The real prevalence of BoHV-5 is unknown because outbreaks are sporadic and routine serologic tests do not discriminate between antibodies against BoHV-5 and BoHV-1. Due to that, BoHV-5 sanitary and economic impact may be underestimated. This, in addition to the scarce information available on BoHV-5 biology, has prompted the present study.
The in vitro characterization performed here suggests that the A663 strain (BoHV-5 subtype "b") possesses diminished capacities concerning cell lysis and cell-to-cell spread in comparison with the N569 strain (BoHV-5 subtype "a"). In addition, the in vivo properties of A663 and N569 were analyzed and, as a result, we observed that both BoHV-5 strains induced a similar degree of virulence in cattle.
Regarding the in vitro characterization, the results obtained from the one-step growth kinetic showed that the production of total infectious viral particles was similar between both BoHV-5 strains. However, the release of infectious viral particles to the extracellular media was lower for the A663 strain. The one-step growth kinetic was performed at high MOI, thus, all cells were infected and cell-to-cell spread was not being evaluated. Hence, the lower release of A663 infectious viral particles to the extracellular media suggested a lower lytic capacity for this strain. This suggestion was directly confirmed by the lysis plaque size assay, which showed that lysis plaques generated by A663 were 90% smaller than those generated by N569. Indeed, this dramatic difference led us to suggest that the A663 strain has a less effective cell-to-cell spread in comparison with N569. In this respect, multiple-steps growth kinetic (performed at low MOI) showed the viral titres in both extracellular and total fractions were larger for N569 than for A663, supporting the contribution of cell-to-cell spread to the lysis plaque area reduction observed for the A663 strain (data not shown). To further assess this, the infection plaque size assay was performed and showed that A663 plaques were 80% smaller than those generated by N569 strain. In this context, we concluded that both lytic capacity and cell-to-cell spread are diminished in the A663 strain in comparison with N569. The fact that A663 and N569 reach the same viral titres in their total fractions in the one-step growth kinetics (from 3 to 24 hpi) indicates that viral particles are equally produced by both strains, however, A663 viral particles could be retained for longer times inside the cell due to their difficulties in cell-to-cell spread and cell lysis.
Some aspects of the in vitro viral replication discussed here could have some relevance in vivo, for example: i) in vitro total production of infectious virus could be related with in vivo levels of viral excretion, or ii) in vitro cell-to-cell spread could have consequences on viral dissemination and infection of different animal tissues, or iii) the lytic capacity could be associated with the level of tissues damage and therefore with the clinical signs observed after viral infection. All these correlations could help understanding the fact that "b" subtype of BoHV-5 has only been reported in Argentina [21, 34] and that it seems to have disappeared since it has not been found circulating during the last years . As suggested by our in vitro results, A663 strain could have diminished capacities to be maintained in the cattle population, in contrast with "a" subtypes of BoHV-5, which circulation among Argentinean cattle herds have been recently reported for first time .
The differences observed in the lysis and the infection plaque sizes between both BoHV-5 strains are larger than expected and have not been reported for wild type strains of the same alphaherpesvirus previously. Despite the fact that the viruses used to perform these assays had the same number of cell culture passages (n = 8), the previous passages to this experience on the original viral stocks were unknown. In our work, passage 8 of N569 strain was obtained after successive passages in cell culture from virus excreted by an animal infected with a N569 virus of unknown, and thus high, number of passages. For this strain, no difference in lysis plaque size was observed between passage 8 and the virus of high number of passages used to infect the bovine (data not shown). Concerning A663 strain, passage 8 of A663 strain was also obtained after successive passages in cell culture from virus secreted by an animal infected with this strain, but in this case, the strain used to infect the animal had a low number of passages. Contrary to N569, a sharp difference between A663 passage 8 and A663 of high number of passages was observed in lysis plaque size, being those caused by A663 passage 8 much smaller. This indicates that plaque size increased with the number of passages in cell culture of the non-adapted to cell culture A663 strain (data not shown). Taking all this into consideration, we are not able to exclude differences between both strains strictly due to cell culture adaptation. Additionally, the genetic background of N569 and A663 strains is unknown and we are not able to rule out the possibility of genomic differences having an impact on the in vitro behavior of both BoHV-5 strains.
Concerning the in vivo characterization, in spite of the in vitro differences discussed above, we observed that both BoHV-5 strains induced a similar degree of virulence in cattle. Almost all animals infected with the N569 or A663 strains showed moderate to high levels of viral excretion confirming the virulence of the viruses used. Individual susceptibility or technical problems during the viral inoculation could account for the lack of infection of the animal 191 from the N569 group.
Only one animal infected with A663 strain developed clinical signs associated with BoHV-5 infection. This difficulty in reproducing clinical encephalitis in BoHV-5 infected animals was also faced by others [29, 35]. Indeed, BoHV-5 infection could induce different degrees of disease. In this respect, Meyer et al. , infected 3-month-old calves with N569 BoHV-5 strain and the animals developed severe neurological signs. Additionally, 6 to 8-month-old animals experimentally infected with the A663 strain developed a moderate BoHV-5-induced neurological disease . In other studies, experimental infections in 4 to 6-month-old calves with BoHV-5 TX74 and EC-1 field isolates [25, 29] and Brazilian isolate SV-507 , have been performed without subsequent development of neurological symptoms, although histopathological changes in the CNS were observed. In accordance with these studies, we did not observe neurological signs in most of the infected animals but histopathological examination of CNS samples evidenced the development of BoHV-5-associated encephalitis. Because the age of the animals and the viral dose used in this study were similar to those of previous studies in which clinical signs were observed [26, 27], the lack of neurological signs reported here could be due to the individual immunological status of cattle.
In addition, in agreement with results presented elsewhere , latency was detected in all the animals infected with A663 strain and two out of four animals infected with N569. The negative result obtained from trigeminal ganglion belonging to animal 199 could be due to technical problems with conservation and/or processing of the sample. In the case of animal 191, as discussed above, the reason could have been the lack of infection, as all the other parameters assayed were also negative for this animal.
At this point, we could suggest that although the development of neurological signs is not always achieved under experimental infection with BoHV-5, presence of lesions in the CNS and establishment of latency are more frequently observed.
Finally, we showed that four out of four and three out of four animals seroconverted after infection with A663 and N569 BoHV-5 strains, respectively. In addition, three out of four and two out of four animals infected with A663 and N569, respectively, presented IgA mucosal antibodies. This constitutes the first report of IgA in nasal secretions of animals infected with BoHV-5 and is in agreement with previous reports showing induction of nasal IgA in calves after infection with different BoHV-1 strains . Concerning animals 628 and 192, the lack of IgA detection could be due to the sensitivity of the ELISA test available. In the case of animal 191, as discussed above, the reason could have been the lack of infection.