Journal of Wildlife Diseases, 44(3), 2008, pp. 760–765 # Wildlife Disease Association 2008

Ocelots on Barro Colorado Island Are Infected with Feline Immunodeficiency but Not Other Common Feline and Canine

Samuel P. Franklin,1,4 Roland W. Kays,2 Ricardo Moreno,3 Julie A. TerWee,1 Jennifer L. Troyer,1 and Sue VandeWoude11Department of Microbiology, Immunology, and Pathology, Colorado State University, Fort Collins, Colorado 80523, USA; 2New York State Museum and Science Services, Albany, New York 12230, USA; 3Smithsonian Tropical Research Institute, Apartado Postal 0843-03092, Balboa, Ancon, Republic of Panama; 4Corresponding author (email: [email protected])

ABSTRACT: Transmission of pathogens from greater in such locations. However, domestic animals to wildlife populations (spill- without appropriate baseline data, it is over) has precipitated local wildlife extinctions difficult to determine whether the pres- in multiple geographic locations. Identifying such events before they cause population ence of a pathogen represents an intro- declines requires differentiating spillover from duction from domestic animals, signify- endemic disease, a challenge complicated by a ing a potential threat to the wildlife lack of baseline data from wildlife populations population, or whether the pathogen has that are isolated from domestic animals. We been historically present in the wild tested sera collected from 12 ocelots (Leopar- dus pardalis) native to Barro Colorado Island, population. Panama, which is free of domestic animals, for Exposure to multiple pathogens typical- antibodies to feline herpes virus, feline calici- ly considered to reside in domestic animal virus, feline corona virus, feline panleukopenia reservoirs have been documented in virus, canine distemper virus, and feline ocelots (Leopardus pardalis), including immunodeficiency virus (FIV), typically a species-specific infection. Samples also were feline herpes virus (FHV), feline calici- tested for antigens. virus (FCV), feline corona virus (FCoV), Positive tests results were only observed for feline panleukopenia virus (FPV), feline FIV; 50% of the ocelots were positive. We leukemia virus (FeLV), and canine dis- hypothesize that isolation of this population has temper virus (CDV; Schmitt et al., 2003; prevented introduction of pathogens typically attributed to contact with domestic animals. de Carvalho Ruthner Batista et al., 2005; The high density of ocelots on Barro Colorado Filoni et al., 2006; Fiorello et al., 2007). Island may contribute to a high prevalence of Fiorello et al. (2007) reported that from a FIV infection, as would be expected with sample of 10 ocelots sampled in Kyaa-Iya increased contact rates among conspecifics in del Gran Chaco National Park (Bolivia) a geographically restricted population. Key words: Barro Colorado Island, FIV, and the adjacent area, seven and 10 had Leopardus, ocelot, serology. antibodies to CDV and FCV, respectively. Although this national park is approxi- The spillover of pathogens from do- mately 40 km from human settlement, mestic animal species has been a source villagers frequently hunt in areas adjacent of numerous outbreaks in wildlife pop- to the national park with dogs (Canis ulations with disastrous consequences familiaris). This makes it difficult to (Daszak et al., 2000). Evaluating the risk determine whether the exposure of these of spillover in wildlife populations is ocelots to CDV and FCV represents the complicated by a lack of baseline data presence of an endemic wildlife disease, from populations isolated from domestic or whether it occurs as a result of direct or animals (Munson and Karesh, 2002). Most indirect contact with domestic animals in surveys are performed in populations areas adjacent to the national park. in proximity to human settlements and Unlike the aforementioned diseases, domestic animals because the threat of wild felid infection with feline immuno- adventitial disease is presumed to be deficiency viruses (FIV; family Retrovir-

760 SHORT COMMUNICATIONS 761 idae, genus ) is not suggestive of by a crocodile. Additional felid species cross-species transmission. Domestic residing on BCI include margay (Leopar- (Felis catus) FIV has been identified in a dus wiedii) and jaguarundi (Herpailurus wild felid only once (Nishimura et al., yagouaroundi). Puma (Puma concolor) are 1999), and transmission of different strains detected regularly, and jaguars (Panthera of FIV among captive or free-ranging onca) rarely; neither of these species nondomestic felids has been documented reside permanently on the island (Moreno on few occasions (Carpenter et al., 1996; et al., 2006). Troyer et al., 2005; Franklin et al., 2007a). Ocelots were trapped with metal and Rather, species-specific FIV strains have wooden box-traps from January 2001 to been identified in almost all cases of May 2004 and sedated with either a nondomestic cat infections where FIV tiletamine and zolezapam premixture (Tel- genotype analyses were performed; this azolH, Fort Dodge Animal Health, Fort relationship has been demonstrated with a Dodge, Iowa, USA) or ketamine hydro- virus that was isolated from an ocelot chloride and xylazine. Whole blood sam- (Troyer et al., 2005). ples were collected in untreated serum We tested 12 ocelots from an estimated tubes. Samples were spun at 25 3 G for total population of 30 animals (Ziegler, 10 min, and the serum was separated from 2002) from Barro Colorado Island (BCI), the coagulated cells. Samples were stored Panama, for antibodies to FHV, FCV, at 270 C or 220 C until serology was FCoV, FPV, CDV, and FIV. Samples also performed with serum for five common were tested for FeLV antigen. Our objec- feline and canine viruses at the Colorado tive was to acquire baseline antibody State University Diagnostic Laboratory prevalence data from a population of (see Table 1 for a list of pathogens and ocelots isolated from domestic animals. assays used). Barro Colorado Island (1,600 ha; 9u99N, A triple chemiluminescent immunoblot 79u519W) is a hilltop that was isolated with antigen preparations from three from the mainland in 1914 when the distinct strains of feline lentivirus (domes- Chagres River was dammed to create tic cat FIV; puma lentivirus, PLV; and Lake Gatun as part of the Panama Canal. African lion lentivirus) was used to test for The minimum distance between the island antibodies to FIV (Franklin et al., 2007a). and the mainland is 200 m, although small Use of this multi-antigen-based immuno- islands break up this interval in some blot has been shown to enhance sensitivity places. There are no domestic animals without loss of specificity in the detection permitted on BCI, and poaching of native of wild felid FIV (Franklin et al., 2007b). A fauna is limited or nonexistent because the commercial enzyme-linked immunosor- island is heavily guarded (Wright et al., bent assay (ELISA) (FIV/FeLV Combo 2000). The area of BCI was not well SNAPTM test;IdexxInc.,Westbrook, developed before the creation of the canal, Maine, USA) was used as a comparison and it is unlikely that there were any to FIV immunoblot and for detection of domestic animals at this site before 1913. FeLV antigen. Deoxyribonucleic acid was Movement of ocelots between the island extracted using a standard phenol chloro- and the mainland has not been studied in form protocol (Sambrook and Russell, detail, although a low level of ocelot 2001) using the coagulated blood cell emigration from BCI has been document- volume (,2 ml). Nested PCR was per- ed in that one radiocollared BCI ocelot formed using degenerate primers de- was tracked to the mainland. Other signed from the conserved reverse tran- predators may likely limit cross-water scriptase region of FIV pol,using movement of ocelots as evidenced by the GenBank sequences of FIV (accession killing of another radiocollared BCI ocelot nos. M25381 and U11820), PLV (acces- 762 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 3, JULY 2008

TABLE 1. Results of serologic testing of 12 Barro Colorado Island ocelots.

Pathogen Method n + Equivocal 2

Feline calicivirus (FCV) Serum Neutralization (SN) 12 0 NAa 12 Canine distemper virus (CDV) SN 12 0 NAa 12 Feline herpes virus (FHV) SN 12 0 NAa 12 Feline corona virus (FCoV) Hemagglutination Inhibition (HI) 12 0 NAa 12 Feline panleukopenia (FPV) Immunofluorescent Antibody (IFA) 12 0 NAa 12 Feline leukemia virus (FeLV) Antigen ELISAb 12 0 0 12 Feline lentivirus Immunoblot 12 6 1 5 Feline lentivirus Antigen ELISAb 12 5 2 5 a NA 5 not applicable. b FIV/FeLV Combo SNAP test (Idexx). sion no. U03982), and FIV-Oma (Otoco- animals. Alternatively, the BCI ocelot lobus manul; accession no. U56928) population, and that of other sympatric (Troyer et al., 2005). Positive and negative felids, may be too small to ensure controls, including FIV-positive and persistence of these viruses. Sampling of -negative laboratory domestic and additional ocelots on BCI, and sampling of FIV-positive pumas, bobcats, and African ocelots and sympatric domestic animals on lions (Panthera leo) were run concurrent- the mainland, is needed to further inves- ly. The immunoblot, ELISA, DNA extrac- tigate these possibilities. Unlike the other tion techniques, and PCR protocols and viruses we investigated, the prevalence of the relative sensitivity and specificity of FIV was higher in our sample than each are discussed in detail in Franklin et reported previously for ocelots from other al. (2007b). studies. Six of the 12 BCI ocelots tested Except for FIV, all test results were were positive (50%) compared with 0% negative (Table 1). These negative results for 38 captive ocelots from Brazil (Filoni are inconsistent with previous studies in et al., 2003), 0% for 10 wild ocelots in which exposure to all six viruses other than Bolivia (Fiorello et al., 2007), and 6% for FIV is reported (Schmitt et al., 2003; de 90 free-ranging, wild-born, and captive Carvalho Ruthner Batista et al., 2005; ocelots (Troyer et al., 2005). These previ- Filoni et al., 2006; Fiorello et al., 2007). ous studies used a virtually identical We do not believe that this disparity immunoblot protocol or the same com- between studies can be attributed to small mercial ELISA test used in this study. sample size, because with one exception We were unable to amplify genomic our sample size was larger than these FIV sequences from any of the seroposi- previous studies. In addition, very high tive animals using degenerate PCR prim- antibody prevalence rates for some of ers for FIV. These negative results are these pathogens have been reported, consistent with reported difficulties in including a 70% and 100% antibody attempts to amplify and sequence virus prevalence rate for CDV and FCV, from ocelots (Troyer et al., 2005) and respectively, in ocelots from Bolivia (Fior- other nondomestic cat species. PCR am- ello et al., 2007). Moreover, the diagnostic plification of FIV sequences from nondo- tests used in our study and these previous mestic cats is not sensitive because se- studies also were similar. quence variation requires use of These findings support the hypothesis degenerate primers that do not bind viral that these pathogens are not endemic to sequence efficiently (Troyer et al., 2005; BCI. The absence of these pathogens may Brennan et al., 2006; Franklin et al., be the result of isolation from domestic 2007b). Furthermore, proviral load may SHORT COMMUNICATIONS 763 be low, thus decreasing the probability of infections in wild pumas (Biek et al., 2003, adequate binding between sufficient num- 2006a, b), other reports of captive, or in bers of primer and target sequences one study free-ranging animals, have (Brennan et al., 2006). Successful ampli- detected end-stage or subclinical immu- fication of proviral sequences from puma, nologic dyscrasias (Poli et al., 1995; Bull et bobcat, and domestic cat samples run in al., 2003; Brennan et al., 2006; Roelke et parallel to the DNA samples from BCI al., 2006). The PCR amplification and ocelots suggest that there were not inher- sequencing are needed to characterize the ent problems with our PCR assay and that virus infecting the BCI ocelots as is antibody-positive ocelots were not infect- further evaluation of its pathogenicity. ed with a FIV strain associated with other Development of a primer set from the hosts, most notably domestic cats. Because ocelot sequence reported by Troyer et al. cross-species transmission of FIV strains is (2005) would be a possible strategy to rare, we suspect that the virus in our study increase sensitivity of this assay. animals is most likely an ocelot-specific In summary, we hypothesize that isola- virus and that our assay lacked the tion from domestic animals has protected sensitivity to detect it due to the afore- this population from pathogens that are mentioned reasons. normally present in domestic animal The high density of ocelots on BCI may reservoirs, whereas a high density of promote more frequent contact among ocelots on BCI may have increased individuals and enhance FIV transmission, transmission of a species-specific FIV. explaining the high antibody prevalence Further study of BCI ocelots and of observed in this study. The BCI popula- ocelots and domestic animals on the tion has been estimated at 30 ocelots adjacent mainland is warranted to test based on camera-trapping surveys (Zieg- these hypotheses. ler, 2002), and because BCI is only We thank B. Powers and the CSU 1,600 ha, this represents a higher density Diagnostic Laboratory for performing (,2 individuals/km2) than reported from FHV, FCV, FCoV, FPV, and CDV assays, other locations where densities range from and M. Lappin for assistance with ELISA 0.077 to approximately 0.8 ocelots/km2 testing. D. Bogan, R. Mares, C. Fiorello, (Emmons, 1988; Trolle and Kery, 2003; and M. Wikelski were instrumental in Maffei et al., 2005; Trolle and Kery, 2005; ocelot sampling. Capture of animals was Di Bitetti et al., 2006). Density-related approved by the Animal Care and Use transmission is supported by the finding Committee at the Smithsonian Tropical that African lion densities are positively Research Institute. Funding was provided correlated with FIV prevalence (Winter- by the Peninsula Foundation, the National bach et al., 2006). Interestingly, only one Geographic Society, The National Science of five (20%) male BCI ocelots was Foundation, a Merck Merial summer positive for FIV infection, whereas five research fellowship, and a College of or six of the seven females (71–86%) were Veterinary Medicine and Biomedical Sci- FIV immunoblot positive (P.0.05). No ences Summer International Research FIV gender bias has been previously Fellowship. reported, but this possibility should be further evaluated because it may provide LITERATURE CITED insight into the mode of FIV transmission BIEK, R., A. G. RODRIGO,D.HOLLEY,A.DRUMMOND, within this population. C. R. ANDERSON,H.A.ROSS, AND M. POSS. 2003. The implications of FIV infection in the Epidemiology, genetic diversity, and evolution of endemic feline immunodeficiency virus in a BCI ocelot population are unknown. population of wild cougars. Journal of Virology Although studies have not detected clini- 77: 9578–9589. cal affects of FIV in naturally occurring ———, A. J. DRUMMOND, AND M. POSS. 2006a. A 764 JOURNAL OF WILDLIFE DISEASES, VOL. 44, NO. 3, JULY 2008

virus reveals population structure and recent FRANKLIN, S. P., J. L. TROYER,J.A.TERWEE,L.M. demographic history of its carnivore host. LYREN,W.M.BOYCE,S.P.D.RILEY,M.E. Science 311: 538–541. ROELKE,K.R.CROOKS, AND S. VANDEWOUDE. ———, T. K. RUTH,K.M.MURPHY,C.R.ANDERSON, 2007a. Frequent transmission of immunodefi- AND M. POSS. 2006b. Examining effects of ciency viruses among bobcats and pumas. persistent retroviral infection on fitness and Journal of Virology 81: 10961–10969. pathogen susceptibility in a natural feline host. ———, ———, ———, ———, R. W. KAYS,S.P.D. Canadian Journal of Zoology-Revue Canadienne RILEY,W.M.BOYCE,K.R.CROOKS, AND S. De Zoologie 84: 365–373. VANDEWOUDE. 2007b. Variability in assays used BRENNAN, G., M. D. PODELL,R.WACK,S.KRAFT,J.L. for detection of lentiviral infection in bobcats TROYER,H.BIELEFELDT-OHMANN, AND S. VANDE- (Lynx rufus), pumas (Puma concolor), and ocelots WOUDE. 2006. Neurologic disease in captive lions (Leopardus pardalis). Journal of Wildlife Diseas- (Panthera leo) with low-titer lion lentivirus es 44: 700–710. infection. Journal of Clinical Microbiology 44: MAFFEI, L., A. J. NOSS,E.CUELLAR, AND D. I. RUMIZ. 4345–4352. 2005. Ocelot (Felis pardalis) population densi- BULL, M. E., S. KENNEDY-STOSKOPF,J.F.LEVINE,M. ties, activity, and ranging behaviour in the dry LOOMIS,D.G.GEBHARD, AND W. A. F. TOMPKINS. forests of eastern Bolivia: Data from camera 2003. Evaluation of T lymphocytes in captive trapping. Journal of Tropical Ecology 21: 349–353. African lions (Panthera leo) infected with feline MORENO, R. S., R. W. KAYS, AND R. SAMUDIO. 2006. immunodeficiency virus. American Journal of Competitive release in diets of ocelot (Leopar- Veterinary Research 64: 1293–1300. dus pardalis) and puma (Puma concolor) after CARPENTER, M. A., E. W. BROWN,M.CULVER,W.E. jaguar (Panthera onca) decline. Journal of JOHNSON,J.PECON-SLATTERY,D.BROUSSET, AND Mammalogy 87: 808–816. S. J. O’BRIEN. 1996. Genetic and phylogenetic MUNSON, L., AND W. B. KARESH. 2002. Disease divergence of feline immunodeficiency virus in monitoring for the conservation of terrestrial the puma (Puma concolor). Journal of Virology animals. In Conservation medicine: Ecological 70: 6682–6693. health in practice, A. A. Aguirre, R. S. Ostfeld, DASZAK, P., A. A. CUNNINGHAM, AND A. D. HYATT. G. M. Tabor, C. A. House and M. C. Pearl 2000. Emerging infectious diseases of wildlife: (eds.). Oxford University Press, New York, New Threats to biodiversity and human health. York, pp. 118–129. Science 287: 443–449. NISHIMURA, Y., Y. GOTO,K.YONEDA,Y.ENDO,T. DE CARVALHO RUTHNER BATISTA, H. B., F. K. MIZUNO,M.HAMACHI,H.MARUYAMA,H.KI- VICENTINI,A.C.FRANCO,F.R.SPILKI,J.C.R. NOSHITA,S.KOGA,M.KOMORI,S.FUSHUKU,K. SILVA,C.H.ADANIA, AND P. M. ROEHE. 2005. USHINOHAMA,M.AKUZAWA,T.WATARI,A.HASE- Neutralizing antibodies against feline herpesvi- GAWA, AND H. TSUJIMOTO. 1999. Interspecies rus type 1 in captive wild felids of Brazil. Journal transmission of feline immunodeficiency virus of Zoo and Wildlife Medicine 36: 447–450. from the domestic cat to the Tsushima cat (Felis DI BITETTI, M. S., A. PAVIOLO, AND C. DE ANGELO. bengalensis euptilura) in the wild. Journal of 2006. Density, habitat use and activity patterns Virology 73: 7916–7921. of ocelots (Leopardus pardalis) in the Atlantic POLI, A., F. ABRAMO,P.CAVICCHIO,P.BANDECCHI,E. Forest of Misiones, Argentina. Journal of Zool- GHELARDI, AND M. PISTELLO. 1995. Lentivirus ogy 270: 153–163. infection in an African lion: A clinical, patholog- EMMONS, L. H. 1988. A field study of ocelots (Felis ical and virological study. Journal of Wildlife pardalis) in Peru. Revue D Ecologie-La Terre et Diseases 31: 70–74. La Vie 43: 133–157. ROELKE, M. E., J. PECON-SLATTERY,S.TAYLOR,S. FILONI, C., C. H. ADANIA,E.L.DURIGON, AND J. L. CITINO,E.BROWN,C.PACKER,S.VANDEWOUDE, CATAO-DIAS. 2003. Serosurvey for feline leuke- AND S. J. O’BRIEN. 2006. T-lymphocyte profiles mia virus and in captive small in FIV-infected wild lions and pumas reveal Neotropic felids in Sao Paulo state, Brazil. CD4 depletion. Journal of Wildlife Diseases 42: Journal of Zoo and Wildlife Medicine 34: 65–68. 234–248. ———, J. L. CATAO-DIAS,G.BAY,E.L.DURIGON,R. SAMBROOK, J., AND D. W. RUSSELL. 2001. Preparation S. P. JORGE,H.LUTZ, AND R. HOFMANN- and analysis of eukaryotic genomic DNA. In LEHMANN. 2006. First evidence of feline herpes- Molecular cloning: A laboratory manual. Cold virus, calicivirus, parvovirus, and Ehrlichia Spring Harbor Press, Cold Spring Harbor, New exposure in Brazilian free-ranging felids. Journal York, pp. 23–25. of Wildlife Diseases 42: 470–477. SCHMITT, A. C., D. REISCHAK,C.L.CAVLAC,C.H.L. FIORELLO, C. V., A. J. NOSS,S.L.DEEM,L.MAFFEI, MONFORTE,F.T.COUTO,A.B.P.F.ALMEIDA,D. AND E. J. DUBOVI. 2007. Serosurvey of small G. G. SANTOS,L.SOUZA,C.ALVES, AND K. carnivores in the Bolivian Chaco. Journal of VECCHI. 2003. Infeca˜o pelos vı´rus da leucemia Wildlife Diseases 43: 551–557. felina e da peritonite infecciosa felina em felı´deo SHORT COMMUNICATIONS 765

selvagem de vida livre e de cativeiro da regia˜odo WINTERBACH, C. W., P. J. FUNSTON,G.HEMSON,H. Pantanal matogrossense. Acta Scientiae Veter- WINTERBACH,M.E.ROELKE,J.L.TROYER, AND S. inariae 31: 185–188. J. O’BRIEN. 2006. Seroprevalence of feline ———, AND ———. 2005. Camera-trap study of immunodeficiency virus in African lions: Is it ocelot and other secretive mammals in the host density dependent? In 8th international northern Pantanal. Mammalia 69: 409–416. feline research symposium, cat geno- TROLLE, M., AND M. KERY. 2003. Estimation of ocelot mics and infectious. Diseases in the 21st century. density in the Pantanal using capture-recapture Washington, DC, 8–11 October 2006, pp. 102. analysis of camera-trapping data. Journal of WRIGHT, S. J., H. ZEBALLOS,I.DOMINGUEZ,M.M. Mammalogy 84: 607–614. GALLARDO,M.C.MORENO, AND R. IBANEZ. 2000. TROYER, J. L., J. PECON-SLATTERY,M.E.ROELKE,W. Poachers alter mammal abundance, seed dis- JOHNSON,S.VANDEWOUDE,N.VAZQUEZ-SALAT,M. persal, and seed predation in a Neotropical BROWN,L.FRANK,R.WOODROFFE,C.WINTER- forest. Conservation Biology 14: 227–239. BACH,H.WINTERBACH,G.HEMSON,M.BUSH,K. ZIEGLER, C. 2002. A magic web: The forest of Barro A. ALEXANDER,E.REVILLA, AND S. J. O’BRIEN. Colorado Island. Oxford University Press, New 2005. Seroprevalence and genomic divergence York. of circulating strains of feline immunodeficiency virus among Felidae and Hyaenidae species. Journal of Virology 79: 8282–8294. Received for publication 23 February 2007.