- Email: [email protected]
Detection of antibodies to Francisella tularensis in cats
Research in Veterinary Science 82 (2007) 22–26 www.elsevier.com/locate/rvsc
Detection of antibodies to Francisella tularensis in cats L. Magnarelli a
a,*
, S. Levy b, R. Koski
c
The Connecticut Agricultural Experiment Station, P.O. Box 1106, New Haven, CT 06504, USA b Durham Veterinary Hospital, 178 Parmelee Hill Road, Durham, CT 06422, USA c 2 L Diagnostics, LLC, 300 George Street, New Haven, CT 06511, USA Accepted 20 June 2006
Abstract Blood samples were obtained from privately owned cats in Connecticut and New York State, USA in 1985–1990, and analyzed for evidence of Francisella tularensis, the etiologic agent of tularemia. Of the 91 sera tested by microagglutination (MA) methods, 11 (12%) contained antibodies to F. tularensis. Analyses of the same sera by indirect fluorescent antibody (IFA) staining methods revealed 22 (24%) positives. There was good agreement in results of both tests (73% concordance). However, we measured higher titers (1:80 to 1:640) with IFA analysis than by MA methods (1:80 to 1:160). Both tests were suitable for general screening purposes. The DNA of F.tularensis was not detected in the 24 antibody-positive sera tested. Cats living in Connecticut and New York State were naturally exposed to F.tularensis or a closely related organism. With exposure to ticks, other biting arthropods, mice, and rabbits, cats are at risk for acquiring F.tularensis infections and can be an important source of information on the presence of this agent in nature. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Cats; Tularemia; Francisella tularensis; Antibodies; Ticks
1. Introduction Tularemia is caused by Francisella tularensis, a highly pathogenic, non-spore-forming coccobacillus that can infect several species of vertebrate hosts. Infection has been documented in humans as well as in Eastern cottontail rabbits (Sylvilagus floridanus), mice, and domesticated mammals, including cats, throughout North America, Eurasia, parts of the Middle East, and the north coast of Africa (Hayes, 2005). The American dog tick (Dermacentor variabilis) is an important vector in the United States, but other Dermacentor, as well as Amblyomma, Haemaphysalis, and Ixodes ticks, maintain enzootic infections in nature (Hopla and Hopla, 1994). Ixodes pacificus and Ixodes scapularis ticks, known to transmit bacterial or protozoan agents that cause Lyme borreliosis, granulocytic anaplasmosis, or human babesiosis, parasitize a wide range of vertebrate hosts and are suspected vectors of F. tularensis (Hayes, *
Corresponding author. Tel.: +1 203 974 8440; fax: +1 203 974 8502. E-mail address: [email protected] (L. Magnarelli).
0034-5288/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.rvsc.2006.06.003
2005), but confirmatory evidence of transmission is lacking. F. tularensis infections can also be acquired from biting insects, by direct contact with blood or other tissues or fluids of infected mammals, by ingestion of contaminated food or water, or by inhaling the pathogen. Although human cases of tularemia are seldom reported in Connecticut and other areas of northeastern United States, repeated outbreaks have occurred recently on Martha’s Vineyard, an island off the coast of Massachusetts (Feldman et al., 2003) and on Long Island, New York (Hayes, 2005). Case reports from Oklahoma (Baldwin et al., 1991; Woods et al., 1998) and New Mexico (Rohrback, 1988) describe tularemia infections in cats. Severity of disease can range from mild to fatal with associated non-specific signs, such as pyrexia, anorexia, lymphadenopathy, or oral ulcers. When allowed outdoors, cats are free roaming and can have extensive exposure to ticks and contact with rodents. Cats, therefore, have great potential for acquiring different pathogens. The main purposes of the present study were to analyze cat sera for antibodies to F. tularensis
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
by means of microagglutination (MA) and indirect fluorescent antibody (IFA) staining methods, compare serologic test results, and to determine if these animals were exposed to this bacterium in two widely separated regions located in Connecticut and New York State. 2. Materials and methods 2.1. Sources of cat sera Veterinarians collected blood samples from 91 privately owned cats during the period 1985 through 1990. Serum samples were analyzed earlier for antibodies to Borrelia burgdorferi and Anaplasma phagocytophilum (Magnarelli et al., 1990, 2005). These samples had been frozen at 60 °C at the Connecticut Agricultural Experiment Station. Because of frequent outdoor activity, the subjects were originally included in a passive surveillance program for borreliosis in northeastern United States to further investigate the rising prevalence of tick-borne infections. Twelve cats showed clinical signs of fever, lameness, anorexia, or fatigue, while the remaining 79 cats appeared healthy at times of examination. Eighty-one sera represented cats living in 15 towns located in south central and southeastern Connecticut. The remaining 10 sera were from cats living in six towns in the lower Hudson Valley region of New York State. The Eastern cottontail rabbit, an important reservoir for F. tularensis, and American dog ticks are common in these towns. Broad areas of coastal northeastern United States are considered highly endemic for one or more tick-associated diseases. Details on times of sample collection, tick parasitism of cats, and clinical records have been reported (Magnarelli et al., 1990); this previous study also revealed no specific or potentially cross-reactive antibodies against Leptospira interrogans serovars canicola or icterohaemorrhagiae. However, it was necessary to further assess immunoassay specificities by testing cat sera with and without antibodies to F. tularensis. Nine sera containing antibodies to F. tularensis and three test sera with no detectable antibodies to this agent were screened in the present study, along with negative control sera, for antibodies to whole-cell B. burgdorferi and A. phagocytophilum by ELISA or IFA staining methods described earlier (Magnarelli et al., 2005). Three other cat sera containing antibodies to B. burgdorferi or A. phagocytophilum (titers P 1,280) were tested for F. tularensis antibodies by MA and IFA staining methods. 2.2. Microagglutination assay A commercial test (Difco Laboratories, Detroit, Michigan, USA) containing a phenolized-inactivated suspension of F. tularensis, was used to detect serum agglutinins. We relied on a rapid slide procedure for preliminary screening and followed the manufacturer’s directions, including the
23
use of a commercially prepared F. tularensis hyperimmune antiserum prepared in rabbits as a positive control. Additional human and rabbit positive controls were received from May Chu of the Centers for Disease Control and Prevention in Fort Collins, Colorado, USA and Susan Wong of the New York State Health Department. Antibody-positive cat sera were provided by Frederick Leighton (via Susan Wong) of the Canadian Cooperative Wildlife Center (Saskatoon, Saskatchewan, Canada). To determine conservative cut-off values for positive results, 10 negative control cat sera, which were determined to be seronegative for B. burgdorferi and A. phagocytophilum antibodies, were purchased (Equitech-Bio, Inc., Kerrville, Texas, USA; Jackson Immuno Research, West Grove, Pennslyvania, USA) and tested at different dilutions with antigen. Since there was no reactivity of negative control cat sera at dilutions of P 1:80, reactivity of test sera at or above this dilution was considered positive. Cat sera positive at a serum dilution of 1:80 were titrated and re-tested on a different day to determine antibody titers and to check reproducibility of results. 2.3. Indirect fluorescent antibody (IFA) staining methods Procedures described before to test cat sera for antibodies against B. burgdorferi and A. phagocytophilum (Magnarelli et al., 1990, 2005) were used in the present study. Formalin-treated, unstained F. tularensis (LVS) cells were provided by May Chu of the Centers for Disease Control and Prevention (Fort Collins, Colorado) and shipped under a federal permit to the Connecticut Agricultural Experiment Station. The non-viable antigen was mixed with 5% uninfected (normal) yolk sac, a step used in a microimmunofluorescence test to detect rickettsial antibodies (Philip et al., 1976). Yolk sac was used to enhance antigen adherence to glass microscope slides and facilitate later microscopic examinations of reactions. Slides were incubated overnight at 37 °C and fixed in cold acetone for 10 min. Following the application of diluted test sera to antigen and subsequent washing procedures, a polyvalent fluorescein isothiocyanate-conjugated goat anti-cat immunoglobulin (Kirkegaard and Perry Laboratories, Gaithersburg, Maryland, USA) was diluted in phosphate buffered saline solution (PBSS) to 1:20 and used as the second antibody. After slides were washed in PBSS and dried, cover slips were mounted with buffered glycerol, and preparations were examined by fluorescence microscopy. The same positive and negative cat sera included in MA tests were used to verify the reactivity of antigen and conjugate or to check for false positive reactions. In analyses of the 10 negative cat sera, there were no false positive reactions when sera were tested at dilutions of 1:80 or greater. Therefore, distinct fluorescence of antigen with test sera at or above this dilution indicated the presence of antibodies against F. tularensis. Titration endpoints were determined for all positive samples.
24
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
2.4. DNA analyses Thiry-seven cat sera, determined to be positive for antibodies to F. tularensis by either or both serologic methods (n = 24) or negative (n = 13) were screened for the DNA of F. tularensis at L2 Diagnostics laboratories following published procedures (Fulop et al., 1996). Sera from three ill cats, all positive by MA testing with antibody titers ranging from 1:160 to 1:640, were included in the DNA analyses. Blood components containing macrophages and other cells were unavailable for testing. Sera, which represented cats living in 12 towns in southern Connecticut, were extracted using the DNeasy kit (Qiagen, Inc., Valencia, California, USA) following the manufacturer’s directions for isolating DNA from whole non-nucleated blood. The analyses were based on a sequential, nested polymerase chain reaction (PCR) protocol for the FopA gene, which encodes an outer membrane protein that elicits a dominant antibody response. This gene was selected because antibodies to FopA are frequently present in human convalescent sera and appears to be well-conserved in the genus Francisella (Fulop et al., 1996). The primers used in the present study had nucleotide sequences from FopA that are unique to this genus (Fulop et al., 1996). The expected PCR product was amplified from positive control DNA but not from material extracted from the cat sera. One negative control serum was included in analyses. 2.5. Statistical analyses A z-test with the Yates correction (SigmaStat, SPSS, Chicago, Illinois, USA) was used to determine significant differences in proportions of positive results. Values of P < 0.05 were considered significant. 3. Results Cat sera contained antibodies to F. tularensis, but PCR analyses for DNA were negative. Of the 91 sera analyzed by MA methods, 11 (12%) were considered positive (Table 1). Testing by IFA methods revealed 22 (24%) positives. When the assay seropositivity rates for 57 sera were evaluated for 1989, the difference in proportions of positive sera were statistically significant (P = 0.033). Sera positive by either serologic method were obtained from cats in the following Connecticut towns: Branford (n = 2), Cheshire, Durham (n = 7), Guilford (n = 14), Hamden, Killingworth, Madison (n = 3), Middlefield, and Middletown. The towns of Branford, Madison, and Durham border Guilford in south central Connecticut. There were two additional seropositives from Pleasantville and Thornwood, New York. In analyses of the 12 sera from ill cats, one serum was positive by MA at a dilution of 1:160, while two other sera were positive by the IFA method at dilutions of 1:160 and 1:320. The remaining 9 sera from ill cats were negative in both tests. There were
Table 1 Serum antibodies to F. tularensis in cats, as detected by microagglutination (MA) and indirect fluorescent antibody (IFA) staining methods Sampling
Total sera tested
State
Years
Connecticut
1985 1986 1987 1988 1989 1990 1985 1987 Totals
New York
2 5 5 11 57 1 5 5 91
No. (%) positive
a
MA
IFA
0 0 1(20) 1(9) 7(12) 0 2(40) 0 11(12)
0 0 1(20) 4(36) 17(30) 0 0 0 22(24)
a
Serum antibody titers of 1:80 or greater were considered positive by MA and IFA staining methods.
no significant differences (P = 0.582) in seropositivity rates between the ill cats (25%) and the healthy cats (38%). Titers of serum antibodies against F. tularensis were greater when the IFA method was used, compared to those determined by performing the MA test. Frequency distributions of antibody titers for the former were 1:80 (n = 9), 1:160 (n = 8), 1:320 (n = 2), and 1:640 (n = 3), whereas those for the agglutination test were 1:80 (n = 8) and 1:160 (n = 3). The highest IFA titers were recorded for sera representing cats that lived in Branford, Durham, Guilford, and Killingworth. When results were compared for both antibody tests, there was good agreement (73% concordance). Four sera were positive in both assays, while 62 sera were negative. Results for the remaining 25 sera were discordant (18 seropositives by IFA only and seven seropositives by the MA test only). There was minimal evidence of cross reactivity or coinfections. Nine cat sera positive by IFA staining procedures for F. tularensis antibodies and three sera negative in these tests were screened for antibodies against wholecell B. burgdorferi by an ELISA and whole-cell A. phagocytophilum by IFA staining methods. Titration endpoints for F. tularensis antibodies ranged from 1:80 to 1:640 by IFA methods; five sera were negative in tests for B. burgdorferi and A. phagocytophilum antibodies. Three other sera with antibodies to F. tularensis (IFA titers = 1:160 or 1:640) reacted positively only to A. phagocytophilum antigen at titers of 1:80, 1:320, and 1:2,560. The remaining positive serum with F. tularensis antibody (titer = 1:320) also had antibodies to both B. burgdorferi and A. phagocytophilum at high titers (1:640). The three sera lacking antibodies to F. tularensis were negative in tests for B. burgdorferi and A. phagocytophilum antibodies. An additional 3 cat sera, with antibody titers of 1:1,280 or greater to whole-cell B. burgdorferi or A. phagocytophilum, were negative by IFA and MA test methods for F. tularensis antibodies.
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
4. Discussion Serologic analyses revealed that cats living in Connecticut and New York State were exposed to F. tularensis or a closely related organism. It is unknown, however, when these animals were infected or whether they were bitten by infected ticks or had eaten infected mice, rabbits, or other mammals. It is likely that F. tularensis subspecies tularensis and, possibly, F. tularensis subspecies holarctica are present along with F. philomiragia in Connecticut and other areas of northeastern United States. Human isolates of the latter pathogen are reported for Connecticut (Versage et al., 2003). Based on genetic analyses or culturing Francisella pathogens from humans and other mammals (Farlow et al., 2005; Hayes, 2005; Versage et al., 2003; Dennis et al., 2001), there is convincing evidence of widespread distribution of different subspecies in the United States. Moreover, the presence of endosymbionts, related to or belonging to the genus Francisella (Niebylski et al., 1997; Sun et al., 2000; Goethert and Telford, 2005), are broadly distributed in ticks. However, the veterinary and public health significance of these microbes is unknown. There is no evidence for transmission of endosymbionts to vertebrate hosts. The paucity of reported human cases of tularemia and lack of documented F. tularensis infections in domesticated animals in Connecticut and other states is probably due, in part, to unawareness of the disease, underdiagnosis, or asymptomatic infections. Therefore, application of a newly developed multitarget real-time TaqMan PCR assay (Versage et al., 2003), which reportedly can differentially diagnose F. tularensis and F. philomiragia infections, may help clarify the prevalence of these pathogens. Unlike F. tularensis, F. philomiragia is not known to be transmitted by arthropods. Some human infections of the latter pathogen were associated with neardrowning events (Versage et al., 2003; Wenger et al., 1989), which may indicate that water exposure is a risk factor. Since most cats in our study were healthy at the times blood samples were obtained and there was no DNA of F. tularensis detected, we conclude that the antibodies present were residual from prior exposure and probably do not reflect active infections. Serologic evidence of exposure may not be tightly linked to infection. If ill cats live in areas where tularemia occurs, blood samples containing macrophages and other cells should be included in analyses for DNA because F. tularensis appears to be an obligate intracellular pathogen. A variety of serologic tests are available for laboratory diagnosis of tularemia (Hayes, 2005; Dennis et al., 2001). The tube agglutination or microagglutination tests (Sato et al., 1990) have been used widely for preliminary screening of sera for F. tularensis antibodies, and fluorescentlabeled antibodies are useful in performing another rapid diagnostic procedure for detecting F. tularensis in mammalian tissues or bodily fluids. In our studies, conservative grading was used in MA and IFA tests to determine positive antibody results. Both antibody tests were suitable for
25
general screening of sera, even though differences in assay sensitivity were apparent. Our results on agglutination antibody titers were comparable to those previously reported for cats infected with F. tularensis (Woods et al., 1998; Rohrback, 1988). In our experience, however, judging positive antibody reactions was more easily accomplished when microscopically examining fluorescein-stained preparations. Tests on specificity were conducted to assess possible cross-reactivity between antibodies and antigens of F. tularensis, B. burgdorferi, and A. phagocytophilum and to detect co-infections. Although the latter 2 organisms have widespread occurrence in ticks and mammals in northeastern United States, they are unrelated to F. tularensis. In seroanalyses for tularemia, cross-reactivity can occur in tests with Brucella, Legionella, Yersinia, and Mycoplasma organisms (Hayes, 2005), but these microbes are not known to infect cats in Connecticut. Moreover, studies conducted in the Austrian and Slovakian borderland (Vyrostekova et al., 2002) revealed co-infections of B. burgdorferi sensu lato (i.e., Borrelia garinii) and F. tularensis in abundant rodent species; simultaneous infections were reported for 2 species of voles during an epizootic of tularemia. Based on our results and published information, we conclude that there is minimal serologic cross-reactivity between F. tularensis, B. burgdorferi, and A. phagocytophilum and that the presence of coexisting antibodies to these agents in cats, albeit infrequent, probably represents distinct infections. In view of the extensive distributions of F. tularensis subspecies and related organisms in the environment, multiple modes of pathogen transmission, and current concerns about potential use of F. tularensis as a biological weapon (Farlow et al., 2005; Hayes, 2005; Dennis et al., 2001; Sato et al., 1990), further field and clinical investigations of F. tularensis and related species are warranted. Our results support those of others (Baldwin et al., 1991; Woods et al., 1998; Rohrback, 1988), which show that cats may be at risk for naturally acquiring F. tularensis because of their free-roaming behavior and contact with rodents and ticks. In widespread outbreaks of tularemia in Sweden (Eliasson et al., 2002), human cases were correlated with mosquito bites as the main risk factor, but contact with cats was also judged to be important. Therefore, tularemia should be considered in the differential diagnosis of unexplained febrile illnesses in cats living in tick-infested habitats where this disease is endemic. Moreover, detection of Francisella infections in cats also may help improve diagnosis and treatment for these animals, determine the epidemiological significance of pathogens in selected sites or regions, and provide geographic baseline estimates of tularemia infections in nature. Acknowledgements We are grateful to Tia Mastrone and Deborah Beck for technical assistance. This work was supported, in part, by
26
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
federal Hatch funds administered by the United States Department of Agriculture and grants from the United States Centers for Disease Control and Prevention (CCU111188-01 and HR8/CCH113382-01), Emerging Infections Program. References Baldwin, C.J., Panciera, R.J., Morton, R.J., Cowell, A.K., Waurzyniak, B.J., 1991. Acute tularemia in three domestic cats. Journal of the American Veterinary Medical Association 199, 1602–1605. Dennis, D.T., Inglesby, T.V., Henderson, D.A., Bartlett, J.G., Ascher, M.S., Eitzen, E., Fine, A.D., Friedlander, A.M., Hauer, J., Layton, M., Lillibridge, S.R., McDade, J.E., Osterholm, M.T., O’Toole, T., Parker, G., Perl, T.M., Russell, P.K., Tonat, K., 2001. Tularemia as a biological weapon–medical and public health management. Journal of the American Medical Association 285, 2763–2773. Eliasson, H., Lindback, J., Nuorti, J.P., Arneborn, M., Giesecke, J., Tegnell, A., 2002. The 2000 tularemia outbreak: a case-control study of risk factors in disease-endemic and emergent areas, Sweden. Emerging Infectious Diseases 8, 956–960. Farlow, J., Wagner, D.M., Dukerich, M., Stanley, M., Chu, M., Kubota, K., Petersen, J., Keim, P., 2005. Francisella tularensis in the United States. Emerging Infectious Diseases 11, 1835–1841. Feldman, K.A., Stiles-Enos, D., Julian, K., Matyas, B.T., Telford III, S.R., Chu, M.C., Petersen, L.R., Hayes, E.B., 2003. Tularemia on Martha’s Vineyard: seroprevalence and occupational risk. Emerging Infectious Diseases 9, 350–354. Fulop, M., Leslie, D., Titball, R., 1996. A rapid, highly sensitive method for the detection of Francisella tularensis in clinical samples using the polymerase chain reaction. American Journal of Tropical Medicine and Hygiene 54, 364–366. Goethert, H.K., Telford III, S.R., 2005. A new Francisella (Beggiatiales: Francisellaceae) inquiline within Dermacentor variabilis say (Acari:Ixodidae). Journal of Medical Entomology 42, 502–505. Hayes, E.B., 2005. Tularemia. In: Goodman, J.L., Dennis, D.T., Sonenshine, D.E. (Eds.), Tick-borne Diseases of Humans. ASM Press, Washington DC, pp. 07–217.
Hopla, C.E., Hopla, A.K., 1994. Tularemia. In: Beran, G.W., Steele, J. (Eds.), CRC Handbook of Zoonoses, Section A, second ed. CRC Press, Boca Raton, pp. 113–126. Magnarelli, L.A., Anderson, J.F., Levine, H.R., Levy, S.A., 1990. Tick parasitism and antibodies to Borrelia burgdorferi in cats. Journal of the American Veterinary Medical Association 197, 63–66. Magnarelli, L.A., Bushmich, S.L., IJdo, J.W., Fikrig, E., 2005. Seroprevalence of antibodies against Borrelia burgdorferi and Anaplasma phagocytophilum in cats. American Journal of Veterinary Research 66, 1895–1899. Niebylski, M., Peacock, M.G., Fischer, E., Porcella, S.F., Schwan, T.G., 1997. Characterization of an endosymbiont infecting wood ticks,Dermacentor andersoni, as a member of the genus Francisella. Applied Environmental Microbiology 63, 3933–3940. Philip, R.N., Casper, E.A., Ormsbee, R.A., Peacock, M.G., Burgdorfer, W., 1976. Microimmunofluorescence test for the serological study of rocky mountain spotted fever and typhus. Journal of Clinical Microbiology 3, 51–61. Rohrback, B.W., 1988. Tularemia. Journal of the American Veterinary Medical Association 193, 428–432. Sato, T., Fujita, H., Ohara, Y., Homma, M., 1990. Microagglutination test for early and specific serodiagnosis of tularemia. Journal of Clinical Microbiology 28, 2372–2374. Sun, L.V., Scoles, G.A., Fish, D., O’Neill, S.L., 2000. Francisella-like endosymbionts of ticks. Journal of Invertebrate Pathology 76, 301– 303. Versage, J.L., Severin, D.D.M., Chu, M.C., Petersen, J.M., 2003. Development of a multitarget real-time TaqMan PCR assay for enhanced detection of Francisella tularensis in complex specimens. Journal of Clinical Microbiology 41, 5492–5499. Vyrostekova, V., Khanakah, G., Kocianova, E., Gurycova, D., Stanek, G., 2002. Prevalence of coinfection with Francisella tularensis in reservoir animals of Borrelia burgdorferi sensu lato. Wien Klin Wochenschr 114, 482–488. Wenger, J.D., Hollis, D.G., Weaver, R.E., Baker, C.N., Brown, G.R., Brenner, D.J., Broome, C.V., 1989. Infection caused by Francisella philomiragia (formerly Yersinia philomiragia): a newly recognized pathogen. Annals of Internal Medicine 110, 888–892. Woods, J.P., Crystal, M.A., Morton, R.J., Panciera, R.J., 1998. Tularemia in two cats. Journal of the American Medical Association 212, 81–83.
Detection of antibodies to Francisella tularensis in cats L. Magnarelli a
a,*
, S. Levy b, R. Koski
c
The Connecticut Agricultural Experiment Station, P.O. Box 1106, New Haven, CT 06504, USA b Durham Veterinary Hospital, 178 Parmelee Hill Road, Durham, CT 06422, USA c 2 L Diagnostics, LLC, 300 George Street, New Haven, CT 06511, USA Accepted 20 June 2006
Abstract Blood samples were obtained from privately owned cats in Connecticut and New York State, USA in 1985–1990, and analyzed for evidence of Francisella tularensis, the etiologic agent of tularemia. Of the 91 sera tested by microagglutination (MA) methods, 11 (12%) contained antibodies to F. tularensis. Analyses of the same sera by indirect fluorescent antibody (IFA) staining methods revealed 22 (24%) positives. There was good agreement in results of both tests (73% concordance). However, we measured higher titers (1:80 to 1:640) with IFA analysis than by MA methods (1:80 to 1:160). Both tests were suitable for general screening purposes. The DNA of F.tularensis was not detected in the 24 antibody-positive sera tested. Cats living in Connecticut and New York State were naturally exposed to F.tularensis or a closely related organism. With exposure to ticks, other biting arthropods, mice, and rabbits, cats are at risk for acquiring F.tularensis infections and can be an important source of information on the presence of this agent in nature. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Cats; Tularemia; Francisella tularensis; Antibodies; Ticks
1. Introduction Tularemia is caused by Francisella tularensis, a highly pathogenic, non-spore-forming coccobacillus that can infect several species of vertebrate hosts. Infection has been documented in humans as well as in Eastern cottontail rabbits (Sylvilagus floridanus), mice, and domesticated mammals, including cats, throughout North America, Eurasia, parts of the Middle East, and the north coast of Africa (Hayes, 2005). The American dog tick (Dermacentor variabilis) is an important vector in the United States, but other Dermacentor, as well as Amblyomma, Haemaphysalis, and Ixodes ticks, maintain enzootic infections in nature (Hopla and Hopla, 1994). Ixodes pacificus and Ixodes scapularis ticks, known to transmit bacterial or protozoan agents that cause Lyme borreliosis, granulocytic anaplasmosis, or human babesiosis, parasitize a wide range of vertebrate hosts and are suspected vectors of F. tularensis (Hayes, *
Corresponding author. Tel.: +1 203 974 8440; fax: +1 203 974 8502. E-mail address: [email protected] (L. Magnarelli).
0034-5288/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.rvsc.2006.06.003
2005), but confirmatory evidence of transmission is lacking. F. tularensis infections can also be acquired from biting insects, by direct contact with blood or other tissues or fluids of infected mammals, by ingestion of contaminated food or water, or by inhaling the pathogen. Although human cases of tularemia are seldom reported in Connecticut and other areas of northeastern United States, repeated outbreaks have occurred recently on Martha’s Vineyard, an island off the coast of Massachusetts (Feldman et al., 2003) and on Long Island, New York (Hayes, 2005). Case reports from Oklahoma (Baldwin et al., 1991; Woods et al., 1998) and New Mexico (Rohrback, 1988) describe tularemia infections in cats. Severity of disease can range from mild to fatal with associated non-specific signs, such as pyrexia, anorexia, lymphadenopathy, or oral ulcers. When allowed outdoors, cats are free roaming and can have extensive exposure to ticks and contact with rodents. Cats, therefore, have great potential for acquiring different pathogens. The main purposes of the present study were to analyze cat sera for antibodies to F. tularensis
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
by means of microagglutination (MA) and indirect fluorescent antibody (IFA) staining methods, compare serologic test results, and to determine if these animals were exposed to this bacterium in two widely separated regions located in Connecticut and New York State. 2. Materials and methods 2.1. Sources of cat sera Veterinarians collected blood samples from 91 privately owned cats during the period 1985 through 1990. Serum samples were analyzed earlier for antibodies to Borrelia burgdorferi and Anaplasma phagocytophilum (Magnarelli et al., 1990, 2005). These samples had been frozen at 60 °C at the Connecticut Agricultural Experiment Station. Because of frequent outdoor activity, the subjects were originally included in a passive surveillance program for borreliosis in northeastern United States to further investigate the rising prevalence of tick-borne infections. Twelve cats showed clinical signs of fever, lameness, anorexia, or fatigue, while the remaining 79 cats appeared healthy at times of examination. Eighty-one sera represented cats living in 15 towns located in south central and southeastern Connecticut. The remaining 10 sera were from cats living in six towns in the lower Hudson Valley region of New York State. The Eastern cottontail rabbit, an important reservoir for F. tularensis, and American dog ticks are common in these towns. Broad areas of coastal northeastern United States are considered highly endemic for one or more tick-associated diseases. Details on times of sample collection, tick parasitism of cats, and clinical records have been reported (Magnarelli et al., 1990); this previous study also revealed no specific or potentially cross-reactive antibodies against Leptospira interrogans serovars canicola or icterohaemorrhagiae. However, it was necessary to further assess immunoassay specificities by testing cat sera with and without antibodies to F. tularensis. Nine sera containing antibodies to F. tularensis and three test sera with no detectable antibodies to this agent were screened in the present study, along with negative control sera, for antibodies to whole-cell B. burgdorferi and A. phagocytophilum by ELISA or IFA staining methods described earlier (Magnarelli et al., 2005). Three other cat sera containing antibodies to B. burgdorferi or A. phagocytophilum (titers P 1,280) were tested for F. tularensis antibodies by MA and IFA staining methods. 2.2. Microagglutination assay A commercial test (Difco Laboratories, Detroit, Michigan, USA) containing a phenolized-inactivated suspension of F. tularensis, was used to detect serum agglutinins. We relied on a rapid slide procedure for preliminary screening and followed the manufacturer’s directions, including the
23
use of a commercially prepared F. tularensis hyperimmune antiserum prepared in rabbits as a positive control. Additional human and rabbit positive controls were received from May Chu of the Centers for Disease Control and Prevention in Fort Collins, Colorado, USA and Susan Wong of the New York State Health Department. Antibody-positive cat sera were provided by Frederick Leighton (via Susan Wong) of the Canadian Cooperative Wildlife Center (Saskatoon, Saskatchewan, Canada). To determine conservative cut-off values for positive results, 10 negative control cat sera, which were determined to be seronegative for B. burgdorferi and A. phagocytophilum antibodies, were purchased (Equitech-Bio, Inc., Kerrville, Texas, USA; Jackson Immuno Research, West Grove, Pennslyvania, USA) and tested at different dilutions with antigen. Since there was no reactivity of negative control cat sera at dilutions of P 1:80, reactivity of test sera at or above this dilution was considered positive. Cat sera positive at a serum dilution of 1:80 were titrated and re-tested on a different day to determine antibody titers and to check reproducibility of results. 2.3. Indirect fluorescent antibody (IFA) staining methods Procedures described before to test cat sera for antibodies against B. burgdorferi and A. phagocytophilum (Magnarelli et al., 1990, 2005) were used in the present study. Formalin-treated, unstained F. tularensis (LVS) cells were provided by May Chu of the Centers for Disease Control and Prevention (Fort Collins, Colorado) and shipped under a federal permit to the Connecticut Agricultural Experiment Station. The non-viable antigen was mixed with 5% uninfected (normal) yolk sac, a step used in a microimmunofluorescence test to detect rickettsial antibodies (Philip et al., 1976). Yolk sac was used to enhance antigen adherence to glass microscope slides and facilitate later microscopic examinations of reactions. Slides were incubated overnight at 37 °C and fixed in cold acetone for 10 min. Following the application of diluted test sera to antigen and subsequent washing procedures, a polyvalent fluorescein isothiocyanate-conjugated goat anti-cat immunoglobulin (Kirkegaard and Perry Laboratories, Gaithersburg, Maryland, USA) was diluted in phosphate buffered saline solution (PBSS) to 1:20 and used as the second antibody. After slides were washed in PBSS and dried, cover slips were mounted with buffered glycerol, and preparations were examined by fluorescence microscopy. The same positive and negative cat sera included in MA tests were used to verify the reactivity of antigen and conjugate or to check for false positive reactions. In analyses of the 10 negative cat sera, there were no false positive reactions when sera were tested at dilutions of 1:80 or greater. Therefore, distinct fluorescence of antigen with test sera at or above this dilution indicated the presence of antibodies against F. tularensis. Titration endpoints were determined for all positive samples.
24
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
2.4. DNA analyses Thiry-seven cat sera, determined to be positive for antibodies to F. tularensis by either or both serologic methods (n = 24) or negative (n = 13) were screened for the DNA of F. tularensis at L2 Diagnostics laboratories following published procedures (Fulop et al., 1996). Sera from three ill cats, all positive by MA testing with antibody titers ranging from 1:160 to 1:640, were included in the DNA analyses. Blood components containing macrophages and other cells were unavailable for testing. Sera, which represented cats living in 12 towns in southern Connecticut, were extracted using the DNeasy kit (Qiagen, Inc., Valencia, California, USA) following the manufacturer’s directions for isolating DNA from whole non-nucleated blood. The analyses were based on a sequential, nested polymerase chain reaction (PCR) protocol for the FopA gene, which encodes an outer membrane protein that elicits a dominant antibody response. This gene was selected because antibodies to FopA are frequently present in human convalescent sera and appears to be well-conserved in the genus Francisella (Fulop et al., 1996). The primers used in the present study had nucleotide sequences from FopA that are unique to this genus (Fulop et al., 1996). The expected PCR product was amplified from positive control DNA but not from material extracted from the cat sera. One negative control serum was included in analyses. 2.5. Statistical analyses A z-test with the Yates correction (SigmaStat, SPSS, Chicago, Illinois, USA) was used to determine significant differences in proportions of positive results. Values of P < 0.05 were considered significant. 3. Results Cat sera contained antibodies to F. tularensis, but PCR analyses for DNA were negative. Of the 91 sera analyzed by MA methods, 11 (12%) were considered positive (Table 1). Testing by IFA methods revealed 22 (24%) positives. When the assay seropositivity rates for 57 sera were evaluated for 1989, the difference in proportions of positive sera were statistically significant (P = 0.033). Sera positive by either serologic method were obtained from cats in the following Connecticut towns: Branford (n = 2), Cheshire, Durham (n = 7), Guilford (n = 14), Hamden, Killingworth, Madison (n = 3), Middlefield, and Middletown. The towns of Branford, Madison, and Durham border Guilford in south central Connecticut. There were two additional seropositives from Pleasantville and Thornwood, New York. In analyses of the 12 sera from ill cats, one serum was positive by MA at a dilution of 1:160, while two other sera were positive by the IFA method at dilutions of 1:160 and 1:320. The remaining 9 sera from ill cats were negative in both tests. There were
Table 1 Serum antibodies to F. tularensis in cats, as detected by microagglutination (MA) and indirect fluorescent antibody (IFA) staining methods Sampling
Total sera tested
State
Years
Connecticut
1985 1986 1987 1988 1989 1990 1985 1987 Totals
New York
2 5 5 11 57 1 5 5 91
No. (%) positive
a
MA
IFA
0 0 1(20) 1(9) 7(12) 0 2(40) 0 11(12)
0 0 1(20) 4(36) 17(30) 0 0 0 22(24)
a
Serum antibody titers of 1:80 or greater were considered positive by MA and IFA staining methods.
no significant differences (P = 0.582) in seropositivity rates between the ill cats (25%) and the healthy cats (38%). Titers of serum antibodies against F. tularensis were greater when the IFA method was used, compared to those determined by performing the MA test. Frequency distributions of antibody titers for the former were 1:80 (n = 9), 1:160 (n = 8), 1:320 (n = 2), and 1:640 (n = 3), whereas those for the agglutination test were 1:80 (n = 8) and 1:160 (n = 3). The highest IFA titers were recorded for sera representing cats that lived in Branford, Durham, Guilford, and Killingworth. When results were compared for both antibody tests, there was good agreement (73% concordance). Four sera were positive in both assays, while 62 sera were negative. Results for the remaining 25 sera were discordant (18 seropositives by IFA only and seven seropositives by the MA test only). There was minimal evidence of cross reactivity or coinfections. Nine cat sera positive by IFA staining procedures for F. tularensis antibodies and three sera negative in these tests were screened for antibodies against wholecell B. burgdorferi by an ELISA and whole-cell A. phagocytophilum by IFA staining methods. Titration endpoints for F. tularensis antibodies ranged from 1:80 to 1:640 by IFA methods; five sera were negative in tests for B. burgdorferi and A. phagocytophilum antibodies. Three other sera with antibodies to F. tularensis (IFA titers = 1:160 or 1:640) reacted positively only to A. phagocytophilum antigen at titers of 1:80, 1:320, and 1:2,560. The remaining positive serum with F. tularensis antibody (titer = 1:320) also had antibodies to both B. burgdorferi and A. phagocytophilum at high titers (1:640). The three sera lacking antibodies to F. tularensis were negative in tests for B. burgdorferi and A. phagocytophilum antibodies. An additional 3 cat sera, with antibody titers of 1:1,280 or greater to whole-cell B. burgdorferi or A. phagocytophilum, were negative by IFA and MA test methods for F. tularensis antibodies.
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
4. Discussion Serologic analyses revealed that cats living in Connecticut and New York State were exposed to F. tularensis or a closely related organism. It is unknown, however, when these animals were infected or whether they were bitten by infected ticks or had eaten infected mice, rabbits, or other mammals. It is likely that F. tularensis subspecies tularensis and, possibly, F. tularensis subspecies holarctica are present along with F. philomiragia in Connecticut and other areas of northeastern United States. Human isolates of the latter pathogen are reported for Connecticut (Versage et al., 2003). Based on genetic analyses or culturing Francisella pathogens from humans and other mammals (Farlow et al., 2005; Hayes, 2005; Versage et al., 2003; Dennis et al., 2001), there is convincing evidence of widespread distribution of different subspecies in the United States. Moreover, the presence of endosymbionts, related to or belonging to the genus Francisella (Niebylski et al., 1997; Sun et al., 2000; Goethert and Telford, 2005), are broadly distributed in ticks. However, the veterinary and public health significance of these microbes is unknown. There is no evidence for transmission of endosymbionts to vertebrate hosts. The paucity of reported human cases of tularemia and lack of documented F. tularensis infections in domesticated animals in Connecticut and other states is probably due, in part, to unawareness of the disease, underdiagnosis, or asymptomatic infections. Therefore, application of a newly developed multitarget real-time TaqMan PCR assay (Versage et al., 2003), which reportedly can differentially diagnose F. tularensis and F. philomiragia infections, may help clarify the prevalence of these pathogens. Unlike F. tularensis, F. philomiragia is not known to be transmitted by arthropods. Some human infections of the latter pathogen were associated with neardrowning events (Versage et al., 2003; Wenger et al., 1989), which may indicate that water exposure is a risk factor. Since most cats in our study were healthy at the times blood samples were obtained and there was no DNA of F. tularensis detected, we conclude that the antibodies present were residual from prior exposure and probably do not reflect active infections. Serologic evidence of exposure may not be tightly linked to infection. If ill cats live in areas where tularemia occurs, blood samples containing macrophages and other cells should be included in analyses for DNA because F. tularensis appears to be an obligate intracellular pathogen. A variety of serologic tests are available for laboratory diagnosis of tularemia (Hayes, 2005; Dennis et al., 2001). The tube agglutination or microagglutination tests (Sato et al., 1990) have been used widely for preliminary screening of sera for F. tularensis antibodies, and fluorescentlabeled antibodies are useful in performing another rapid diagnostic procedure for detecting F. tularensis in mammalian tissues or bodily fluids. In our studies, conservative grading was used in MA and IFA tests to determine positive antibody results. Both antibody tests were suitable for
25
general screening of sera, even though differences in assay sensitivity were apparent. Our results on agglutination antibody titers were comparable to those previously reported for cats infected with F. tularensis (Woods et al., 1998; Rohrback, 1988). In our experience, however, judging positive antibody reactions was more easily accomplished when microscopically examining fluorescein-stained preparations. Tests on specificity were conducted to assess possible cross-reactivity between antibodies and antigens of F. tularensis, B. burgdorferi, and A. phagocytophilum and to detect co-infections. Although the latter 2 organisms have widespread occurrence in ticks and mammals in northeastern United States, they are unrelated to F. tularensis. In seroanalyses for tularemia, cross-reactivity can occur in tests with Brucella, Legionella, Yersinia, and Mycoplasma organisms (Hayes, 2005), but these microbes are not known to infect cats in Connecticut. Moreover, studies conducted in the Austrian and Slovakian borderland (Vyrostekova et al., 2002) revealed co-infections of B. burgdorferi sensu lato (i.e., Borrelia garinii) and F. tularensis in abundant rodent species; simultaneous infections were reported for 2 species of voles during an epizootic of tularemia. Based on our results and published information, we conclude that there is minimal serologic cross-reactivity between F. tularensis, B. burgdorferi, and A. phagocytophilum and that the presence of coexisting antibodies to these agents in cats, albeit infrequent, probably represents distinct infections. In view of the extensive distributions of F. tularensis subspecies and related organisms in the environment, multiple modes of pathogen transmission, and current concerns about potential use of F. tularensis as a biological weapon (Farlow et al., 2005; Hayes, 2005; Dennis et al., 2001; Sato et al., 1990), further field and clinical investigations of F. tularensis and related species are warranted. Our results support those of others (Baldwin et al., 1991; Woods et al., 1998; Rohrback, 1988), which show that cats may be at risk for naturally acquiring F. tularensis because of their free-roaming behavior and contact with rodents and ticks. In widespread outbreaks of tularemia in Sweden (Eliasson et al., 2002), human cases were correlated with mosquito bites as the main risk factor, but contact with cats was also judged to be important. Therefore, tularemia should be considered in the differential diagnosis of unexplained febrile illnesses in cats living in tick-infested habitats where this disease is endemic. Moreover, detection of Francisella infections in cats also may help improve diagnosis and treatment for these animals, determine the epidemiological significance of pathogens in selected sites or regions, and provide geographic baseline estimates of tularemia infections in nature. Acknowledgements We are grateful to Tia Mastrone and Deborah Beck for technical assistance. This work was supported, in part, by
26
L. Magnarelli et al. / Research in Veterinary Science 82 (2007) 22–26
federal Hatch funds administered by the United States Department of Agriculture and grants from the United States Centers for Disease Control and Prevention (CCU111188-01 and HR8/CCH113382-01), Emerging Infections Program. References Baldwin, C.J., Panciera, R.J., Morton, R.J., Cowell, A.K., Waurzyniak, B.J., 1991. Acute tularemia in three domestic cats. Journal of the American Veterinary Medical Association 199, 1602–1605. Dennis, D.T., Inglesby, T.V., Henderson, D.A., Bartlett, J.G., Ascher, M.S., Eitzen, E., Fine, A.D., Friedlander, A.M., Hauer, J., Layton, M., Lillibridge, S.R., McDade, J.E., Osterholm, M.T., O’Toole, T., Parker, G., Perl, T.M., Russell, P.K., Tonat, K., 2001. Tularemia as a biological weapon–medical and public health management. Journal of the American Medical Association 285, 2763–2773. Eliasson, H., Lindback, J., Nuorti, J.P., Arneborn, M., Giesecke, J., Tegnell, A., 2002. The 2000 tularemia outbreak: a case-control study of risk factors in disease-endemic and emergent areas, Sweden. Emerging Infectious Diseases 8, 956–960. Farlow, J., Wagner, D.M., Dukerich, M., Stanley, M., Chu, M., Kubota, K., Petersen, J., Keim, P., 2005. Francisella tularensis in the United States. Emerging Infectious Diseases 11, 1835–1841. Feldman, K.A., Stiles-Enos, D., Julian, K., Matyas, B.T., Telford III, S.R., Chu, M.C., Petersen, L.R., Hayes, E.B., 2003. Tularemia on Martha’s Vineyard: seroprevalence and occupational risk. Emerging Infectious Diseases 9, 350–354. Fulop, M., Leslie, D., Titball, R., 1996. A rapid, highly sensitive method for the detection of Francisella tularensis in clinical samples using the polymerase chain reaction. American Journal of Tropical Medicine and Hygiene 54, 364–366. Goethert, H.K., Telford III, S.R., 2005. A new Francisella (Beggiatiales: Francisellaceae) inquiline within Dermacentor variabilis say (Acari:Ixodidae). Journal of Medical Entomology 42, 502–505. Hayes, E.B., 2005. Tularemia. In: Goodman, J.L., Dennis, D.T., Sonenshine, D.E. (Eds.), Tick-borne Diseases of Humans. ASM Press, Washington DC, pp. 07–217.
Hopla, C.E., Hopla, A.K., 1994. Tularemia. In: Beran, G.W., Steele, J. (Eds.), CRC Handbook of Zoonoses, Section A, second ed. CRC Press, Boca Raton, pp. 113–126. Magnarelli, L.A., Anderson, J.F., Levine, H.R., Levy, S.A., 1990. Tick parasitism and antibodies to Borrelia burgdorferi in cats. Journal of the American Veterinary Medical Association 197, 63–66. Magnarelli, L.A., Bushmich, S.L., IJdo, J.W., Fikrig, E., 2005. Seroprevalence of antibodies against Borrelia burgdorferi and Anaplasma phagocytophilum in cats. American Journal of Veterinary Research 66, 1895–1899. Niebylski, M., Peacock, M.G., Fischer, E., Porcella, S.F., Schwan, T.G., 1997. Characterization of an endosymbiont infecting wood ticks,Dermacentor andersoni, as a member of the genus Francisella. Applied Environmental Microbiology 63, 3933–3940. Philip, R.N., Casper, E.A., Ormsbee, R.A., Peacock, M.G., Burgdorfer, W., 1976. Microimmunofluorescence test for the serological study of rocky mountain spotted fever and typhus. Journal of Clinical Microbiology 3, 51–61. Rohrback, B.W., 1988. Tularemia. Journal of the American Veterinary Medical Association 193, 428–432. Sato, T., Fujita, H., Ohara, Y., Homma, M., 1990. Microagglutination test for early and specific serodiagnosis of tularemia. Journal of Clinical Microbiology 28, 2372–2374. Sun, L.V., Scoles, G.A., Fish, D., O’Neill, S.L., 2000. Francisella-like endosymbionts of ticks. Journal of Invertebrate Pathology 76, 301– 303. Versage, J.L., Severin, D.D.M., Chu, M.C., Petersen, J.M., 2003. Development of a multitarget real-time TaqMan PCR assay for enhanced detection of Francisella tularensis in complex specimens. Journal of Clinical Microbiology 41, 5492–5499. Vyrostekova, V., Khanakah, G., Kocianova, E., Gurycova, D., Stanek, G., 2002. Prevalence of coinfection with Francisella tularensis in reservoir animals of Borrelia burgdorferi sensu lato. Wien Klin Wochenschr 114, 482–488. Wenger, J.D., Hollis, D.G., Weaver, R.E., Baker, C.N., Brown, G.R., Brenner, D.J., Broome, C.V., 1989. Infection caused by Francisella philomiragia (formerly Yersinia philomiragia): a newly recognized pathogen. Annals of Internal Medicine 110, 888–892. Woods, J.P., Crystal, M.A., Morton, R.J., Panciera, R.J., 1998. Tularemia in two cats. Journal of the American Medical Association 212, 81–83.