The phylogenetic relationship based on the deduced amino acid sequences of the H protein of fourteen CDV strains were analyzed as shown in Fig. 1. As a result, strains 007Lm, 55L, 66L, 009L, M25CR, 011C, 50Con and 50Cbl were classified into Asia 2 group and strains Ac96I, Th12, 50Sc, 81ND, 82Con and 83mLN were classified into Asia 1 group. Among the Asia 2 strains, 007Lm, 66L, 009L, M25CR, and 011C had identical amino acid sequences of the H gene, although strain 009L differed from strains 007Lm, 66L, M25CR, and 011C in its nucleotide sequence of the H gene (99% identity). In addition, these four strains differed from strains 55L, 50Con, and 50 Cb1 in both amino acid and nucleotide sequences (99% identity of both amino acid and nucleotide sequences). On the other hand, among the Asia 1 strains, 50Sc, 82Con, and 83mLN had identical amino acid sequences of the H gene although strain 82Con differed from strains 50Sc and 83mLN in the nucleotide sequences of the H gene (99% identity). However, these three strains differed from strains Ac96I, Th12, and 81ND in both amino acid (identity 99% with Ac96I and 81ND, and 98% with Th12) and nucleotide sequence (identity 99% with Ac96I and 81ND, and 98% with Th12).

Sequence analyses of the F gene revealed a new initiation codon and extra 27 nucleotides upstream of the usual F gene open reading frame (ORF) in all Asia 2 isolates. To characterize this nucleotide sequence, which extended from 4908 to 4934, various CDV strains as well as the present fourteen strains were compared about the nucleotide sequences from 4901 to 4940 as shown in Fig. 2. Interestingly, only Asia 2 isolates have a nucleotide change from ⁴⁹⁰⁹G to ⁴⁹⁰⁹T which led to the expansion of the F ORF. All Asia 2 isolates (007Lm, 55L, 66L, 009L, M25CR, 011C, 50Con and 50Cbl) had an identical 27 nucleotide sequence. In addition, other nucleotide differences were found among Asia 2 isolates such as ⁴⁹⁰⁷T/⁴⁹⁰⁷C that characterized 50Con and 50Cbl from other strains, also ⁴⁹²⁰G/⁴⁹²⁰A and ⁴⁹³⁰T/⁴⁹³⁰G characterized all Asia 2 strains although American and European as well as Yanaka (Asia 1) strains have the same nucleotide at position 4920 as Asia 2 isolates. The ⁴⁹²⁶A and ⁴⁹²⁸A were shared by all strains in compared to Onderstepoort strain (Fig. 2).

The F genes of fourteen Asian CDV isolates were sequenced and the deduced amino acids were aligned to detect the genetic variations among Asian isolates as shown in Fig. 3. The F gene product is cleaved by cellular proteases of signal peptidase and furin into three regions; signal peptide, F2 and F1 [2]. The cleavage sites, AQIHW in the C-terminus of signal peptide region and RRQRR in the N-terminus of F1 region [16-18], were highly conserved in all Asian isolates as shown in Fig. 3.

The signal peptide region is an important region for location of the precursor F0 into golgi network to cleave into F1 and F2 for fusion activity [19] and is a highly divergent region [2]. As shown in Fig. 3 all Asia 2 isolates have extra 9 amino acids upstream of the N-terminus of this region which is a characteristic for Asia 2 isolates. The amino acid variations were 30 - 32 % and 34 - 35 % while the nucleotide differences were 15 - 16 % and 18 - 19 % for Asia 2 and Asia 1, respectively in comparison with the Onderstepoort strain in the signal peptide region. Both Asia 1 and Asia 2 isolates have common amino acids which differ from Ondestepoort. Moreover, each group has its specific amino acids. However, within the same group there is amino acid(s) characteristic to individual strain as in 007Lm ¹¹⁶Y/¹¹⁶C and inTh12 ⁹T/⁹S and ²⁶R/²⁶K (Fig. 3).

In the F2 region (aa136-224), amino acid differences were found as ²⁰⁸N/²⁰⁸K and ²¹⁶V/²¹⁶L in Asia 1 isolates, whereas ¹⁸⁶D/¹⁸⁶G, ¹⁹³S/¹⁹³N, ¹⁹⁵V/¹⁹⁵I and ²¹⁶V/²¹⁶L in Asia 2 isolates but strains 50Con and 50Cbl have the same amino acid at positions 193 and 195 as Onderstepoort strain (Fig. 3).

The membrane anchored F1 subunit contains the fusion peptide (FP) domain (hydrophobic) at the N-terminus, trans-membrane (TM) domain (hydrophobic) and the cytoplasmic tail (CT) domain at the C-terminus. The fusion peptide domain was highly conserved among all CDV strains. On the other hand, amino acid changes were found in the TM domain as ⁶¹⁶I/⁶¹⁶S in all Asia 1 isolates and ⁶²⁷C/⁶²⁷Y in all Asia 2 isolates. In the CT domain, six amino acid changes were observed within a span of 33 amino acid sequence. Common amino acid changes in all Asian isolates were found as ⁶⁴⁰N/⁶⁴⁰H and ⁶⁴⁶T/⁶⁴⁶A. Specific amino acid changes to all Asia 1 isolates were found as ⁶³⁴R/⁶³⁴Q, ⁶³⁷F/⁶³⁷L and ⁶³⁹H/⁶³⁹Q while those specific to Asia 2 isolates were found as ⁶³⁷H/⁶³⁷L and ⁶⁵⁶R/⁶⁵⁶K. Strains 50Con and 50Cbl had the same amino acids at positions 637 and 656 as Onderstepoort strain.

Adjacent to these domains, heptad repeats were designated as HRA (aa 250-307), HB (aa 328-374) and HRB (aa 557- 601), respectively [6,20,21] as shown in Fig. 3. Amino acid changes were found in HB domain as ³⁶⁶N/³⁶⁶G in all Asian isolates and as ⁶⁰⁰K/⁶⁰⁰R in HRB domain in Asia 1 isolates except for Yanaka strain while HRA domain was conserved

In other regions than the above described domains, common amino acid changes in all Asian isolates but different from Onderstepoort strain were found as ³¹⁷K/³¹⁷R, ⁴³¹V/⁴³¹I and ⁵⁵⁶S/⁵⁵⁶G. Group specific change(s) was found as ³⁹⁵V/³⁹⁵I in Asia 2 isolates while those were found as ³⁰⁹L/³⁹⁰F, ⁴²⁹K/⁴²⁹R, ⁴⁶⁶I/⁴⁶⁶L and ⁶⁰⁷S/⁶⁰⁷G in Asia 1 isolates, but strains Ac96I and Th12 have no amino acid difference at position 607 in compare to Onderstepoort strain. However, an unique amino acid to one or more strains in the same group was detected such as ⁵⁴⁶S/⁵⁴⁶G for M25CR and 011C, ⁴⁷⁸G/⁴⁷⁸C for 81ND and ⁴⁸²W/⁴⁸²L for 50Sc strains as shown in Fig 3.

The F protein of Asian isolates had seven potential glycosylation sites (Fig. 3). Four of them were recognized at positions 141-143, 173-175, 179-181 in the F2 region and 517-519 in the F1 region as reported previously [16-18,21]. Interestingly, the extra three potential glycosylation sites according to the consensus amino acid sequence for N-glycosylation site (N-X-S/T) were found in Asia 1 but not Asia 2 isolates, two at positions 62-64 and 108-110 in the signal peptide region were conserved in all Asia 1 strains while the site 605-607 in F1 region was found in some Asia 1 strains (50Sc, 81ND, 82Con and 83mLN). These seven glycosylation sites are shared by Taiwanese field isolates [22].

Cysteine amino acids are an important factor for the intra molecular disulfide bond and the steric structure of protein. As a result, a total 18 cysteine residues were detected in the F gene product; among them, fourteen residues were located at identical positions in all CDV strains (Fig. 3). Characteristic cysteine residues were located at positions 67, 116, 478 and 627.

The identities between the amino acid and nucleotide sequences of Asia 2 and these of Onderstepoort were 91 % except for 50Con and 50Cbl (92 %), whereas identities between the amino acid and nucleotide sequences of Asia1 and Onderstepoort were 90 % and 91 %, respectively as shown in Table 1. Strains 66L and 009L, 011C and M25CR as well as strains 82Con and 83mLN showed 100 % identities in both amino acid and nucleotide sequences. The similarity of strain Th12 was 99 and 98 % to other Asia 1 strains in amino acids and nucleotides, respectively. While the similarity was 91 and 93 % to all Asia 2 isolates in amino acids and nucleotide sequences except for strains 50Con and 50Cbl amino acid identity was 92 % (Table 1).

The Phylogenetic analysis of F genes revealed that Asia 2 strains clustered into four clades; clade 50Con and 50 Cbl, clade 007Lm, clade 011C and M25CR, clade 66L, 55L and 009L as shown in Fig. 4. On the other hand, Asia 1 isolates have five clades including clade Ac96I, clade Th12, clade 81ND, clade 50Sc, clade 82Con and 83mLN. Interestingly, Asia 1 isolates were appeared to be closer to the European (5804) strain than American strains (A75/17, 98-2645, 00-2601, and 01-2689) when the phylogenetic relationship of F gene (Fig. 4) was compared with that of H gene (Fig. 1). In addition to this, the identical strains in H gene sequences such as 66L, 009L, M25CR, 011C and 007Lm could be distinguished into three distant clades of Asia 2, and strains 82Con, 50Sc and 83meLN into two distant clades of Asia 1 by F gene sequences analysis as shown in Fig. 4.

The phylogenetic relationship among various CDV strains based on the deduced amino acid sequences of the signal peptide region (Fig. 5) showed similar but not identical classification to that of F gene (Fig. 4).

Vero.DogSLAMtag cells were established as described previously [30]. Cells were passaged and maintained in Dulbecco's modified Eagle's medium (D-MEM; autoclavable; Nissui Pharmaceutical Co. Ltd., Tokyo, Japan) supplemented with 10 % fetal bovine serum in a CO₂ incubator at 37°C.

Fourteen CDV strains; 007Lm, 55L, 66L, 009L, M25CR, 011C, 50Con, 50Cbl, Ac96I, Th12, 50Sc, 81ND, 82Con and 83mLN, were isolated and propagated, one or a few times, in Vero.dogSLAMtag cells and stored at -80°C until use. Specimens were collected from diseased dogs as summarized in Table 2.

Vero.DogSLAMtag cells were infected with virus suspensions at MOI = 0.01 and incubated for 18 - 24 hours. When the CPE almost covered the cultures, total RNA was extracted using a MagExtractor™ RNA Extraction Kit (Toyobo Co., Ltd. Osaka, Japan) according to the manufacturer's instructions. Reverse transcription and PCR amplification (RT-PCR) were carried out using a ReverTra-Plus-™-RT-PCR Kit (Toyobo Co., Ltd. Osaka, Japan). The primers used were as follows: 5'ACTTGCCCGATCTCAAGCTA 3' and 5' ATGCTGGAGATGGTTT AATTCAATCG 3'. The forward represents nucleotides 4754 - 4773 of the M-F region in the positive sense and the reverse represents nucleotides 8969 - 8994 of the H-L region in the negative sense. The amplified PCR products (4240 bp) were purified by using a Gene Clean II kit (Biogene, Inc., USA) after agarose gel (0.7 %) electrophoresis, and sequenced directly using a Big Dye^® Terminator v.3.1 cycle sequencing kit (Applied Biosystems, Inc., CA, USA), with appropriate primers designed according to an overlapping strategy Table 3. The sequences were aligned by CLUSTAL W (1.83) Multiple Sequence Alignments (DDBJ) and phylogenetic analysis was carried out by the neighbor-joining method in Mega 3.1 program.