Natural Infections with Borrelia Spirochetes in Two Dogs from Florida

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, Feb. 1994, p /94/$ Copyright 1994, American Society for Microbiology Vol. 32, No. 2 Natural Infections with Borrelia Spirochetes in Two Dogs from Florida EDWARD B. BREITSCHWERDT,l* WILLIAM L. NICHOLSON,1t ANITA R. KIEHL,2 CHRIS STEERS,3 DONALD J. MEUTEN,4 AND JAY F. LEVINE4 Departments of Companion Animal and Special Species Medicine' and Microbiology, Pathology, and Parasitologyj College of Veterinary Medicine, North Carolina State University, Raleigh, North Carolina 27606; Doctors and Physicians Laboratory, Leesburg, Florida ; and the Suburban Animal Hospital, Gainesville, Florida Received 1 June 1993/Returned for modification 12 August 1993/Accepted 4 November 1993 Spirochetemia is a rarely reported observation in dogs. We describe the clinical, hematologic, and immunodiagnostic features of two spirochetemic dogs from northern Florida and the subsequent isolation and preliminary characterization of a Borrelia species from one dog in which culture of a sample for spirochetes was attempted. Results of polyacrylamide gel electrophoresis, monoclonal antibody testing, and PCR analysis indicate that the Florida isolate is not Borrelia burgdorferi, the only other member of the genus that has been isolated in Florida. Our findings also indicate that a member of the genus Borrelia potentially causes disease in dogs in Florida and that serologic cross-reactivity of the Florida canine Borrelia isolate with B. burgdorfieri probably contributes to the inaccurate diagnosis of canine Lyme disease in the region. Arthropod-borne spirochetes of the genus Borrelia cause a wide range of clinical manifestations in both humans and animals (4, 7, 13). Human diseases are classified as illnesses caused by either relapsing fever borreliae or the Lyme disease agent, Borrelia burgdorfeni. Epidemic or louse-borne relapsing fever is caused by Borrelia recurrentis, whereas endemic relapsing fever is associated with at least 15 Borrelia species that are transmitted by soft ticks of the genus Ornithodoros (7, 13). Numerous vertebrates, including chipmunks, squirrels, rabbits, rats, mice, owls, and lizards, may serve as natural reservoirs for relapsing fever borreliae. In the United States, Borrelia hernsii (4), Borrelia turicatae (4), Borrelia parkeri (4), Borrelia coriaceae (11), Borrelia anserina (15), and Borrelia burgdorferi (1) are known to exist. In the western and southwestern United States, endemic tick-borne relapsing fever is caused by B. hermsii, B. turicatae, B. parkeri, and, because of the distribution of the vector tick, possibly B. mazzottii (7). B. coriaceae has been isolated from Ornithodoros conaceus, which primarily feeds on cattle and deer in the western United States (7, 11). B. ansenna, the cause of avian spirochetosis, is associated with marked morbidity and mortality in domestic poultry and has a worldwide distribution (4, 7). In humans, B. burgdorferi is associated with arthritis, myocarditis, and neurological problems (13). The clinical and pathogenic features of naturally occurring B. burgdorfen infections in dogs are still being clarified (16, 17); however, experimental inoculation of dogs with B. burgdorferi results in fever, lethargy, inappetence, and polyarthritis (3, 21). Spirochetemia has rarely been reported in dogs. In 1979, Schalm (20) described spirochetemia in a 6-year-old spayed female Blue Tic Hound from Bakersfield, Calif., and in 1990, * Corresponding author. Mailing address: Department of Companion Animal and Special Species Medicine, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough Street, Raleigh, NC Phone: 919/ Fax: 919/ t Present address: Department of Immunology and Infectious Diseases, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD Moreland and colleagues (18) described spirochetemia in a 2-year-old male Siberian husky from Texas. Illness in the Blue Tic Hound was characterized by lethargy, fever, lameness, leukocytosis, spirochetemia, and an uneventful recovery following unspecified antibiotic therapy. The Siberian husky's clinical course consisted of lethargy, fever, weight loss, emesis, anterior uveitis, hepatosplenomegaly, nonregenerative anemia, neutropenia, and thrombocytopenia. Although tetracycline therapy resulted in rapid clinical improvement, the disease course was complicated by concurrent infection with Ehrlichia canis. The unusual finding of spirochetemia in two sick dogs, residing in the same geographic region and the subsequent isolation and preliminary characterization of the observed Borrelia isolate from one dog in which culture of a sample for spirochetes was attempted, prompted this report. CASE REPORTS Dog 1. A previously healthy 3-year-old male Akita was examined on 23 January 1990 at Suburban Animal Hospital because of a 4-day history of anorexia, depression, polydipsia, polyuria, and shifting leg lameness. The dog was allowed to roam and played frequently among acres of pine trees intermingled within marshland that was adjacent to the owner's home located in Alachua County, Fla. Tick infestation was a recurrent problem, and several ticks had recently been removed from the dog. The dog weighed 2.3 kg less than when he was examined 5 months earlier. The dog's rectal temperature was 39.4 C, and the dog was photophobic. There was bilateral mucoid ocular discharge, episcleral injection, dull, lusterless irides, and bilateral miosis, consistent with acute anterior uveitis. Mild lymphadenomegaly and splenomegaly were palpable. Conscious proprioceptive deficits were evident in the left hind limb. Hematologic abnormalities included a microcytic, normochromic anemia (hematocrit, 33.4%; mean corpuscular volume, 58.8 fl; mean corpuscular hemoglobin concentration, 33.2 g/dl) with slight poikilocytosis and an estimated de-

2 VOL. VOL. 32, 32, ~~~~~~~BORRELIA INFECTIONS IN DOGS IN FLORIDA 353 -Affi.. Adbk 91% Aft-h-IMRML- TRI", Akt crease in platelet number. The leukocyte count was 14,300/ pl, with 12,460 segmented neutrophils, 143 band neutrophils, 858 lymphocytes, 286 monocytes, and 286 eosinophils per PI. Numerous spirochetes were observed by the clinical pathologist (A.R.K.) on a Wright-Giemsa-stained blood smear (Fig. 1A). The stained blood film and acute- and convalescent-phase serum samples were mailed to North Carolina State University for comment on the spirochetes that were found to occur between the blood cells. Quantitation of 100 randomly selected oil immersion fields identified an average of 1.4 spirochetes per x 1,000 magnification field. Organisms (n = 10 spirochetes per blood film) measured an average of FIG. 1. Spirochetes observed on a Wright-Giemsa-stained blood smear from dog 1 (A [bar, 30.8 p.m] and B [bar, 43.5 klm]) and dog 2 (C [bar, 30.8 plm]) ± ~tm in length and 0.5 p~m in width and contained an average of 3.8 coils per spirochete. Most spirochetes were loosely coiled, while occasional organisms were tightly coiled. Spirochetes were observed in a small number of neutrophils. Biochemical abnormalities in serum included slightly high serum alkaline phosphatase activity (94 IU/liter; reference range, 5 to 61 IU/liter), serum amylase activity (2,100 U/liter; reference range, 300 to 2,000 U/liter), and serum lipase activity (426 U/dl; reference range, 10 to 180 UIdl). The urine specific gravity was 1.030, and urinalysis was unremarkable except for the presence of course granular casts (<1 cast per x450 magnification field). Treatment consisted of tetracycline, 1 g every 8 h orally for 14 days, and triple antibiotic-corticosteroid ophthalmic ointment every 6 h. Seven days later, the dog was alert and active and his appetite was normal. The anterior uveitis had resolved and the conscious proprioceptive deficits could no longer be elicited. The spleen and lymph nodes remained palpably enlarged. A complete blood count was found to be essentially unchanged from the previous one, except that the platelet estimation was considered adequate and the lymphocyte count had increased (4,940/pl). Reevaluation was recommended, but because the dog seemed healthy, the owner did not return until 28 days later, at which time the dog was lame in the left rear leg. The left stifle joint was swollen. There were several areas of acute moist dermatitis involving the face and neck. The dog's body weight had increased by 1 kg, and his rectal temperature was normal (38.40C). Thrombocytopenia (platelet count, 90,000/ pj) was the only hematologic abnormality. Indirect immunofluorescent-antibody (IFA) tests for E. canis, Rickettsia nickettsii, and B. burgdorferi were negative. Treatment consisted of ampicillin, 1 g every 8 h for 3 weeks.

3 354 BREITSCHWERDT ET AL. J. CLIN. MICROBIOL. With the exception of rupture of the left anterior cruciate ligament, the dog remained healthy for 3 months and was subsequently lost to follow-up. Complete blood counts, repeated 5 and 51 days after the initiation of ampicillin therapy, were normal except for thrombocytopenia (121,000 and 84,000 platelets per pl, respectively); however, moderate platelet clumping was reported and may have contributed to the lower-than-normal platelet counts. Dog 2. A 12-year-old male Labrador retriever was examined on 19 May 1992 because of weight loss of 2 weeks' duration and anorexia of 2 days' duration. The dog roamed on a large farm bordered by a river and a cypress swamp located in Sumter County, Fla., where the dog had resided since moving from Colorado 6 years earlier. The owner reported that the dog was persistently infested with ticks and fleas. During the physical examination, a stick lodged in the caudal oropharynx was removed, and the dog's appetite returned to normal within 24 h. Hematologic abnormalities included a normocytic, hypochromic anemia (hematocrit, 33.7%; mean corpuscular volume, 62.6 fl; mean corpuscular hemoglobin concentration, 30.3 g/dl) and eosinophilia (5,092 cells per pl). Biochemical abnormalities included slightly elevated creatinine phosphokinase and lactate dehydrogenase activities (339 U/liter; [normal, 50 to 280 U/liter] and 690 U/liter [normal, 0 to 367 U/liter], respectively). Dot enzyme-linked immunosorbent assay (ELISA) results for antibodies to B. burgdorferi (Pitman-Moore, Mundelein, Ill.) and Dirofilaria immitus (Pet-CHEK; Idexx Labs, Westbrook, Maine) were negative. During the next month the dog appeared healthy, regained weight, and developed acute moist dermatitis around the muzzle, which resolved spontaneously. Seven weeks after the initial evaluation that revealed an oral foreign body, the owner noticed lethargy, a markedly stiff gait, pallor, and panting. The dog's rectal temperature was 38.7 C. Hematologic abnormalities included a normocytic hypochromic anemia (hematocrit, 18.3%; mean corpuscular volume, 63.1 fl; mean corpuscular hemoglobin concentration, 30.6 g/dl) with slight spherocytosis, lymphopenia (475 cells per pl), and thrombocytopenia (22,000/pl). An ELISA (Pitman-Moore) was positive for antibodies to B. burgdorferi. Numerous spirochetes were seen on a Wright-Giemsa-stained blood smear (Fig. 1B). Stained blood smears, acute- and convalescent-phase serum samples, and sterile blood obtained pretreatment for spirochete culture were sent to North Carolina State University for additional study. Quantitation of 100 randomly selected oil immersion fields identified an average of 0.3 spirochetes per x 1,000 magnification field (n = 10 spirochetes per blood film) that measured an average of 13.3 ± 2.7 p,m in length and 0.3 p,m in width and contained an average of 5.7 coils per spirochete. In contrast to dog 1, most spirochetes were tightly coiled, while occasional organisms were loosely coiled. Radiographs of the vertebral column and the carpal joints revealed periosteal bony proliferation compatible with degenerative osteoarthritis. Treatment consisted of tetracycline hydrochloride per os, 500 mg every 8 h for 7 days, and prednisone per os, 5 mg three times daily on day 1, twice daily on day 2, and once daily on day 3. The dog appeared to be clinically healthy by the second day of therapy. Eight days after the discovery of spirochetemia, the dog remained healthy except for an intermittent ocular discharge. Anemia persisted (hematocrit, 22%), but the platelet count (458,000/,l) had normalized. Aspiration cytology of the right carpal joint was unremarkable. An immunofluorescence assay for E. canis antibodies was negative, but a dot ELISA for B. burgdorferi antibodies (Pitman-Moore) was positive. Doxycycline, 500 mg per os every 12 h, was again administered for 14 days. The dog remained healthy during the 7-month follow-up period. MATERLILS AND METHODS Serology. Acute- and convalescent-phase sera from the two spirochetemic dogs were examined by IFA testing against the Florida canine Borrelia (FCB) isolate (see spirochete isolation below), B. burgdorferi (B31; ATCC 35210), B. anserina (Esparto; John Barnes, North Carolina State University) B. hennsii (HS1 serotype C; ATCC 35209), B. coniaceae (C053; ATCC 43381), R nickettsii, and E. canis and by microscopic hemagglutination testing against Leptospira interrogans serovars, Leptospira canicola, Leptospira grippotyphosa, Leptospira icterohaemorrhagiae, Leptospira pomona, Leptospira hardjo, Leptospira autumnalis, and Leptospira bratislava. Spirochete isolation. Blood was obtained from dog 2 on 9 July 1992 and was received at North Carolina State University and inoculated into culture on 14 July. One drop was harvested from the buffy coat layer of the centrifuged EDTA blood sample containing spirochetes of various sizes and was placed into 8 ml of modified Barbour-Stoenner-Kelly (BSK-H and BSK-II) culture media (Sigma Chemical Co., St. Louis, Mo.) that were incubated at 35 C. Cultures were examined daily by dark-field microscopy. Preliminary taxonomic characterization. (i) Monoclonal antibody testing. Monoclonal antibodies H9724 (5) directed against the 41-kDa flagellin of B. burgdorfiei, a genusspecific monoclonal antibody, and H5332 (6) directed against the 31-kDa outer surface protein (OspA) of B. burgdorferi, a species-specific antigen, were graciously provided by A. G. Barbour, University of Texas Health Science Center, San Antonio. Organisms grown in BSK-II medium were washed twice in phosphate-buffered saline (PBS) and dotted onto each well of a 24-well Teflon-printed slide (Cel-Line Associates, Newfield, N.J.). After air drying, the slides were fixed in acetone for 10 min. Slides were rinsed in PBS, and the monoclonal antibody was applied at a 1/5 dilution in PBS. Slides were incubated in a moisture chamber at 37 C for 30 min and washed three times in PBS (5 min per wash, with agitation), and an optimal dilution (1/200) of fluorescein isothiocyanate (FITC)-labeled goat anti-mouse immunoglobulin G (heavy and light chain specific; Kirkegaard and Perry Laboratories) was applied to each well. The slides were incubated at 37 C for 30 min and washed as described above, and coverslips were affixed by using antifade mounting medium (0.3 M DABCO [1,4-diazabicyclo(2.2.2)-octane] [Aldrich Chemical Co.] in 9:1 glycerin-pbs). The slides were examined under epifluorescence illumination at x400 magnification. PCR analysis of the cultured organism was performed by E. K. Hofmeister, Johns Hopkins University School of Hygiene and Public Health, using a previously described primer targeted at highly conserved regions of the ospa gene of B. burgdorferi (10). (ii) Electrophoresis. Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) was performed on whole-cell lysates of Borrelia sp. strain FCB-1 and B. burgdorferi B31 by a modification of the methods of Laemmli (14). Cultures of each isolate were centrifuged at 10,000 x g for 10 min at 4 C and were washed three times with PBS (ph 7.38). The pellet was resuspended in PBS, and the

4 VOL. 32, 1994 protein concentration was determined by the Bradford method (Bio-Rad Protein Assay; Bio-Rad Laboratories, Richmond, Calif.) after lysis with formic acid. Samples were mixed with an equal volume of sample buffer containing 60 mm Tris hydrochloride (ph 6.8), 2% SDS, 5% 2-3-mercaptoethanol, 10% glycerol, and % (wt/vol) bromophenol blue. The mixture was boiled for 5 min and was applied to the gel at 43,ug of spirochetal protein per well. The lysates were electrophoretically separated in a discontinuous SDS-polyacrylamide gel by using a 4.5% acrylamide stacking gel and a 12.5% acrylamide resolving gel. The relative molecular masses of the proteins were determined by coelectrophoresis of known standards (GIBCO BRL, Grand Island, N.Y.). Electrophoresis was performed at 100 V for 18 h. The completed gel was stained with 0.1% Coomassie blue R-250 in 40% methanol-10% acetic acid aqueous solution overnight and was destained in methanolacetic acid. (iii) EFA screen for reactivity to the Florida Borrelia isolate. Sera from dogs residing in the same geographic region of Florida in which the propositus cases were identified submitted to a commercial diagnostic laboratory between September and December 1992 were stored frozen (-20 C) for screening against the Florida Borrelia isolate by an IFA test. Sera were serially diluted twofold in PBS (ph 7.3) beginning at a reciprocal dilution of 16. Samples reacting at reciprocal IFA titers of 64 were further diluted to obtain an endpoint titer. Organisms grown in BSK-II medium were washed twice in PBS and were dotted onto each well of a 24-well Teflon-printed slide (Cel-Line Associates). After air drying, the slides were fixed in acetone for 10 min. Slides were rinsed in PBS, and the serial dilutions of samples were applied. Slides were incubated in a humid chamber at 37 C for 30 min and were washed three times in PBS (5 min per wash, with agitation), and an optimal dilution (1/100) of FITC-labeled, goat anti-dog immunoglobulin G (gamma chain; Kirkegaard and Perry Laboratories) was applied to each well. The slides were incubated at 37 C for 30 min and washed as before, and coverslips were affixed by using anti-fade mounting medium (0.3 M DABCO in 9:1 glycerin- PBS). The slides were examined under epifluorescence illumination at x400 magnification. Sera from unvaccinated specific-pathogen-free dogs originating from New York State were used as negative controls. Sera from dog 2 and sera from a B. burgdorferi-positive dog from Connecticut were used as positive controls. RESULTS Serology. Acute- and convalescent-phase sera from dogs 1 and 2 did not contain detectable antibodies to E. canis, spotted fever group rickettsiae, or Leptospira serovars. Both dogs seroconverted to the FCB isolate and, in general, but to a lesser degree, to the other Borrelia species tested (Table 1). Reciprocal IFA titers against the homologous organism were 32 on 7 March 1990 and 128 on 20 April 1990 for dog 1 and 256 on 10 July 1992, 2,048 on 17 July 1992, 512 on 24 July 1992, and 128 on 5 October 1992 for dog 2. Similarly, acute-phase sera from both dogs contained IFAs to B. burgdorferi, and dog 2 seroconverted to the B. burgdorfen antigen, with titers being less than or identical to the titers for the homologous organism. Spirochete isolation. Twenty-four hours after culture inoculation, two to three nonmotile spirochetes per x400 magnification field were observed by dark-field microscopy. Within 48 h, translational, rotational, and flexing movements BORRELIA INFECTIONS IN DOGS IN FLORIDA 355 Species TABLE 1. IFA titers to various species of Borrelia Reciprocal titer' Dog 1 Dog 2 Dog 3 Dog 4 FCB isolate B. burgdorferi ,024 B. conaceae B. hernsii B. anserina <16 < ' Convalescent-phase serum samples from spirochetemic dogs 1 and 2, dog 3 (a high-titer serosurvey sample), and dog 4 (presumably exposed to B. burgdorferi in Connecticut) were simultaneously tested against various Borrelia antigens. were observed, and by postinoculation day 6, multiplication was noted. Preliminary taxonomic characterization. The FCB isolate was bound by monoclonal antibody H9724, but was not bound by MAB monoclonal antibody PCR analysis amplification of the ospa gene of B. burgdorferi failed to detect B. burgdorfiei DNA. The SDS-PAGE protein profile of the Florida isolate (FCB-1) differed markedly from that of the type strain (B31) of the Lyme disease agent, B. burgdorfeni (Fig. 2). Major protein bands were located in the 39- and 42-kDa regions. Other less intense banding was noted in the 60- to 66-kDa regions. IFA screen for reactivity to the FCB isolate. Of the 99 serum samples that were tested by IFA for FCB antigen, 82 (81%) samples did not contain reactive antibodies (reciprocal titer, <16), 5 samples were reactive at a reciprocal titer of 64, 6 samples were reactive at a reciprocal titer of 128, 4 samples were reactive at a reciprocal titer of 256, and 2 samples were MW FCB B31 ~-I FIG. 2. SDS-PAGE protein profile of whole-cell lysates stained with Coomassie blue R-250. MW, molecular mass standards (in kilodaltons); B31, B. burgdorferi B31. _

5 356 BREITSCHWERDT ET AL. TABLE 2. Comparison of serologic reactivity to the FCB Borrelia isolate and B. burgdorferi in 20 seropositive samples from Florida dogs Sample no. FCB isolate Reciprocal titer B. burgdorferi < < < < < < < < reactive at a reciprocal titer of 512. Of these 17 seropositive samples, 4 were also reactive by IFA at reciprocal titers of >64 to B. burgdorferi antigen (Table 2). DISCUSSION Detection of spirochetemia on a Wright-Giemsa-stained blood smear from a sick dog from Florida initiated efforts to culture and characterize the isolate. The binding of monoclonal antibody H9724 (5) to the isolate indicates that the spirochete is a member of the genus Borrelia. Lack of binding by monoclonal antibody H5332 (6) to the organism, failure to amplify the ospa gene for B. burgdorfeni (10), results of the gel analysis, and documentation of clinical and hematologic abnormalities not previously associated with B. burgdorfen infection in dogs indicate that the Florida canine spirochete is not B. burgdorferi. With the exception of a case report (18) in which definitive characterization of the cultured spirochete was not detailed, visible spirochetemia and hematologic abnormalities, such as anemia and thrombocytopenia, have not been reported in association with experimental or naturally occurring B. burgdorferi infections in dogs (2, 3, 21). Experimentally, spirochetemia and illness have been induced in dogs with several relapsing fever Borrelia species (8). However, several factors suggest that this isolate is not a relapsing fever Borrelia isolate. Although Omithodoros ticks feed indiscriminately on many different kinds of animals, including humans, indigenous foci of relapsing fever have not been reported in people living in the eastern United States (7). When guinea pigs were inoculated with the isolate, a relapsing fever pattern was not observed (unpublished data). In addition, antibody titers to the homologous organism (i.e., the FCB isolate) in the two spirochetemic dogs, and in particular dog 2, were higher when compared with the lower cross-reactive titers to B. burgdorfeni or to the relapsing fever Borrelia species tested. These observations suggest that this canine isolate may represent a newly recognized species of Borrelia. Because neither dog developed antibodies to serovars of L. interrogans in either acuteor convalescent-phase serum samples, exposure to leptospires is unlikely. Additional studies are in progress to J. CLIN. MICROBIOL. further characterize the taxonomic relationship of this isolate to known Borrelia species. The clinical and laboratory evaluations of only two dogs limit any detailed conclusions regarding the potential clinical importance of this spirochete. In both dogs, the duration of illness was brief, suggesting an acute-phase disease. It was fortuitous that dog 2 was tested for antibodies to B. burgdorferi by dot ELISA (manufacturer's reported sensitivity of 96% and specificity of 100%) when he was examined for an oral foreign body and 7 weeks later when spirochetemia was detected. The initial result was negative, but the subsequent evaluation was positive, supporting the supposition of acutephase disease and indicating, as supported by the results presented in Tables 1 and 2, a potential problem with serologic cross-reactivity. Since spirochetemia was recognized 31 months apart in two dogs residing in different counties located 45 miles (72 km) apart, it is probable that an established endemic focus of this organism exists in the region. This conclusion is further supported by the limited serosurvey of 99 sick dogs from the region that identified a seroprevalence of 17%. In comparison with a previous serosurvey (9) that identified an overall prevalence rate of B. burgdorferi antigen in 1,002 dogs from North Carolina of 3.2%, the high seroprevalence in the present study was unexpected. Both dogs described in this report had similar clinical and hematologic abnormalities. The unusual finding of spirochetemia in two dogs from the same geographic region during a brief period of time and the serologic responses of both dogs to the Borrelia organism isolated from dog 2 suggest that these dogs were exposed to the same organism. Although the possibility exists that spirochetemia may have been coincidental to another disease process, several factors support a pathogenic role for this spirochete in dogs. Following treatment with tetracycline hydrochloride, an antibiotic that is efficacious for treating spirochetal disease (12, 13), there was rapid improvement in clinical signs, including fever and lethargy in both dogs, stiffness in dog 2, and anterior uveitis and lameness in dog 1. Although not sequentially characterized, the rapid resolution of laboratory abnormalities, particularly the severe thrombocytopenia, following the administration of tetracycline also supports a pathogenic role for this organism. Perhaps the most convincing evidence of pathogenicity is that seroconversion was followed by a sequential decrease in antibody titer to the organism in dog 2 after antimicrobial therapy. Despite the visualization of large numbers (average counts, 0.3 to 1.4 organisms per x 1,000 magnification field) of spirochetes on Wright-Giemsa-stained blood smears from both of these dogs, it should be emphasized that spirochetemia was not readily discernible to an experienced hematologist during the initial blood smear examinations. Once spirochetes were seen on the blood smear, it was easy to find additional organisms in other microscopic fields. Because of the difficulty in visualizing the organism, spirochetemia in dogs associated with this Borrelia species could remain undetected. This emphasizes the need for future studies to characterize the extent and duration of the spirochetemia, the pathogenic role of the organism, the kinetics of the humoral antibody response, and the development of reagents to facilitate the diagnosis of naturally occurring disease. Historically, both dogs experienced recurrent tick infestation, a frequently encountered problem for dogs throughout the year in the region described here. A common feature for both dogs included repeated exposure to swamp and marsh-

6 VOL. 32, 1994 land habitats that contained numerous small mammals, including small rodents, and reptiles, such as turtles, snakes, and alligators. Although a nonspecific and frequently encountered problem in warm, moist environments, acute moist dermatitis involving the face was observed in both dogs either 1 month prior to or 1 month after the onset of illness. This observation might reflect facial contact with dirt or irritant substances that might occur during digging in burrows or could represent a cutaneous manifestation of the spirochetal disease (19, 22). We conclude that a member of the genus Borrelia may cause disease in dogs in Florida. The extent to which this Borrelia isolate contributes to the incidence of unexplained tetracycline-responsive fever, lameness, anterior uveitis, or thrombocytopenia in dogs located in the southeastern United States deserves further consideration. Importantly, serologic cross-reactivity with B. burgdorferi following exposure to our isolate could contribute to the inaccurate diagnoses of canine Lyme disease in Florida and perhaps other regions of the United States. Studies related to the geographic distribution, vector transmission, and pathogenicity of this isolate in various species are in progress. ACKNOWLEDGMENTS We thank Laurie Champion, Laura Cullins, and Peter Howard for excellent technical assistance and John Mounger and Mark Yates for providing information related to the clinical evaluation of dog 2. REFERENCES BORRELA INFECrIONS IN DOGS IN FLORIDA Anderson, J. F., S. W. Barthold, and L. A. Magnarelli Infectious but nonpathogenic isolate of Borrelia burgdorfen. J. Clin. Microbiol. 28: Appel, M. J. G Canine Lyme disease: toward satisfying Koch's postulates, p In R. W. Kirk and J. D. Bonagura (ed.), Current Veterinary Therapy XI, Small Animal Practice. The W. B. Saunders Co., Philadelphia. 3. Appel, M. J., S. Allan, R. H. Jacobson, T. L. Lauderdale, Y. F. Chang, S. J. Shin, T. W. Thomford, R J. Todhunter, and B. A. Summers Experimental Lyme disease in dogs produces arthritis and persistent infection. Infect. Dis. 167: Barbour, A. G Antigenic variation of a relapsing fever Borrelia species. Annu. Rev. Microbiol. 44: Barbour, A. G., S. F. Hayes, R. A. Heiland, M. E. Schrumpf, and S. L. Tessier A Borrelia genus-specific monoclonal antibody binds to a flagellar epitope. Infect. Immun. 52: Barbour, A. G., S. L. Tessier, and W. J. Todd Lyme disease spirochetes and Lodes tick spirochetes share a common surface antigenic determinant as defined by a monoclonal antibody. Infect. Immun. 41: Burgdorfer, W., and T. G. Schwan Borrelia, p In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 8. Felsenfeld, Borreliae, human relapsing fever, and parasite-vector-host relationships. Bacteriol. Rev. 29: Greene, R. T., J. F. Levine, E. B. Breitschwerdt, and H. A. Berkhoff Antibodies to Borrelia burgdorferi in dogs in North Carolina. Am. J. Vet. Res. 49: Hofmeister, E. K., R. B. Markham, J. E. Childs, and R. R. Arthur Comparison of polymerase chain reaction and culture for detection of Borrelia burgdorferi in naturally infected Peromyscus leucopus and experimentally infected C.C-17 scid/ scid mice. J. Clin. Microbiol. 30: Johnson, R. C., W. Burgdorfer, R. S. Lane, A. G. Barbour, S. F. Hayes, and F. W. Hyde Borrelia conaceae sp. nov.: putative agent of epizootic bovine abortion. Int. J. Syst. Bacteriol. 37: Johnson, R. C., C. Kodner, and M. Russell In vitro and in vivo susceptibility of the Lyme disease spirochete Borrelia burgdorfien, to four antimicrobial agents. Antimicrob. Agents Chemother. 31: Johnson, W. D Borrelia species, p In G. L. Mandell, R. G. Douglas, and J. E. Bennett (ed.), Principles and practice of infectious diseases, 3rd ed. Churchill Livingstone, New York. 14. Laemmli, U. K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227: Levine, J. F., W. L. Dykstra, W. L. Nicholson, R. L. Walker, and G. Massey Attenuation of Borrelia anserina by serial passage in liquid medium. Res. Vet. Sci. 48: Lissman, B. A., E. M. Bosler, H. Camay, B. B. Ormiston, and J. L. Benach Spirochete-associated arthritis (Lyme disease) in a dog. J. Am. Vet. Med. Assoc. 185: Magnarelli, L. A., J. F. Anderson, and A. B. Schreier Persistence of antibodies to Borrelia burgdorferi in dogs of New York and Connecticut. J. Am. Vet. Med. Assoc. 196: Moreland, K. J., E. A. Wilson, and R. B. Simpson Concurrent Ehrlichia canis and Borrelia burgdorferi infections in a Texas dog. J. Am. Anim. Hosp. Assoc. 26: Pfister, K., B. Bigler, J. Newswadba, and A. Aeschlimann Borrelia burgdorfeni infections of dogs in Switzerland. Zentrablatt. Mikrobiol. Bakteriol. Hyg. Suppl. 18: In G. Stanely, W. Kristoferitsch, M. Pletschette, A. G. Barbour, and H. Flamm (eds). Lyme Borreliosis II. Gustav Verlag, New York. 20. Schalm, 0. W Uncommon hematologic disorders: spirochetosis, trypanosomiasis, leishmaniasis, Pelger-Huet anomaly. Canine Pract. 6: Wasmoen, T. L., R. W. Sebring, B. M. Blumer, L. G. Chavez, H. J. Chu, and W. M. Acree Examination of Koch's postulates for Borrelia burgdorferi as the causative agent of limb/joint dysfunction in dogs with borreliosis. J. Am. Vet. Med. Assoc. 201: Weber, A., U. Heim, and R. Schafer Occurrence of antibodies to Borrelia burgdorferi in dogs of a small animal practice in Northern Bavaria. Berl. Munch. Tierarztl. Wochenschr. 104: