BVDV is one of the most important infectious agent in livestock industry because it causes high economic impact, affecting reproduction and facilitating secondary infections [1]. BVDV infects the fetus of a pregnant cow with multiple consequences such as abortion, mummification, being stillborn or born weak. If cattle are infected in early pregnancy with a non-cytopathogenic strain, the virus induces immunotolerance in the fetus. These animals are born persistently infected (PI) and continually shed large amounts of the virus into the environment, representing the main source of infection in a herd [2].

BVDV is not limited to cattle; there is evidence of susceptibility to infection in more than 50 species of the order Artiodactyla [3]. In addition to cattle, persistent infection has been identified in sheep (Ovis orientalis aries), goat (Capra aegagrus hircus), swine (Sus scrofa domestica), alpaca (Vicugna pacos), eland (Taurotragus oryx), Java mouse deer (Tragulus javanicus), mule deer (Odocoileus hemionus), white-tailed deer (Odocoileus virginianus) and mountain goat (Oreamnos americanus) [4–12]. The ability to establish persistent infection in other animal species emphasizes the importance of surveillance in wild species, which may act as reservoirs, hindering control and eradication programs of BVDV within a country.

Zoos keep a large number of domestic and wild species susceptible to BVDV in high proximity or contact between them. Cases of illness or persistent infection with BVDV have been reported in some animals in zoos [7, 8, 13, 14].

In this study, a serological and virological analysis was performed in a captive population of non-bovine artiodactyl species, with antecedents of abortions whose pathology suggests an infectious etiology, in order to detect BVDV infection in the Zoo.

There were different results between S1 and S2 within camelid species. The three guanacos evaluated in S1 were seronegative, however, in S2 one of these animals showed antibodies. In alpacas, none of the 13 individuals sampled in S1 showed antibodies against pestiviruses, whereas in S2 three of 14 alpacas were positive. This group of 14 individuals corresponded to 10 animals present in the previous sampling and 4 new alpacas, of which two and one were seropositive, respectively.

Six of the seven pudús evaluated in S1 were seropositive, while in S2, of six animals evaluated, five that remained from the previous sampling and one new pudú born in the zoo (Pudú 8) were positive to antibodies against pestivirus. The same pudú which was seronegative in S1 was also in S2 (Pudú 7), which made him suspect of being a PI animal (Fig. 1). The two blood samples from the seronegative pudú, obtained nine months apart (January and October 2013), were negative for antibodies against pestivirus but positive for the presence of pestivirus RNA. Virus isolation in MDBK cells indicated that both viruses were non-cytopathogenic strains because no cytopathic effect was observed in culture cells. The PI animal showed no physiological evidence or clinical signs of disease during the study.

Nucleotide identities among the nucleotide sequence of the two viruses showed 100% identity, confirming that it was the same virus. The nucleotide sequence of this “Pudú-Zoo” isolate was submitted to GenBank (accession number KX023408). Phylogenetic analysis showed that the nucleotide sequence of this isolate belonged to a pestivirus closely related to the BVDV subgenotype 1b circulating among Chilean cattle. The percentage of identity between the nucleotide sequence of the Pudú-Zoo isolate and the other BVDV-1b isolates analyzed ranged from 93 to 96%. The highest genomic identity of the nucleotide sequence (96%) of the Pudú/Zoo isolate was seen with the Chilean cattle isolates CHL-113, CHL-170, and CHL-615 (CHL-group); the lowest genomic identity (93%) was seen with cattle isolates NY-1, CP7 and CHL-565(INT-group) (Fig. 2).

The detection of identical pestivirus in two blood samples taken more than three weeks apart in the absence of antibodies against the virus showed a persistent infection of pestivirus in a captive pudú (Pudu puda) in Chile. This is the first report of PI in a deer endemic to South America. This study also reports the first evidence of pestivirus infection in a captive population of wild artiodactyls of South America.

All the animals that resulted positive for antibodies against pestivirus were more than 10 months old at the time of first sampling and were kept in contiguous cages, where direct contact between alpacas and guanaco was possible, while the pudús were housed a few meters apart from the South American camelids. Some animals showed seroconversion in the second sampling, showing the transmission of the virus within the zoo. The most likely pestivirus infecting these animals was BVDV because of the presence of a PI pudú and because BVDV is the only pestivirus reported in Chile, although another pestivirus is not discounted. Even though the main route of transmission of pestivirus is direct contact, the presence of PI animals also facilitates indirect transmission, which can occur through fomites, mechanical vectors or aerosols for a limited time and over short distances [23–25], which may explain the infection of alpacas and guanacos from the pudús; however, other routes of infection for these groups of south American camelids cannot be discounted. The origin of the virus in pudús is also unknown. The PI pudú was born in the zoo, but the identity of his mother was not recorded. Of the six pudús that were seropositive in S1, one was a male born in the zoo and the remaining five came from a captive center in southern Chile, including one female who was pregnant at the time of entering the zoo (Pudú 3). The infection of the mother and the generation of the PI specimen thus may have occurred within the zoo or previously in the other captive center in southern Chile.

Infection with pestiviruses had not been reported in guanacos until now, while in alpacas the prevalence of infection has been reported between 0.9–11.5%, except one case with a herd that had a PI alpaca, where 85% of the animals were seropositive (17 of 20) [5]. Although the proportion of seropositive animals observed in the camelids of our study is lower, the presence of a PI animal among them cannot be discounted. Virus detection in seronegative individuals will be necessary, especially considering that all cases of PI alpacas reported have been by infection with the BVDV-1b subgenotype, which was the subgenotype isolated from the PI pudú. No clinical signs consistent with a pestivirus infection were recorded in the zoo camelids during the inter-sampling period, although they may have gone unnoticed. Acute BVDV infection in South American camelids is almost undetectable, except for vague signs of disease such as anorexia and lethargy [26].

Including our results, antibody titers against pestivirus following natural infection in alpacas ranged from 20 to 480 [26]. However, real titers are uncertain due to possible differences between the infecting virus and the strain used in the diagnostic technique [27].

Neutralizing antibodies against pestivirus in pudús showed an increase at the second sampling, suggestive of reinfection. Titers between sampling periods are comparable because all samples were analyzed simultaneously.

The only previous report of BVDV infection in pudús is from Pizarro-Lucero et al. (2005) [28], who isolated an non-cytopathogenic strain from a free-ranging pudú found dead in southern Chile. This animal had neutralizing antibodies against pestivirus and ulcerative lesions in the nose, mouth, gingival mucosa, interdigital spaces and esophagus, very similar to lesions observed in mucosal disease in cattle. Nevertheless, no cytopathic strain was isolated to associate it with this pathological condition. Pudús did not show clinical signs related to BVDV infection during our study. However, the population of pudús had a history of several deaths by unknown causes. Three pudús included in this study died, two before (Pudú 1, with necropsy findings suggesting infectious disease but no ancillary test to confirm it, and Pudú 2 with necropsy findings not suggesting infectious disease) and one after S2 (Pudú 8 with necropsy findings not suggesting infectious disease). In addition, pudús showed abortions with congenital malformations suggestive of BVDV infection.

Both the pestivirus reported in this study (Pudú-Zoo isolate) and by Pizarro-Lucero et al. (2005) (Pudú-Chillan isolate) were of the BVDV-1b subgenotype. The BVDV-1b subgenotype is one of the most prevalent subgenotypes in cattle in Chile. It is divided into two groups: one group containing only viruses isolated from Chilean cattle (CHL group) and the other group containing viruses isolated from cattle of Chile and other countries (INT group) [29]. Interestingly, the nucleotide sequence from the Pudú-Chillan isolate belongs to the INT group, and the nucleotide sequence from the Pudú-Zoo isolate belongs to the CHL group (Fig. 2). These results suggest that viruses that belong to the BVDV-1b subgenotype have the ability to infect pudús, and transmission of BVDV from cattle to pudú or vice versa could be occurring.

Additional studies should be performed to learn the effects of BVDV infection on pudú health. This virus has shown evidence of causing reproductive disease in white-tailed deer after experimental inoculation and is considered an important emerging disease in the South American camelid industry [30–32]. BVDV has been identified recently as an etiological agent of abortion in sheep flocks, suggesting high potentiality as a threat for new domestic and wild species [33].

The pudú was recently taken off the International Union of Conservation on Nature (IUCN) red list of threatened species; however, enormous fires occurred lately in much of its distribution range in Chile, with consequent mortality and loss of habitat that may change its vulnerability and conservation status [34]. The threat of diseases increases as species move toward extinction [35] and infectious agents that potentially affect individuals and populations through their impact on reproduction, i.e. BVDV and others, could play a role as a threat to conservation of this endemic species of South America [36, 37].

This study showed for the first time a persistent infection of pestivirus in pudú (Pudu puda), a wild deer endemic to South America. The results add a new wildlife species to the list of those with persistent infections of BVDV and a new case report within a zoo.

The proper detection of pathogens that cause persistent infection such as BVDV is also very important for the health of ecosystems, particularly in South American zoos, many of which act as rescue centers for injured or confiscated animals, which are later released to their natural environment.

A PI animal represents a high risk of illness or generation of new PIs for the rest of the captive population in any animal breeding or exhibition center, and the usual management of these animals is their elimination by euthanasia, emphasizing the threat of BVDV for captive endangered species.

Further studies are needed to determine the level of BVDV infection of wild populations of pudús present in southern Chile and the epidemiology of the virus, considering livestock and other susceptible species in the region such as sheep, huemuls (Hippocamelus bisulcus) and red deer (Cervus elaphus).