results obtained after experimental exposure can be extrapolated to natural exposures since experimentally exposed

JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1988, p. 648653 00951137/88/0464806$02.00/0 Copyright 1988, American Society for Microbiology Vol. 26, No. 4 Immunoblot Analysis of Immunoglobulin G Response to the Lyme Disease Agent (Borrelia burgdorferi) in Experimentally and Naturally Exposed Dogs RUSSELL T. GREENE,'* RICHARD L. WALKER,2 WILLIAM L. NICHOLSON,2 HANS W. HEIDNER,2 JAY F. LEVINE,2 ELIZABETH C. BURGESS, MICHAEL WYAND,4 EDWARD B. BREITSCHWERDT,1 AND HERMAN A. BERKHOFF2 Departments of Companion Animal and Special Species Medicine' and Microbiology, Pathology and Parasitology,2 North Carolina State University School of Veterinary Medicine, Raleigh, North Carolina 27606; Research Animal Resources Center and Department of Medicine, School of Veterinary Medicine, University of WisconsinMadison, Madison, Wisconsin 537063; and New England Regional Primate Research Center, Southborough, Massachusetts 017724 Received 17 August 1987/Accepted 8 January 1988 Immunoblots were used to study the immunoglobulin G response to Borrelia burgdorferi in experimentally and naturally exposed dogs. Adsorption studies confirmed that the antibodies were specific for B. burgdorferi. Experimentally exposed dogs were asymptomatic. Naturally exposed dogs included both asymptomatic animals and animals showing signs compatible with Lyme disease. Naturally exposed dogs were from four geographic regions of the country. No differences were detected between immunoblot patterns of naturally exposed symptomatic or asymptomatic dogs from different areas of the country. The immunoblot patterns obtained with sera from experimentally exposed dogs were different from those obtained with sera from naturally exposed dogs and were characterized by reactivity toïfewer and different protein bands. Immunoblot analysis using an OspAproteinproducing Escherichia coli recombinant showed that experimentally exposed dogs produced antibodies to OspA, whereas naturally exposed dogs did not. Modifications of the immune response over time, different routes of antigen presentation, and strain variation are factors postulated to account for thé observed differences. Lyme disease is a tickborne, spirochetal disease with clinical manifestations in humans including dermatitis, carditis, neuritis, and arthritis (17). Initial reports of Lyme disease in dogs described arthritis as the only clinical manifestation (10, 12, 13), but a more recent report described skin lesions, lameness, vomiting, and abortion (8). Since the causative organism, Borrelia burgdorferi, has been difficult to isolate from affected human and canine patients, serology has been widely used as a diagnostic aid (6, 8, 10, 12, 13, 17). The exact nature of the canine humoral response is poorly defined. An elevated serologic titer in the presence of compatible clinical signs has been used to diagnose Lyme borreliosis (8, 10, 13). However, previous reports have shown that high antibody titers may develop in asymptomatic dogs (10, 13). In addition, B. burgdorferi has been isolated from the blood of some dogs with low antibody titers (6, 12). These findings emphasize the difficulties associated with making a definitive diagnosis of canine borreliosis. In humans, it takes approximately 1 to 1.5 months after exposure for specific immunoglobulin G (IgG) titers to become detectable (7, 16). These titers slowly increase during the progression of the disease and often peak months to years after clinical remission. Immunoblots have been used to demonstrate expansion of the IgG response, involving up to 11 antigens late in the disease (1, 7). Indepth knowledge of the canine serologic response is lacking. In experimental studies, IgG titers increased within 21 days and remained elevated for up to 10 months after dogs were infected with B. burgdorferi (5; R. T. Greene, J. F. Levine, E. B. Breitschwerdt, R. L. Walker, H. A. Berkhoff, J. Cullen, and W. L. Nicholson, Am. J. Vet. Res., in press). * Corresponding author. 648 In natural infections it is difficult to determine the time of exposure, the interval before a specific IgG response becomes detectable, and the total length of time that the IgG response remains elevated. It is not known whether the results obtained after experimental exposure can be extrapolated to natural exposures since experimentally exposed dogs have failed to demonstrate clinical signs. Accordingly, immunoblots were performed on canine sera positive for antibodies to B. burgdorferi by enzymelinked immunosorbent assay (ELISA) and indirect immunofluorescence assay (IFA). Naturally exposed dogs were both asymptomatic and symptomatic, whereas experimentally exposed dogs were asymptomatic. Serum samples from naturally exposed dogs were collected in four geographical regions of the country. The goals of this study were (i) to determine whether clinical signs could be related to differences in immunoblot patterns, (ii) to determine whether immunoblot pattern differences existed among dogs from different geographical regions, and (iii) to evaluate whether immunoblot pattern differences existed between naturally and experimentally exposed dogs. (This report represents a portion of a thesis submitted by R. T. Greene as partial fulfillment of the requirements for the Ph.D. degree from North Carolina State University, Raleigh.) MATERIALS AND METHODS Experimentally infected animals. Eight laboratoryreared, male Beagles were experimentally exposed to B. burgdorferi. The experimental design and clinical and serological results have been described elsewhere (Greene, Levine, Breitschwerdt, Walker, Berkhoff, Cullen, and Nicholson, in press). Briefly, infected nymphal ticks were allowed to feed

VOL. 26, 1988 on the skin of one group of four dogs. After 56 days, when none of these dogs seroconverted, they were inoculated intravenously with 3 x 108 spirochetes. The isolate of B. burgdorferi was obtained by dissection of a nymphal Ixodes dammini tick. The second group of four dogs initially received a subcutaneous inoculation of 500 organisms. After 56 days, when only two of these dogs had seroconverted, they were inoculated intraperitoneally with 3 x 108 spirochetes. The sera used in this study were obtained from these dogs 15 to 30 days after the second exposure. The IFA titers were.1:512, and the ELISA values (a ratio that relates the absorbance of the test serum with that of positive and negative control sera, such that the ELISA values for previously unexposed dogs are often less than 10) were.50.0 (Greene, Levine, Breitschwerdt, Walker, Berkhoff, Cullen, and Nicholson, in press). These dogs remained asymptomatic for Lyme disease for the duration of the study. Naturally exposed animals. Positive serum samples were obtained from symptomatic dogs in endemic areas. The animals were presented to their local veterinarians with the primary complaint of lameness. The diagnosis of Lyme disease was made based on clinical signs of fever and lameness or joint pain and a positive serologic response. The endemic areas included Connecticut (12 dogs), Wisconsin (7 dogs), and Maryland (10 dogs) (14). Sera from these dogs had IFA titers.1:512 and ELISA values 50.0. Positive serum samples from dogs without clinical signs were obtained when serologic surveys were performed on dogs from North Carolina (four dogs), Maryland (six dogs), and Wisconsin (three dogs). All sera from these dogs had IFA titers 1:512 and ELISA values 50.0. Bacterial strains. The bacterial strain used as an antigen source for the immunoblots was B. burgdorferi B31 (ATCC 35210; American Type Culture Collection, Rockville, Md.). In the adsorption studies, Borrelia anserina (provided by H. J. Barnes, North Carolina State University, Raleigh), Leptospira interrogans serovar canicola (provided by C. Willis, Rollins Animal Disease Diagnostic Laboratory, Raleigh, N.C.), and Escherichia coli ATCC 35218 were used. To evaluate the reactivity of sera against the major outer surface protein (OspA), E. coli DHSa (Bethesda Research Laboratories, Inc., Gaithersburg, Md.) containing plasmid ptrh44 [DH5at(pTRH44)] which codes for the production of OspA was used (9). E. coli DH5a without ptrh44 served as a negative control (both strains were provided by A. G. Barbour, University of Texas Health Science Center, San Antonio). Serologic assays. The IFA and ELISA procedures have been described elsewhere (R. T. Greene, J. F. Levine, E. B. Breitschwerdt, and H. A. Berkhoff, Am. J. Vet. Res., in press; Greene, Levine, Breitschwerdt, Walker, Berkhoff, Cullen, and Nicholson, in press). Monoclonal antibodies. Genus (H9724) and species (H5332)specific monoclonal antibodies (MAbs) (provided by A. G. Barbour) and a MAb directed against an outer surface protein (HT5S) (provided by T. G. Schwan, Rocky Mountain Laboratories, Hamilton, Mont.) were used to locate major proteins in the immunoblots. H9724 recognizes a 41kilodalton (kda) flagellar protein, H5332 is directed against the 31kDa OspA protein, and HT5S recognizes the 34kDa major outer surface protein (OspB) (2, 3, 4). Electrophoresis and immunoblotting. Sodium dodecyl sulfatepolyacrylamide gel electrophoresis was performed on wholecell lysates of B. burgdorferi by a modification of methods described by Laemmli (11). For a wholecell lysate, 2weekold cultures of B. burgdorferi were centrifuged at IgG RESPONSE TO B. BURGDORFERI IN DOGS 649 10,000 x g for 30 min at 4 C and washed three times with phosphatebuffered saline (ph 7.38) (Greene, Levine, Breitschwerdt, and Berkhoff, in press; Greene, Levine, Breitschwerdt, Walker, Berkhoff, Cullen, and Nicholson, in press). The pellet was resuspended in enough sterile phosphatebuffered saline that 100,ul of a 1:4 dilution of the suspension had an A410 of 0.165 in a microelisa reader (Dynatech Laboratories, Inc., Alexandria, Va.). Samples (30,ul) were diluted in an equal volume of sample buffer containing 60 mm Tris hydrochloride (ph 6.8) (BioRad Laboratories, Richmond, Calif.), 2% sodium dodecyl sulfate (BioRad), 5% 2pmercaptoethanol, 10% glycerol, and 0.00125% (wt/vol) bromophenol blue. This mixture was heated to 95 C for 5 min. The lysates were electrophoresed in a discontinuous sodium dodecyl sulfatepolyacrylamide gel, using a 4% acrylamide stacking gel and a 12% acrylamide resolving gel. The molecular masses of the proteins were determined by coelectrophoresis of known standards (BioRad). Electrophoresis was performed at a constant current of 35 ma for 3 to 4 h. Proteins were electrophoretically transferred from the unstained gel to a nitrocellulose membrane (18). Briefly, transfer to nitrocellulose membranes was effected electrophoretically in Tris (25 mm)glycine (192 mm) buffer (ph 8.3) at 30 V (100 ma) overnight in a Transblot cell (BioRad). After transfer was complete, 4 to 5mm strips of the nitrocellulose membrane were cut and placed into capped tubes (16 by 150 mm; Fisher Scientific Co., Raleigh, N.C.). The strips were then blocked for 1 h at room temperature (ca. 25 C) by using 3 ml of 20 mm Trisbuffered saline (TBS; ph 7.5) with 3% (wt/vol) gelatin, with constant rocking. The strips were washed three times for 5 min per wash with 0.05% Tween TBS. The strips were incubated for 1 h with constant rocking at 25 C with 2 ml of the test serum diluted 1:50 in 1% gelatintbs (antibody buffer). The wash procedure was repeated, and the strips were incubated for 1 h at room temperature with 2 ml of a 1:500 dilution of horseradish peroxidaseconjugated goat antidog IgG (heavy chain specific) (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) in antibody buffer. The strips were washed twice with TweenTBS and then once with TBS. The nitrocellulose strips were developed by the addition of a freshly prepared solution of 0.05% (wt/vol) 4chloro1naphthol (BioRad), 0.015% hydrogen peroxide, and 20% methanol in TBS. The reaction was stopped by washing the strips with cold distilled water. The top and bottom edges of the strips were used for alignment. When immunoblots were performed with MAbs, 2 ml of a 1:50 dilution of the MAb was incubated similarly to the test serum and 2 ml of a 1:3,000 dilution of horseradish peroxidaseconjugated goat antimouse IgG (heavy and light chain specific) (BioRad) was used as a second antibody. Adsorption. For the adsorption procedures, wholecell suspensions of B. burgdorferi, B. anserina, L. interrogans coli were prepared as described serovar canicola, and E. above for electrophoresis. By using the microelisa reader, suspensions were prepared such that 100 pul of 1:4 dilutions of the suspensions had an A410 of 0.222. To demonstrate the specificity of the reactions, serum from one dog from each group was adsorbed with the B. burgdorferi suspension, and one serum sample was adsorbed with all three organisms. A 0.5ml portion of the bacterial suspension to be used for adsorption was centrifuged at 8,200 x g for 10 min at 25 C. The supernatant was removed, and 10 pul of the test serum diluted with 0.5 ml of antibody buffer was used to resuspend the organisms. This suspension

650 GREENE ET AL. was incubated for 30 min at 37 C. The suspension was then centrifuged at 8,200 x g for 10 min, and the supernatant was used to suspend the new pellet of the same organism. A total of three adsorptions with each organism were performed. After the adsorption steps, immunoblotting was performed with the final supernatant, as described above. The 0.5 ml of diluted, adsorbed serum was added to 1.5 ml of antibody buffer to make a total of 2 ml of a 1:50 dilution of the serum sample, as described above for the immunoblot procedure. As a control, the test serum was processed by the adsorption procedure with the adsorption antigen excluded. Use of recombinants to evaluate OspA reactivity. Recommendations for bacterial growth medium and procedures for preparation of recombinant cells were made by A. G. Barbour (personal communication). The E. coli recombinant DH5a(pTRH44) was grown overnight at 37 C in 10 ml of broth containing 1% Casamino Acids (Difco Laboratories, Detroit, Mich.), 0.5% yeast extract (Difco), 52 mm NaCl, 26 mm KCl, 20 mm HEPES (N2hydroxyethylpiperazineN' 2ethanesulfonic acid) (Sigma Chemical Co., St Louis, Mo.), 10 mm MgCl2, and 50,ug of sodium ampicillin per ml. The broth inoculated with the control strain E. coli DHSa did not contain ampicillin. After overnight growth, 1.5ml aliquots were centrifuged at 8,200 x g for 10 min at 25 C. The pellets were suspended in 1 ml of phosphatebuffered saline. The washed cells were then centrifuged as described above, and the supernatant was discarded. The pellets were stored at 100 C for 15 min and then suspended in 200 pl of 2 x electrophoresis sample buffer. The sample was boiled for 5 min, allowed to cool, and then centrifuged at 8,200 x g for 10 min at 25 C. The supernatant was loaded into 4cmwide wells for electrophoresis. Electrophoresis and transfer to nitrocellulose was performed as described above. Before immunoblotting, the positive sera from experimentally and naturally exposed dogs were adsorbed with E. coli DH5a. The concentration of the adsorption antigen and the procedure for the adsorption are described above. RESULTS When sera from naturally exposed dogs were adsorbed with B. burgdorferi, reactivity to most protein bands was eliminated (Fig. 1). At areas corresponding to some bands which reacted strongly with unadsorbed sera, faint reactivity could be seen after adsorption. When B. anserina, L. interrogans serovar canicola, and E. coli were used, there was little to no reduction in the reactivity to any bands (Fig. 1). The immunoblots obtained with sera from naturally exposed dogs from the different areas of the country were similar (Fig. 2 and Table 1). Most sera recognized at least 15 protein bands (Fig. 2). There were approximately 10 polypeptide bands that sera from most dogs recognized strongly (major bands), and an additional 7 to 10 bands showed weak reactivity (minor bands). There were some major variations among dogs, but for a geographical area, all protein bands were adequately represented (Table 1). The major bands recognized by sera from most dogs were approximately 83, 66, 61 to 60, 41 to 39, 31 to 29, 17, and 15 kda. The minor bands corresponded to 90, 75, 50, 48, 34, 27, 24, 22, and 19 kda. There were no apparent differences in serum reactivity between naturally exposed dogs that were exhibiting clinical signs and those that were not (Fig. 2 and 3 and Table 1). All of the sera from the experimentally exposed dogs reacted similarly. In contrast to the naturally exposed dogs, the experimentally exposed dogs had different immunoblot pat j~~~~~~45. J. CLIN. MICROBIOL. terns (Fig. 2 and 3). The sera from the experimentally exposed dogs reacted against a limited number of antigens, often less than 6, whereas the sera from the naturally exposed dogs often reacted with at least 15 protein bands. One major antigen recognized by the sera from the experimentally exposed dogs was a broad band in the 31kDa region (Fig. 2 and 3). The speciesspecific MAb reacted to this protein in a similar manner (Fig. 3). The sera from naturally exposed dogs reacted against a narrower band located near the bottom of this region (Fig. 2 and 3). In addition, sera from the experimentally exposed dogs consistently reacted against a 34kDa band which was only occasionally recognized by the sera from naturally exposed dogs. The reactivity of MAb H5332 in the immunoblot obtained with the recombinant DH5a(pTRH44) demonstrated the OspA band at approximately 31 kda (Fig. 4). No reactivity was seen when H5332 was reacted with DH5a. A thin band in the OspA region was present when sera from experimentally exposed dogs were reacted with DH5a(pTRH44). No corresponding band was seen when the same sera were reacted with DH5a. When serum from a naturally exposed dog was reacted with DH5a(pTRH44), no band was seen in the OspA region (Fig. 4). There was a dark band at approximately 66 kda in the immunoblot with DH5a. A corresponding band, but much lighter, was seen when the serum was reacted with DH5a(pTRH44). DISCUSSION Diagnosis of canine borreliosis is often based on positive serologic response in conjunction with compatible clinical 1 2 3 4 5 Mws " 92.5 66.2 X= x:: ~ ~ 1. Ifs.`A: :. c;[ 4 *: w. Of 21.5 14.5 FIG. 1. Immunoblot analysis of wholecell suspensions of B. burgdorferi after adsorption of sera with organisms. Lanes: 1, no adsorption; 2, adsorbed with B. burgdorferi; 3, adsorbed with B. anserina; 4, adsorbed with L. interrogans serovar canicola; 5, adsorbed with E. coli. MWS, Molecular weight standards (molecular weights in thousands).

::r...v..., s. Ri E :>``.: E:, = ::<:t. w,..x:.,.':: Z,'; W zz, VOL. 26, 1988 IgG RESPONSE TO B. BURGDORFERI IN DOGS 651 No Clinical Signs r= With Clinical Signs EX NC wl W I CT M D a b C d e f k m n o q r s MW1S :: S M... C I Jg,2. s:«i «., um~ ~~~~~. ti ~!amom..... 9 s. :xz to si«si } : m Wws,ç, ^ ".: n ::: w 2i:vv` 1!11 Il}` sx 0s;.. + *:::4:. 45.0.`` 444^ r ' s 2,»:.».S. "' ;X 3 8 ::: 111E.,...«e<s s w ma 7... 2i W.. iei 11 w,k 3o t@f e.i31 *idgub m. r. S 2.'.e.. ':.' ''.`: `2 :es w =.. es... ;.5^.` :, e: ««.... :`;J :::: :i..... W..:.Y `.1 i........ :iâ ::::,l'... ',':"',:,,,.. :,.: z *19 d d".. 2 l z I..: ẕ Mf àrfil' X. M. :w::t 14.5 FIG. 2. Immunoblots of wholecell suspensions of B. burgdorferi obtained with canine sera. Representative blots from each group are shown. EX, experimentally infected dogs; NC, dogs from North Carolina; WI, dogs from Wisconsin; CT, dogs from Connecticut; MD, dogs from Maryland. Lowercase letters represent sera from different dogs. MWS, Molecular weight standards (molecular weights in thousands). Clinical signs included fever, lameness, and joint pain. <.. <:. Downloaded from http://jcm.asm.org/ signs (8, 10, 12, 13). However, elevated IgG antibody titers can be found in asymptomatic dogs (6, 10, 13). There is considerable overlap in the magnitude of the serologic titer between the two groups (10, 13). Adsorption studies confirmed that the antibody response in naturally exposed dogs was specific for B. burgdorferi (Fig. 1). The sampling of the serum of one dog from each region was considered sufficient to determine that the response was specific. Adsorptions performed with B. anserina, L. interrogans serovar canicola, and E. coli revealed minimal antigenic crossreactions. Immunoblots obtained with human sera have shown an expansion of the immune response to antigens of B. burgdorferi as the disease progresses (7). It was considered a possibility that differences in immunoblot profiles may be useful for differentiation of symptomatic and asymptomatic, serologically positive dogs, but the results of the present study revealed no consistent differences. It is important to recognize that asymptomatic dogs were not monitored over a long period. It is possible that at a later time these animals might develop clinical signs consistent with Lyme disease. Similarly, many other causes of lameness, such as degenerative arthritis, infectious arthritis, idiopathic polyarthritis, and myopathies, may not have been adequately ruled out in cases diagnosed as Lyme disease. Although there may be spirochete strain differences in various areas of the country (3), the immunoblots obtained with the canine sera tested were similar for a geographical region. The antigenic reactivity in naturally infected dogs was similar to that observed in immunoblot profiles of humans with chronic Lyme disease (1, 7). Humans with early or mild lesions, (erythema chronicum migrans only) develop higher IgM than IgG activity (7, 16). Of 12 patients with this early lesion, only 3 developed IgG responses, which were limited to the 41kDa polypeptide region (7). This response was no longer present several months later. In chronic borreliosis, the human IgG response has been sequentially characterized, showing that sera first bind the 41kDa antigen during erythema chronicum migrans. Subsequently, in 1 to 5 months a response to the 83, 66, 27, and 15kDa antigens may develop. If the disease is untreated, months to years later mildtomoderate responses in the 75 and 60kDa regions develop, whereas strong reactivities are seen in the 34, 31, 29, and 17kDa regions. This late response is very similar to the immunologic response of the naturally exposed dogs in the present study and suggests that the dogs were chronically exposed. This may have been a result of repeated exposures to infected ticks or persistence of the infection in the dogs. Without a marker for the early signs of the disease in dogs, it is not possible to determine whether temporal changes in the immunoblot patterns parallel the human response. Sera from experimentally exposed dogs had different immunoblot patterns than those from naturally exposed dogs. The response of the experimentally exposed dogs was to a limited number of protein bands. The primary protein bands recognized corresponded to the major outer surface proteins (OspA and OspB). In humans, antibodies are not produced against these proteins until late in the disease. This on May 4, 2018 by guest

652 GREENE ET AL. J. CLIN. MICROBIOL. TABLE 1. Percentage of sera from experimentally and naturally exposed dogs reacting to specific B. burgdorferi bands in the immunoblots % of sera reacting from dogs with: Band (kda) No clinical signs Clinical signs EX (n = 8) NC (n = 4) WI (n = 3) MD (n = 6) WI (n = 7) CT (n = 12) MD (n = 10) 90 0 0 33 67 86 71 80 83 100 100 100 100 100 100 100 75 0 75 67 67 100 86 100 72 0 75 67 33 71 29 80 66 0 100 100 100 71 86 60 61 0 100 100 100 86 93 100 60 0 100 100 100 86 93 90 50 25 100 67 33 100 57 100 48 36 100 67 67 86 79 100 45 25 25 0 33 86 71 20 41 75 100 100 100 100 100 100 39 100 100 100 100 71 100 80 34 100 50 67 67 86 57 30 31 100 100 100 100 100 100 100 29 0 100 100 100 100 100 100 27 0 50 33 0 43 64 40 24 0 50 67 67 71 64 70 22 63 75 33 67 43 57 30 19 0 0 33 67 29 29 60 17 50 50 100 67 100 100 100 15 50 50 100 100 71 100 100 a Either experimentally exposed dogs (EX) or naturally exposed dogs from North Carolina (NC), Wisconsin (WI), Maryland (MD), or Connecticut (CT) were used. disparity could be related to the pathogenesis of the spirochete or to the method of experimental inoculation. In natural Lyme disease, it is thought that these outer membrane proteins are initially masked from the immune system, possibly by a glycocalyx (7). It has been postulated that the organism must be multiplying in the host for these major proteins to be recognized. In preparing the inoculum for experimental infection, the organisms were washed three times in phosphatebuffered saline. This might have removed the glycocalyx. In addition, the spirochetes were artificially inoculated as a bolus. Therefore, antigen presentation may explain why inoculation caused antibody production which was quite different from that which occurs under natural conditions. The serologic pattern of reactivity in the region of the 31kDa protein differed between the experimentally and naturally exposed dogs. For the experimentally exposed dogs, the antibody reactivity appeared as a wide band in the 31kDa region, compared with only a thin band near the leading edge of this region for the naturally exposed dogs (Fig. 2 and 3). An E. coli recombinant which produces OspA was used to evaluate whether the reactivity was to OspA in both instances. The band observed on the immunoblot when sera from experimentally exposed dogs were reacted against DH5a(pTRH44) confirmed that the antibodies were against OspA. The lack of a band when sera from naturally exposed dogs were used suggests that the band in the 31kDa region was directed against another protein of similar molecular mass (Fig. 4). The reason that presumably chronically exposed dogs failed to develop antibodies against a major surface protein is unknown. The dark band seen at approximately 66 kda when the sera from naturally exposed dogs were reacted with DH5a was considered to represent a nonspecific reaction since it was present in DH5a both with and without ptrh44. The difference in the immunoblot patterns in the 34kDa region for experimentally versus naturally exposed dogs is of G(e SP Ex As Sy MWS!:p 925 66.f2 45.0.21.5 14.5 FIG. 3. Immunoblots of wholecell suspensions of B. burgdorferi obtained with genus (H9724) and species (H5332)specific monoclonal antibodies (provided by A. G. Barbour) and canine sera. Lanes: Ge, genus specific; Sp, species specific; Ex, experimentally exposed; As, asymptomatic, naturally exposed; Sy, symptomatic, naturally exposed. MWS, Molecular weight standards (molecular weights in thousands).

VOL. 26, 1988 *e.*^, `: s. a. I, X l : fi M... :.' 2s. S ssa l, E s. i DHScl r}}dsa ptré44 1 2 3 1 2 3.:?... $, b',x, ',...... :.: *;wr` E j''es' ``g`:':.:sex w E.. X,Es s, im#i X M iig? IllGk E.. W?jN R 0 lt #É Fs :....! LG ffi xss Ms. S fi. l e e A :\ Usp " s si e:. 3E..,E S.... s s t s 3 as..... l. s `.: :`:.. : sf :... :.... `. il ] se..... 1 11!!.l FIG. 4. Immunoblot analysis using E. coli DH5cx and recombinant DH5at(pTRH44) (both provided by A. G. Barbour) to detect the presence of antîbodies to OspA. Lanes: 1, speciesspecific monoclonal antibody (H5332), 2, sera from experimentally exposed dogs; 3, sera from naturally exposed dogs. The OspA region is indicated. unknown significance. Sera from ail of the experimentally exposed dogs reacted in this region, whereas the protein was recognized only occasionally by sera from naturally exposed dogs. The 34kDa protein is a major surface protein which may confer serotype specificity to B. burgdorferi (3, 15). Reactivity with HT5S confirmed that this protein is present in the isolate used in the experimental inoculations (data not shown). Therefore, reactivity to this protein may be a reflection of mode of exposure, exposure to different strains of spirochetes, or possibly a response that is temporally related to the disease course. In humans, reactivity to this protein has been reported to develop over time (7). In summary, there was considerable variation in the immunoblot patterns when sera from naturally exposed dogs were reacted against B. burgdorferi. The variation may have been due to temporal differences during infection, spirochete strain variability, or variations in the immune responses of the individual dogs. Within and among geographic regions, the immunoblot patterns were similar. Immunoblots were significantly different for naturally and experimentally exposed dogs. Variation in antigen clearance, route of exposure, frequency of exposure, or duration of infection is postulated as a possible cause of the differences. Sera from naturally exposed dogs did not recognize the major outér surface protein (OspA), although based on immunoblot patterns, it appears that the dogs were often chronically infected or had had repeated exposures. The exact reason for this failure to react with OspA is unknown. Finally, there were no major differences in immunoblot patterns for dogs that were showing clinical signs compatible with Lyme disease and those that were not. It is possible that the asymptomatic dogs used in this study might later develop IgG RESPONSE TO B. BURGDORFERI IN DOGS 653 clinical signs or that symptomatic dogs did not have Lyme disease. Importantly, a positive humoral response in a dog may not be sufficient to differentiate a dog with Lyme disease from one that merely has been exposed to B. burgdorferi. ACKNOWLEDGMENT Resources for this project were provided by the state of North Carolina. LITERATURE CITED 1. Barbour, A. G., W. Burgdorfer, E. Grunwaldt, A. C. Steere. 1983. Antibodies of patients with Lyme disease to components of the Ixodes dammini spirochete. J. Clin. Invest. 72:504515. 2. Barbour, A. G., S. F. Hayes, R. A. Heiland, M. E. Schrumpf, and S. L. Tessier. 1986. A Borreliaspecific monoclonal antibody binds to a flagellar epitope. Infect. Immun. 52:549554. 3. Barbour, A. G., S. L. Tessier, and S. F. Hayes. 1984. Variation in a major surface protein of Lyme disease spirochetes. Infect. Immun. 45:94100. 4. Barbour, A. G., S. L. Tessier, and W. J. Todd. 1983. Lyme disease spirochetes and ixodid tick spirochetes share a common surface antigenic determinant defined by a monoclonal antibody. Infect. Immun. 41:795804. 5. Burgess, E. C. 1986. Experimental infection of dogs with Borrelia burgdorferi. Zentr`albl. Bakteriol. Hyg. Reihe A 263:4954. 6. Burgess, E. C. 1986. Natural exposure of Wisconsin dogs to the Lyme disease spirochete (Borrelia burgdorferi). Lab. Anim. Sci. 36:288290. 7. Craft, J. E., D. K. Fischer, G. T. Shimamoto, and A. C. Steere. 1986. Antigens of Borrelia burgdorferi recognized during Lyme disease: appearance of a new immunoglobulin M response and expansion of the immunoglobulin G response late inthe illness. J. Clin. Invest. 78:934939. 8. Eugster, A. K., and A. B. Angulo. 1986. Lyme disease. Southwest. Vet. 37:2225. 9. Howe, T. R., F. W. LaQuier, and A. G. Barbour. 1986. Organization of genes encoding two outer membrane proteins of the Lyme disease agent Borrelia burgdorferi within a single transcriptional unit. Infect. Immun. 54:207212. 10. Kornblatt, A. N., P. H. Urband, and A. C. Steere. 1985. Arthritis caused by Borrelia burgdorferi in dogs. J. Am. Vet. Med. Assoc. 186:960964. 11. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680685. 12. Lissman, B. A., E. M. Bosler, H. Camay, B. G. Ormiston, and J. L. Benach. 1984. Spirocheteassociated arthritis (Lyme disease) in a dog. J. Am. Vet. Med. Assoc. 185:219220. 13. Magnarelli, L. A., J. F. Anderson, A. F. Kaufmann, L. L. Lieberman, and G. D. Whitney. 1985. Borreliosis in dogs frorn southern Connecticut. J. Am. Vet. Med. Assoc. 186:955959. 14. Schmid, G. P. 1985. The global distribution of Lyme disease. Rev. Infect. Dis. 7:4148. 15. Schwan, T. G., and W. Burgdorfer. 1987. Antigenic changes of B. burgdorferi as a result of in vitro cultivation. J. Infect. Dis. 156:852853. 16. Steere A. C., R. L.. Grodzicki, A. N. Kornblatt, J. E. Craft, A. G. Barbour, W. Burgdorfer, G. P. Schmid, E. Johnson, and S. E. Malawista. 1983. The spirochetal etiology of Lyme disease. N. Engl. J. Med. 308:733740. 17. Steere, A. C., S. E. Malawista, N. H. Bartenhagen, P. N. Spieler, J. H. Newman, D. W. Rahn, G. J. Hutchinson, J. Green, D. R. Snydman, and E. Taylor. 1984. The clinical spectrum and treatment of Lyme disease. Yale J. Biol. Med. 57:453461. 18. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76:43504354.