key: cord-0007618-27875dd6 authors: Shin, Sang J.; Chang, Yung-Fu; Jacobson, Richard H.; Shaw, Eric; Lauderdale, Tsai-Ling; Appel, Max J.; Lein, Donald H. title: Cross-reactivity between B. burgdorferi and other spirochetes affects specificity of serotests for detection of antibodies to the Lyme disease agent in dogs date: 2002-11-13 journal: Vet Microbiol DOI: 10.1016/0378-1135(93)90137-v sha: dcec4338079dfaf6a1748e339ca55efd08944994 doc_id: 7618 cord_uid: 27875dd6 Western immunoblots, the kinetics-based enzyme-linked immunosorbent assay (KELA), and the microagglutination test were used to evaluate cross-reactivity among antibodies to serovars of Leptospira interrogans (leptospiral serovars), and B. burgdorferi from naturally infected dogs, and to Serpulina (Treponema) hyodysenteriae from vaccinated rabbits. Whole-cell lysates from Borrelia spp., leptospiral serovars, and Serpulina spp. were used for SDS-PAGE, western blots, and KELA. Cross-reactivity occurred between the antibodies to B. burgdorferi and leptospiral serovars when tested on the heterologous antigens. Antibodies to leptospiral serovars tended to cross-react more strongly with antigens of B. burgdorferi spp. than did antibodies to B. burgdorferi when tested against antigens of leptospiral serovars. The antibodies against B. burgdorferi showed a lesser degree of cross-reactivity to the antigens of S. hyodysenteriae and S. innocens than they did to leptospiral serovars. We conclude that cross-reactivity occurs between B. burgdorferi and leptospiral serovars. Validation and interpretation of ELISA tests for detection of antibody activity to whole cell lysates of the Lyme agent must take this cross-reactivity into consideration. Conversely, dogs infected with the Lyme agent do not show significant cross-reactivity in the microagglutination test for antibody to the leptospiral serovars. probably useful as sentinel animals for investigations of the epidemiology of Lyme disease (Magnarelli, et al., 1985; Schulze, et al., 1986) . Dogs may develop arthritis after infection (Lissman, et al., 1984; Kornblatt, et al., 1985; Roush, et al., 1989 ) or kidney disease (Grauer, et al, 1988 ) . It has also been reported that febrile spirochetemic dogs showed migratory lameness, and pain and swelling in the carpal joints (Kornblatt, et al., 1985 ) . Recently, surveys using blood samples from dogs with limb/joint disorders or clinical signs of unknown etiology such as fever, anorexia, or fatigue, indicated that 76.3% and 66.5% ofsera were Lyme-antibody-positive for the dogs living in the lower Hudson Valley of New York state and southern Connecticut, respectively (Magnarelli, et al., 1987) . Schulze et al. (1987) reported that 34.7% of Lyme sero-positive dogs from New Jersey were symptomless. The prevalence rate is estimated at approximately 2.7% in North Carolina and 5.5% of sera from dogs in Texas were positive for Lyme antibodies (Cohen, et al., 1990) . Several common serologic techniques have been used for presumptive confirmation of B. burgdorferi infection in dogs in our laboratory and others; these include the indirect immunofluorescence assay (IFA), kinetics-based enzyme-linked immunosorbent assay (KELA), and Western blot (Lindenmayer et al., 1990; Karlsson, 1990; Karlsson et al., 1989; Zoller et al., 1991 ) . The specificity of serotest results for dogs, both in surveillance and clinical applications, thus becomes an important issue. In this paper, we evaluate the protein profiles of Borrelia spp., leptospiral serovars, and Serpulina spp. using the Western blot technique in an attempt to assess the extent and degree of cross reactions among antibodies produced against these organisms. Correlation of Western blot results with those of KELA are also reported. Sixteen isolates of Borrelia burgdorferi, one isolate each of B. anserina, B. hermsii, Leptospira interrogans serovars hardjo, grippotyphosa, pomona, icterohaemorrhagiae, and canicola (leptospiral serovars ), three isolates of Serpulina (Treponema) hyodysenteriae and two isolates of Serpulina (Treponema) innocens were used in these studies (Table 1 ) . All Borrelia species were cultured in BSK medium (Barbour, 1984) . Leptospiral serovars were cultured in leptospira broth (Bovuminar microbiological media solution, PLM5, Intergen Company, Purchase, NY). S. hyodysenteriae and S. innocens were cultured in trypticase soy broth with glucose, containing 5% horse serum, 0.5% yeast extract, 2.0% each ofVPI A (0.04% CaC12, 0.04% MgSO4) and B (0.2% K2HPO4, 2.0% NaHCO3) salt solutions and 0.05% L-cysteine; these orga- Whole-cell lysates of all of the bacterial isolates were prepared. Sixteen isolates ofB. burgdorferi were initially isolated from Ixodes dammini ticks and rodents ( Table 1 ) . B. anserina and B. hermsii were grown in 20 ml of BSK liquid media at 34°C incubator without CO2, washed with phosphate buffer saline solution (PBS; 0.01 M, pH 7.2) three times, and the final pellet was resuspended in SDS-sample buffer (0.01% Tris hydrochloride, 10% glycerol, 2% sodium dodecyl sulfate, 0.5% 2-mercaptoethanol, 0.1% bromphenol blue, pH 8.0), boiled for 3 minutes, and spun 5 min in an Eppendorf centrifuge (13000 rpm). The supernatants were either used immediately or stored at -20 o C until used. Leptospiral serovars were initially obtained from the National Reference Center for Leptospirosis, NVSL (Ames, IA), and have been maintained in our laboratory in PLM5 media. Whole cell lysates were prepared by boiling in the SDS-sample buffer and treated as described above for B. burgdorferi. Lyophilized killed whole cells of S. hyodysenteriae and S. innocens were prepared by boiling in the SDS-sample buffer as described for the B. burgdorferi antigen. Serum samples from dogs were tested for B. burgdorferi antibodies using a computer-assisted, kinetics-based ELISA (KELA). Microdilution plates (Maxisorp F16, Nunc, Denmark) were coated for three hours in a moist chamber at 37 ° using 100 #l/well of supernatant of French pressed B. burgdorferi strain 25550 ( 16000 psi, 3 X ) diluted 1 : 1200 in carbonate buffer (pH 9.6 ). Wells exposed to carbonate buffer and the washing steps served as background control for each sample. Antigen (Ag) coated plates were stored in sealed plastic bags at -20°C and used within a few days of preparation. Prior to testing, Ag coated plates were thawed and the wells were washed 4 times with 0.1 M phosphate buffered saline containing 0.05% Tween (PBST). Serum samples and controls were diluted 1 : 100 in freshly prepared 0.1 M PBST containing 2% nonfat dry milk (Carnation Co., Los Angeles, California) and dispensed ( 100 #1 ) into antigen coated plates in quadruplicate. Plates were incubated at 37 °C for one hour and washed with PBST. One hundred #1 of conjugate consisting of horseradish peroxidase (HRP) linked to goat antidog serum (heavy and light chain-specific) was added to each well and incubated for 30 rain at room temperature. (The antibody concentration in the conjugate was 8.0 mg/ml diluted 1 : 2000 in PBST with 2% milk. ) The plate was then washed with PBST and 100 /tl of freshly prepared substrate (3,3',5,5'-tetramethylbenzidine, 0.4 mg/ml, mixed with equal volume of 0.02% H202 in a citric acid buffer) was added to each well. Plates were placed on a Bio-Tek EL312 reader and three readings (650 nm ) were taken at 1 min intervals with continuous shaking between readings. Results were calculated as the slope of the reaction rate for each sample (Barlough et al., 1986 ) . Several hundred samples from dogs that were either symptomatic or asymptomatic for Lyme disease, were evaluated previously by KELA, IFA, and Western blot. Also, SPF dogs exposed to infected ticks developed antibody responses which, upon Western blotting, had patterns identical to those of naturally-infected dogs (Jacobson et al., 1992 ) . Western blot confirmed that For the detection of antibodies to leptospiral serovars, the microscopic microagglutination technique (MAT) was used as described (USAHA, 1987). Sera Anti-B. burgdorferi sera from dogs naturally exposed to the Lyme disease agent were evaluated individually by KELA and were pooled for use in Western blot. Group 1 consisted of a pool from 32 dogs, group 2 from 18 dogs, group 3 from 14 dogs, group 4 from 16 dogs, and group 5 from 18 dogs. Antibodies to leptospiral serovars were derived from dogs inhabiting areas that were non-endemic for Lyme disease. These dogs had no possible exposure to ticks harboring the B. burgdorferi. The titers were evaluated individually by MAT and sera were pooled for use in Western blots. Three groups of leptospira seropositive samples were pooled. First group containing 12 dogs, the second group containing 15 dogs and the third group containing 19 dogs. Titers of the pooled sera are shown in Table 2 . Rabbit antisera to S. hyodysenteriae were the gift of Dr. Wannemuehler. All sera were kept in -20 ° C until use. Antigens prepared as aforementioned were subjected to SDS-polyacrylamide gel electrophoresis (12% acrylamide) at 200 V for 4-6 hours (Chang et al., 1987 ) . The gels were either silver stained or proteins were transferred to nitrocellulose membranes electrophoretically. Immunoreactive proteins were detected by Western blot analysis (Towbin et al., 1979) as previously described (Chang et al., 1989 ) . Antibody pools, as described above, were used as the first antibody, followed by a second antibody conjugate consisting of goat anti-dog IgG linked to horse-radish peroxidase (Cappel Laboratories, Inc. lot #31340). SDS-PAGE patterns of the spirochetes were compared (Fig. 1 ) . All B. burgdorferi isolates showed many major protein components with patterns, particularly of the major bands (60, 47, 41, 39 and 20 kDa), that are consistent with previously published data . Several isolates (lane 1, 3, 4, 6, 7, 8, 11, 13 and 14) showed strong bands at 34 kDa which may be the outer surface protein B (OspB). Other isolates ( 1-5, Fig. 2) showed a common pattern (62, 46, 43, 39 .5 and 32 kDa) and several different bands among different serovars (40 kDa for serovar grippotyphosa -lane 4; 36 kDa for serovar icterohaemorrhagiae -lane 3; 30 kDa for serovar pomona -lane 1 ). However, the general SDS-PAGE pattern of leptospiral serovars was distinctly different from Borrelia spp. (Fig. 2) . The five groups of pooled sera from dogs naturally infected with the Lyme agent all had strong antibody activity in KELA but had no detectable anti-body to the leptospiral serovars antigens ( Table 2 ). The three groups of pooled sera from dogs naturally infected with leptospiral serovars developed antibody to the 4 leptospiral serovars as expected. These sera also had some reactivity against B. burgdorferi in the KELA test (Table 2 ) . Five leptospiral serovars, 8 isolates of B. burgdorferi, and one isolate each ofB. anserina and B. hermsii were subjected to SDS-PAGE and western blot analysis. The western blots were developed by subjecting one blot (Fig. 3 ) to group 2 pooled sera from dogs infected with the five different leptospiral serovars while the other blot (Fig. 4) was developed by using the group 5 pooled sera from dogs naturally exposed to the Lyme agent. The protein patterns between leptospiral serovars and Borrelia spp. were easily distinguishable. When developing the immunoblots with an antibody pool consisting of an- When sera from dogs infected with the Lyme disease agent were used, two distinct bands appeared for the 5 leptospiral serovars with molecular weight of 32 and 44 kDa (Fig. 4) . The smear staining of 30-20 kDa for lysates of serovar grippotyphosa, pomona, icterohaemorrhagiae, and canicola (lanes 2-5 ) probably represents carbohydrate (Fig. 4) . The other four pools of serum from dogs infected with the Lyme agent, and the two other pools of sera from dogs infected with leptospiral serovars spp. showed western blot patterns very similar to those shown in Figs 2-4 (data not shown). When using Lyme infection-specific antibody for immunoblotting, several faint bands with molecular weights varying from 40, 60 and 70 kDa appeared for isolates ofS. hyodysenteriae (lanes 4 and 5 ), and bands of 42, 60, and 68 kDa in the 2 strains ofS. innocens (Fig. 5A) . Two faint bands (55 and 52 kDa) appeared in three of the B. burgdorferi strains and another 68 kDa band developed in strain 25550 (lane 2) and AH2 (lane 3) when anti-S, hyodysenteriae antibody was used in immunoblotting (Fig. 5B ) . DISCUSSION Western immunoblotting has been used to analyze the antibody response or to confirm Lyme disease infection in human patients (Karlsson et al., 1989; Karlsson, 1990; Zoller et al., 1991 ) . Western blotting has confirmed KELA results in our laboratory and has also been used elsewhere as a confirmatory test for dog Lyme disease Lindenmayer et al., 1990; Jacobson, et al., 1992) . SDS-PAGE of the total lysate of the five leptospiral serovars species showed distinctly different protein profiles when compared to the Borrelia species antigens (Fig. 2 ) . The B. anserina strain also showed a different SDS-PAGE pattern from that ofB. burgdorferi isolates. Based on the protein profiles from SDS-PAGE, differentiation of leptospiral serovars, B. burgdorferi, and B. anserina was not difficult. In western immunoblotting, antibody to B. burgdorferi cross-reacted slightly with leptospiral antigens and antibody to the leptospiral serovars reacted somewhat more vigorously with antigens of B. burgdorferi. The protein profiles were qualitatively different and staining patterns were less intensive in the heterologous systems indicating relatively limited quantities of cross reacting antibodies ( Fig. 3 and Fig. 4 ). When reacted with Borrelia antigen, antibodies from the leptospiral pool showed dominant bands at approximately 41 kDa, which is consistent with the flagellin antigen, and at 60 kDa, which may reflect a heat shock protein of Borrelia; some other protein bands also were developed by antibody to leptospiral serovars (Fig. 3) . When we tested sera from the Lyme pool against antigens of the leptospiral serovars, two main bands appeared with molecular weight of approximately 32 and 44 kDa, respectively. A smear between 30-18 kDa is probably a carbohydrate; further identification of this substance is necessary. These results indicate that the antibody to the leptospiral serovars resulted in stronger cross reactions to B. burgdorferi antigens than did sera from the B. burgdorferi antibody pool when tested against leptospiral antigens. These observations suggest that dogs exposed to leptospiral antigens may have antibodies that will cross-react in serotests for the antibody to the Lyme agent. In KELA, the range of activity observed for several thousand clinical samples and dogs exposed experimentally to infected ticks ranged from 0 to about 600 slope units (Jacobson et al., 1992) . Activity from 0 to 100 units could not be distinguished from background activity and samples in this range have been considered negative in KELA while antibody activity ranging from 100 to 200 units is considered as low-positive based on western immunoblotring confirmation. From Table 2 , it is clear that antibody to the leptospiral serovars is detected as low positive in the KELA serotest for antibody to B. burgdorferi. It is, therefore, probable that the same cross-reactivity occurs in other ELISAs that use whole-cell lysates ofB. burgdorferi. Western blot also showed a weak cross reaction between B. burgdorferi and S. hyodysenteriae or S. innocens (Fig. 5. A and B ) . Such reactions may not be a problem in the diagnosis of canine Lyme disease, since S. hyodysenteriae or S. innocens cause disease only in pigs. Since leptospirosis is an important disease in dogs in the USA, most dogs have received a leptospiral vaccine. Because western blots confirm serological cross-reactions between leptospiral serovars and Borrelia species, they constitute a useful tool in validating the specificity of ELISA tests. Furthermore, since it is not difficult to differentiate between antibodies to the leptospiral serovars and B. burgdorferi in western blots, the technique can be a useful aid in the diagnosis of canine Lyme disease. Because validation of the KELA assay is based on western immunoblotting of sera of known Lyme agent infection status, the cutoff between seropositive and seronegative dogs in the KELA Lyme test is based on objective criteria. The KELA is thus a valuable serotest for the confirmation of canine Lyme disease. Epizootiology of Borrelia in Ixodes tick vectors and reservoir hosts Isolation and cultivation of Lyme disease spirochetes Coronavirus antibody detection in cats by computer-assisted kinetics-based enzyme-linked immunosorbent assay (KELA): Field studies Arthritis and panuveitis as manifestations of Borrelia burgdorferi in a Wisconsin pony Identification and characterization of the Pasteurella haemolytica leukotoxin Cloning and characterization of a hemolysin gene from Actinobacillus (Haemophilus)pleuropneumoniae Clinical and epizootiologic characteristics of dogs seropositive for Borrelia burgdorferi in Texas: 110 cases Renal lesions associated with Borrelia burgdorferi infection in a dog Antibodies to Borrelia burgdorferi in dogs in North Carolina Immunoblot analysis ofimmunoglobulin G response to the Lyme disease agent (Borrelia burgdorferi) in experimentally and naturally exposed dogs Canine antibody responses to Borrelia burgdorferi induced by exposure to infected ticks, injection of viable organisms, or vaccination Comparison of western blot and enzyme-linked immunosorbent assay for diagnosis of Lyme borreliosis Western immunoblot and flagellum enzyme-linked immunosorbent assay for serodiagnosis of Lyme borreliosis Arthritis caused by Borrelia burgdorferi in dogs Comparison of indirect immunofluorescent-antibody assay, enzyme-linked immunosorbent assay, and western immunoblot for the diagnosis of Lyme disease in dogs Spriocheteassociated arthritis (Lyme disease) in a dog Borreliosis in dogs from southern Connecticut Clinical and serologic studies of canine borreliosis Rheumatoid arthritis subsequent to Borrelia burgdorferi infection in two dogs Prevalence of canine Lyme disease from an endemic area as determined by serosurvey Prevalence of canine Lyme disease from an endemic area as determined by serosurvey Lyme arthritis: an epidemic of oligoarticular arthritis in children and adults in three Connecticut communities Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets Validity of Western immunoblot band patterns in the serodiagnosis of Lyme borreliosis USAHA We are grateful to Helen Bell for administrative assistance and to the Biomedical Communications Office of the College of Veterinary Medicine at Cornell University for photographic services. Thanks to Dr. Michael J. Wannemuehler, Department of Verterinary Microbiology and Preventive Medicine, Iowa State University for supplying S. hyodysenteriae, and S. innocens strains and anti-S, hyodysenteriae antibody, and to Dr. John Anderson, Connecticut Agricultural Experiment Station, New Haven, Connecticut for supplying B. anserina and B. hermsii. This work was supported by the Cornell University Alumni Fund and New York State Department of Agriculture and Markets.