2001, Juan Lubroth, FADDL, PIADC, Greenport, NY, USA for supplying O/SKR/ 2000 and Wilna Vosloo, OVI, Onderstepoort, Republic of South Africa, for supplying O/SAR/19/2000. We thank Jason Gregory for technical assistance. This work was partially supported by a grant from the National Research Initiative Competitive Grants program of USDA/CSREES (grant # ), the Agricultural Research Service of the USDA (CRIS Project # D), and the UK Department of the Environment, Food and Rural Affairs (DEFRA) grant no. SE REFERENCES Beard, C. W. & Mason, P. W. (2000). Genetic determinants of altered virulence of Taiwanese foot-and-mouth disease virus. J Virol 74, Clarke, B. E., Brown, A. L., Currey, K. M., Newton, S. E., Rowlands, D. J. & Carroll, A. R. (1987). Potential secondary and tertiary structure in the genomic RNA of foot and mouth disease virus. Nucleic Acids Res 15, Efron, B., Halloran, E. & Holmes, S. (1996). Bootstrap confidence levels for phylogenetic trees. Proc Natl Acad Sci U S A 93, Escarmis, C., Toja, M., Medina, M. & Domingo, E. (1992). Modifications of the 59-untranslated region of foot-and-mouth disease virus after prolonged persistence in cell culture. Virus Res 26, Escarmis, C., Dopazo, J., Davila, M., Palma, E. L. & Domingo, E. (1995). Large deletions in the 59-untranslated region of foot-andmouth disease virus of serotype C. Virus Res 35, Feigelstock, D. A., Mateu, M. G., Valero, M. L., Andreu, D., Domingo, E. & Palma, E. L. (1996). Emerging foot-and-mouth disease virus variants with antigenically critical amino acid substitutions predicted by model studies using reference viruses. Vaccine 14, Frohman, M. A., Dush, M. K. & Martin, G. R. (1988). Rapid production of full-length cdnas from rare transcripts: amplification using a single gene-specific oligonucleotide primer. Proc Natl Acad SciUSA85, Harris, T. J. (1980). Comparison of the nucleotide sequence at the 59 end of RNAs from nine aphthoviruses, including representatives of the seven serotypes. J Virol 36, Kanno, T., Yamakawa, M., Yoshida, K. & Sakamoto, K. (2002). The complete nucleotide sequence of the PanAsia strain of foot-andmouth disease virus isolated in Japan. Virus Genes 25, Journal of General Virology 84

11 Comparisons of PanAsia FMDV genomes Knowles, N. J., Davies, P. R., Henry, T., O Donnell, V., Pacheco, J. M. & Mason, P. W. (2001a). Emergence in Asia of foot-and-mouth disease viruses with altered host range: characterization of alterations in the 3A protein. J Virol 75, Knowles, N. J., Samuel, A. R., Davies, P. R., Kitching, R. P. & Donaldson, A. I. (2001b). Outbreak of foot-and-mouth disease virus serotype O in the UK caused by a pandemic strain. Vet Rec 148, Mason, P. W., Bezborodova, S. V. & Henry, T. M. (2002). Identification and characterization of a cis-acting replication element (cre) adjacent to the internal ribosome entry site of foot-and-mouth disease virus. J Virol 76, Martinez-Salas, E., Saiz, J. C., Davila, M., Belsham, G. J. & Domingo, E. (1993). A single nucleotide substitution in the internal ribosome entry site of foot-and-mouth disease virus leads to enhanced capindependent translation in vivo. J Virol 67, Mathews, D. H., Sabina, J., Zuker, M. & Turner, D. H. (1999). Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol 288, Núñez, J. I., Baranowski, E., Molina, N., Ruiz-Jarabo, C. M., Sánchez, C., Domingo, E. & Sobrino, F. (2001). A single amino acid substitution in nonstructural protein 3A can mediate adaptation of foot-andmouth disease virus to the guinea pig. J Virol 75, Page, R. D. (1996). TREEVIEW: an application to display phylogenetic trees on personal computers. Comput Appl Biosci 12, Pilipenko, E. V., Blinov, V. M., Chernov, B. K., Dmitrieva, T. M. & Agol, V. I. (1989). Conservation of the secondary structure elements of the 59-untranslated region of cardio- and aphthovirus RNAs. Nucleic Acids Res 17, Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, Samuel, A. R. & Knowles, N. J. (2001). Foot-and-mouth disease type O viruses exhibit genetically and geographically distinct evolutionary lineages (topotypes). J Gen Virol 82, Tellier, R., Bukh, J., Emerson, S. U. & Purcell, R. H. (1996). Amplification of the full-length hepatitis A virus genome by long reverse transcription-pcr and transcription of infectious RNA directly from the amplicon. Proc Natl Acad Sci U S A 93, Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997). The CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25, Zuker, M., Mathews, D. H. & Turner, D. H. (1999). Algorithms and thermodynamics for RNA secondary structure prediction: a practical guide. In RNA Biochemistry and Biotechnology, pp Edited by J. Barciszewski & B. F. C. Clark. Dordrecht: Kluwer Academic Publishers