Vero-DST cells were established as described above 10. HmLu cells were derived from hamster lung cells 13. BHK cells were derived from baby hamster kidney cells 14. All 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.

Thirteen strains of CDV were isolated and propagated, one or a few times, in Vero-DST cells and stored at -80°C until use. Specimens were collected from diseased dogs in Japan except for strain Th12, which was sampled in Thailand. Nine strains were newly isolated except for strains of 007 Lm, 009L, 011C 11, and Ac96I 12. The phylogenetic relationship of nine new strains were determined by the sequencing of their (H) genes 15 where strains 55L, M25CR, 50Con and 50Cbl were assigned as members of Asia 2 group, while strains Th12, 50Sc, 81ND, 82Con and 83mLN were assigned as members of Asia 1 group. The vaccine strain of Onderstepoort (Ond) was used as a control.

Vero-DST cells were seeded on 24-well multi-plates (Sumitomo-Bakelite, Tokyo, Japan) at a concentration of 1 × 10⁵ cells per well and incubated overnight. After the removal of media, the cultures were overlaid with 0.1 ml of each serial 10-fold-diluted virus suspensions in duplicate manner, and incubated for 1 h with rocking at 15-min intervals. After virus adsorption, the suspensions were removed by suction and the cultures were washed once with culture media. The cultures were then overlaid with medium (1 ml) containing 1% methyl cellulose-4,000 (Nacalai Tesque, Kyoto, Japan; methyl cellulose media), incubated for 3 days, then fixed with formalin, stained with crystal violet solution, and plaques were counted. Virus titers were shown as mean value after two independent titrations which were carried out in a duplicate form.

Sub-confluent cultures of HmLu or BHK cells were infected with strains Ond, 007Lm, 55L, 009L, M25CR, 011C, 50Con, 50Cbl, Ac96I, Th12, 50Sc, 81ND, 82Con, or 83mLN of CDV at a multiplicity of infection (MOI) = 1 in culture bottles. When the cultures became confluent (usually after 3-day incubation) the supernatant media was harvested, centrifuged at 3,000 rpm for 10 min to remove cell debris, aliquotted, and stored at -80°C until use as cell-free virus. The residual cells were liberated by trypsinization, and one tenth of the liberated cells underwent further cultivation, while the residual nine-tenths were used to liberate the cell-associated virus by freezing at -80°C, thawing under flowing water, and centrifugation at 3,000 rpm for 10 min to remove cell debris. The supernatant fraction was aliquotted and stored at -80°C until use as cell-associated virus. The cycle of passage was repeated, and all cell-free and cell-associated virus samples were titrated on Vero-DST cells.

To find out whether the hamster cell lines can propagate CDV, a study was started with HmLu cells and CDV strains of Ond and 007Lm (Asia 2 group). When the HmLu cells were infected with the Ond strain at a MOI = 1, the virus induced the cytopathic effect (CPE) of giant cell formation, almost throughout the entire culture within 3-days. The obtained titers of cell-free and cell-associated viruses were 1.3 and 0.6 × 10⁶ PFU/ml, respectively, at the 1^st passage. The passage of infected cells resulted in a low cell density and low virus titer, as shown in Fig. 1. The cell-associated virus titer obtained from HmLu cells infected with the 007Lm strain was 2 × 10² PFU/ml at the 1^st passage. The cell-free virus showed a titer that was about 10-fold lower (1.5 × 10¹ PFU/ml) than that of the cell-associated. The cell-free and cell-associated titers increased gradually up to the 9^th passage, and the highest virus titer (1.3 × 10⁶ PFU/ml) was found for the cell-associated virus obtained at the 9^th passage. This score is comparable with that obtained for HmLu cells infected with the Ond strain. The persistent infection lasted during passages, although the CPE of giant cell formation appeared and increased gradually after the 4^th passage. The infected culture could be transferred up to the 15^th passage. On the other hand, the BHK cells infected with strain 007Lm were revealed to be transferred up to the 6^th passage (data not shown). According to these results, seven passages for infected HmLu and five passages for infected BHK cultures were preset for testing residual 12 CDV strains.

As shown in Table 1, all strains of Asia 2 group could grow in both HmLu and BHK cells. Among the strains of Asia 1 group, Ac96I could not grow in either HmLu or BHK cells and Th12 grew in BHK cells but not in HmLu cells. Moreover, strains 50Sc, 81ND, 82Con, and 83mLN could grow in either HmLu or BHK cells. The cultures of BHK cells infected with strains 50Cb1, Th12, or 81ND could not be transferred over 3 passages, the cultures of both HmLu and BHK cells infected with strain 82Con could not be transferred over 4 passages, and the cultures of HmLu cells infected with strains 55L, 009L, 50Sc, 81ND, and 83mLN also could not be transferred over 7 passages because the massive detachment of giant cells occurred, resulting in a low cell density with extensive appearance of CPE. A typical observation was that strain Th12 grew in BHK cells in a very short period of 3 passages but not in HmLu cells.

The culture of HmLu cells infected with strain 007Lm showed no CPE by the 3^rd passage as if strain 007Lm induced persistent infection. However, the culture showed CPE of giant cell formation (Fig. 2A) after the 4^th passage, and the population of giant cells increased gradually with increasing numbers of passages as the giant cells detached from the substrate as floating giant cells. This giant cell formation did not proceed to the formation of typical syncytia of fusion plaques that were observed in infection of Vero-DST cells with strain 007Lm by the 15^th passage although the population of giant cells increased. At the final stage, the culture showed extensive spread of giant cell formation and floating giant cells (round cells) throughout the entire culture (Fig. 2B). With regard to the infected BHK culture, similar process of CPE development was observed (Fig. 2A and 2B). To characterize giant cell formation, the culture of HmLu cells infected with strain 007Lm was liberated by trypsinization at the 8^th passage and plated into dishes to form colonies.

The dishes were incubated for 7 days. Colonies of various sizes and morphologies were obtained and classified into three categories as shown in Fig. 3(A - C). Colonies with normal cell morphology, which represented the majority of colonies, as shown in Fig. 3A, hybrid colonies composed of normal and giant cell morphology are shown in Fig. 3B and 3B' (low incidence), and colonies composed of giant cells are shown in Fig. 3C (low incidence). Among the giant cells, a typical body, which seemed to be a nucleolus, was found as condensed material in the center area surrounded with a suggested nuclear membrane (Fig. 3D; indicated with an arrow). In other giant cells, several condensed materials were found in a circle (Fig. 3E; indicated with arrows).

Strains 55L, 009L, M25CR, 011C, 50Con, 50Cb1, 50Sc, 81ND, 82Con, and 83mLN induced similar CPE of giant cell formation in the infected HmLu culture at the middle and late stage of infection except at the first infection. Similar CPE of giant cell formation were also induced in the culture of BHK cells infected with every Asian isolate except for strain Ac96I.