INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Arthritis is the most frequent and the major complication of tick-borne transmission of Borrelia burgdorferi sensu stricto (B. burgdorferi) (31). Approximately 60% of individuals develop intermittent episodes of arthritis several weeks or months after infection. The brief attacks of arthritis last several days or weeks and generally occur in the larger joints (31, 32). In addition, 10% of arthritogenic patients develop antibiotic-resistant Lyme arthritis (12, 14), which can lead to permanent joint dysfunction (31). Infection with B. burgdorferi also causes moderate to severe arthritis in dogs (2), hamsters (13, 26), mice (4, 25), monkeys (3), and rats (5).

Recently, we showed that vaccination of hamsters with a whole-cell preparation of B. burgdorferi also induced arthritis, specifically severe destructive Lyme arthritis, following challenge with B. burgdorferi (18). Inflammation or swelling in the hind paws of vaccinated hamsters was detected 7 days after infection, peaked on day 10, and gradually decreased. A chronic synovitis characterized by hypertrophic villi, focal erosion of articular cartilage, and a subsynovial mononuclear infiltrate persisted for approximately 1 year. These findings demonstrate that B. burgdorferi possesses antigenic components that can induce arthritis in naturally infected humans (31, 32) and experimentally infected animals (2-5, 13, 25, 26).

Most importantly, some of these antigenic components are feasible candidates for use as a vaccine against infection with B. burgdorferi sensu lato (9, 11, 20, 24). The most promising candidate has been outer surface protein A (OspA) (29, 30). Recently, the Food and Drug Administration (FDA) approved the use of OspA for vaccination of humans despite indirect evidence and concerns that OspA is associated with arthritis (1, 12, 29, 30). In this study, we present direct evidence that vaccination with two preparations of recombinant OspA (rOspA) can induce severe destructive arthritis in hamsters after challenge with the Lyme borreliosis spirochete.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Hamsters. Twelve- to 16-week-old inbred LSH hamsters were obtained from our breeding colony located at the Wisconsin State Laboratory of Hygiene. Hamsters weighing 100 to 150 g were housed three or four per cage at an ambient temperature of 21°C. Food and water were provided ad libitum.

Organisms. Low-passage (<10) B. burgdorferi isolates 297 (from human spinal fluid), S-1-10 (from Ixodes scapularis), and C-1-11 (also from I. scapularis) were grown at 32°C in modified Barbour-Stoenner-Kelly (BSK) medium (6) until reaching a concentration of approximately 10⁷ spirochetes per milliliter. Five-hundred-microliter samples were then dispensed into 1.5-ml screw-cap tubes (Sarstedt, Newton, N.C.) containing 500 µl of BSK medium supplemented with 20% glycerol (Sigma, St. Louis, Mo.), and the tubes were sealed and stored at 70°C. When needed, a frozen suspension of spirochetes was thawed and an aliquot was used to inoculate 4 ml of fresh BSK medium. Spirochetes were enumerated by dark-field microscopy, using a Petroff-Hausser counting chamber. Escherichia coli DH5 (Gibco BRL, Gaithersburg, Md.) was used for cloning experiments.

Amplification and cloning of the ospA gene. Plasmid-enriched DNA was isolated from B. burgdorferi isolate S-1-10 as previously described (20). The DNA was used as a template for the amplification of the ospA gene (GeneAmp; Perkin-Elmer Cetus, Norwalk, Conn.). The amino-terminal primer B1 (5'GCGTGGATCCATGAAAAAATATTTATTGGGAA3') and the carboxy-terminal B2 (5'AATTCCCGGGTTATTTTAAAGCGTTTTTAA3') were used for amplification. Primers were each used at a final concentration of 1.0 µM with an MgCl₂ concentration of 2.5 mM. Thermal cycling parameters were 94°C for 60 s followed by 35 cycles of (i) 94°C for 60 s, (ii) a 2-min ramp to 45°C, (iii) 45°C for 60 s, (iv) a 60-s ramp to 60°C, and (v) 60°C for 6 min. The final extension was done at 60°C for 10 min to fully extend any truncated DNA strands. Amplified DNA was purified with GeneClean (Bio 101, La Jolla, Calif.). After digestion with SmaI and BamHI (Gibco BRL), purified DNA fragments were ligated into pGEX-2T (Pharmacia Biotech, Piscataway, N.J.). The insert and plasmid were ligated with T4 DNA ligase (Gibco BRL), and the ligation mix was used to transform competent E. coli DH5. Transformed E. coli cells were then plated onto 2× tryptone-yeast extract agar medium containing ampicillin (100 µg/ml; Sigma). Colonies expressing rOspA protein were identified by Western blot analysis using B. burgdorferi isolate B31 OspA monoclonal antibody H53332, provided by A. G. Barbour.

rOspA expression and purification. The transformed E. coli organisms containing the ospA gene were grown for 12 h at 37°C in 100 ml of 2× tryptone-yeast extract broth containing 100 µg of ampicillin per ml. Cultures were diluted 1:10 with broth medium and incubated for an additional 1 h. Isopropyl--D-thiogalactopyranoside (final concentration, 0.1 mM) was added, and the culture was incubated for 5 h. After incubation, the suspension of bacteria was centrifuged, resuspended in phosphate-buffered saline (PBS; pH 7.4), and lysed by three 30-s pulses with a sonicator (model W-350; Branson Sonic Power Co., Danbury, Conn.). Sonicated E. coli organisms were mixed with Triton X-100 (10%), diluted 10-fold with PBS, and centrifuged to remove insoluble material. The supernatant was mixed with a 50% slurry of glutathione-Sepharose beads (Pharmacia Biotech) for 5 min at room temperature and washed three times with ice-cold PBS. Fusion proteins were eluted by mixing the beads with 1 ml of 50 mM Tris-HCl (pH 8.0) containing 5 mM reduced glutathione for 2 min and collected after centrifugation for 60 s at 500 × g. The elution procedure was repeated four times. Fractions were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting with specific hamster serum, goat anti-glutathione S-transferase polyclonal antibody (Pharmacia Biotech), or monoclonal antibody H5332. Finally, fractions were concentrated by using a Centriprep-10 concentrator (Amicon, Beverly, Mass.) and dialyzed against PBS (pH 7.2) at 4°C overnight, using a dialysis cassette with a 10-kDa-molecular-mass cutoff (Pierce Chemical Co., Rockford, Ill.).

Vaccination of hamsters. Three groups of eight hamsters each were mildly anesthetized with ether contained in a nose-and-mouth cup and then vaccinated intramuscularly in each hind thigh with 0.25 ml containing 15, 30, or 60 µg of nonlipidated rOspA adsorbed to 1% aluminum hydroxide gel (Reheis Inc., Berkley Heights, N.J.) in PBS. Each hamster received a total of 30, 60, or 120 µg of rOspA. In addition, 20 hamsters were vaccinated in each hind thigh with 0.5 ml of a commercially available canine rOspA vaccine (with or without aluminum hydroxide) (Merial, Athens, Ga.). Controls consisted of nonvaccinated hamsters, hamsters inoculated with 1% aluminum hydroxide gel, hamsters vaccinated with 30, 60, or 120 µg of rOspA adsorbed to 1% aluminum hydroxide gel, and hamsters vaccinated with the canine rOspA.

Infection of hamsters. rOspA- and canine rOspA-vaccinated hamsters were mildly anesthetized with ether and challenged subcutaneously in each hind paw with 0.2 ml of BSK containing 5 × 10⁶ viable B. burgdorferi isolate 297 or C-1-11 organisms. In some studies, hamsters inoculated with B. burgdorferi isolate 297 were reinfected with a similar inoculum 24 h after the initial infection. Nonvaccinated hamsters and those vaccinated with 1% aluminum hydroxide gel served as controls; each group was challenged with 10⁷ viable spirochetes in BSK. Arthritis was not induced in hamsters vaccinated with aluminum hydroxide alone or with E. coli or Staphylococcus epidermidis in alum and then challenged with B. burgdorferi (8).

Assessment of arthritis. The degree of hind-paw swelling was used as an index to evaluate the inflammatory response. Prior to experimentation and cage assignment, hamsters were randomly chosen and their hind paws were measured to establish a baseline. After infection, the hind paws were measured periodically for 20 days with a dial-type Vernifer caliper (Fisher Scientific, Pittsburgh, Pa.) graduated in 0.1-cm increments. Measurements were obtained by mildly anesthetizing each hamster and carefully measuring the width and thickness of each hind paw. The daily mean group value was calculated by dividing the sum of the caliper values of each hind paw by the number of hind paws per group. This average value represented the severity of hind-paw swelling. Detection of arthritis by measurement of hind-paw swelling with a caliper is less variable when hamsters are challenged with 10⁶ spirochetes or more; histopathologic examination is needed to confirm arthritis when hamsters are inoculated with 10² to 10⁵ spirochetes.

Preparation of tissues for histologic examination. Twenty-one days after infection, hamsters were euthanized and their hind legs were amputated at mid-femur, fixed in 10% neutral buffered formalin, placed in decalcifying solution (Lerner Laboratories, Pittsburgh, Pa.) for 24 h, and stored in 10% zinc formalin prior to processing. The hind legs were bisected longitudinally, placed in embedding cassettes (Fisher Scientific), embedded in paraffin, and cut into 6-mm-long sections. The sections were then placed on glass slides and stained with hematoxylin and eosin. The hind legs were randomly selected and cryptically coded for unbiased histopathologic examination by a certified pathologist.

Statistical analyses. The mean caliper values among groups were tested by analysis of variance with Minitab statistical analysis software. The alpha level was set at 0.05 before the experiments were started. The standard error of the mean for each mean caliper group value was also calculated.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Ability of rOspA vaccination to induce arthritis. Hamsters were vaccinated with 30, 60, or 120 µg of rOspA and challenged with B. burgdorferi isolate 297 at 11 and 12 days after vaccination. In addition, seven hamsters were vaccinated with a commercial canine rOspA vaccine and infected with B. burgdorferi isolate 297 (Table 1). Severe swelling of the hind paws was detected in 100, 100, 50, and 57% of hamsters vaccinated with 30, 60, or 120 µg of rOspA or the rOspA canine vaccine, respectively. Although slight swelling of the hind paws (mean ± standard error at baseline, 0.64 ± 0.05) occurred in nonvaccinated hamsters challenged with B. burgdorferi isolate 297, the degree of swelling was considerably lower than that (range, 0.91 ± 0.03 to 0.97 ± 0.04) detected in hamsters vaccinated with 30 or 60 µg of rOspA and challenged with B. burgdorferi isolate 297. When eight hamsters were vaccinated with 120 µg of rOspA and challenged with another B. burgdorferi isolate, C-1-11, that was not vaccine specific, all of the rOspA-vaccinated hamsters developed severe swelling of the hind paws. Similarly, hamsters vaccinated with 30 or 60 µg of rOspA developed severe swelling after challenge with B. burgdorferi isolate C-1-11. Furthermore, severe swelling of the hind paws developed in all hamsters vaccinated with the canine rOspA vaccine mixed with aluminum hydroxide and challenged with B. burgdorferi isolate 297. When these experiments were repeated, similar results were obtained.

Development of arthritis in rOspA-vaccinated hamsters. Two groups of three hamsters each were vaccinated with 30 µg of rOspA (Fig. 1). Eleven and 12 days after vaccination, members of one group of vaccinated hamsters were challenged subcutaneously in the hind paws with 10⁷ viable B. burgdorferi isolate 297 organisms. Swelling of the hind paws was detected 7 days after primary challenge; increased rapidly, with peak swelling occurring on day 11; and gradually decreased. No swelling of the hind paws was detected in nonchallenged hamsters vaccinated with 30 µg of rOspA. Although swelling of the hind paws was detected in nonvaccinated hamsters challenged with B. burgdorferi isolate 297, the severity of swelling was considerably less than that detected in rOspA-vaccinated hamsters challenged with B. burgdorferi isolate 297. No swelling of the hind paws was detected in nonvaccinated, nonchallenged hamsters.

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Development of swelling in the hind paws of rOspA (30 µg)-vaccinated () and nonvaccinated (---) hamsters with () and without () challenge with B. burgdorferi isolate 297.

Histopathology of hind-paw swelling. OspA-vaccinated hamsters challenged with B. burgdorferi isolate 297 showed a diffuse swelling of the hind paws secondary to fibroinflammatory and edematous changes of the soft tissue and joint capsule (Fig. 2A). Prominent focal tenosynovitis with subsynovial inflammation and early pannus formation (Fig. 2A) was also present. The pannus formation encroached on the periphery of the joint, causing osteoclastic reabsorption of bone as well as separation and fragmentation of the subchondrial bone of the joint. Proximal to the tibiotarsal joint, the inflammation showed further encroachment, with compression and atrophy of the bone, producing pyknosis and degeneration of osteocytic nuclei (Fig. 2B). In addition to bone erosion and distortion of the joint, a lymphoplasmacytic infiltrate admixed with a few neutrophils revealed involvement of the tendons of the hind paws. The intertarsal joints showed less inflammation. No granulomata, vasculitis, osteophytes, or loose bodies of the joints were found. By contrast, OspA-vaccinated (Fig. 2D) and nonvaccinated (Fig. 2E) hamsters not challenged with B. burgdorferi showed intact joints and normal capsular and pericapsular soft tissue. Nonvaccinated hamsters challenged with B. burgdorferi isolate 297 showed only mild soft-tissue swelling and mild tenosynovitis (Fig. 2C).

View larger version (147K): [in this window] [in a new window]   FIG. 2.   Histopathology of the hind paws of rOspA-vaccinated (A and B) and nonvaccinated (C) hamsters 21 days after challenge with B. burgdorferi isolate 297. Controls included rOspA-vaccinated (D) and nonvaccinated (E) hamsters.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Public health concerns about the morbidity associated with Lyme borreliosis have stimulated efforts to develop an effective vaccine. Several B. burgdorferi sensu lato proteins, OspA (9, 20, 24), OspB (9, 24), OspC (11, 23, 24), and the 39-kDa protein (27), are capable of inducing a protective antibody response. Of these, OspA has emerged as the leading Lyme borreliosis vaccine candidate (29, 30). Two Lyme borreliosis vaccines based on rOspA have been shown to be protective in recent human clinical trials (29, 30). In addition, an rOspA vaccine has been approved by the USDA for use in dogs. Undoubtedly, these vaccines will be widely used, particularly in regions in which Lyme borreliosis is endemic, such as the upper midwestern and northeastern United States (7).

Although the Lyme borreliosis vaccines developed to date have been reported to be safe (17), there are concerns that rOspA might induce adverse effects, such as arthritis (5, 22, 29, 30). Akin et al. (1) showed that the level of anti-OspA immunoglobulin G, especially that specific to the C-terminal epitope of OspA, correlated with maximum arthritis in naturally infected patients. In addition, the cellular immune response to OspA was elevated in genetically susceptible persons, particularly those with HLA-DR4 specificity (14). These patients also had persistent arthritis despite treatment with antimicrobial agents. Furthermore, Gross et al. (12) identified an immunodominant epitope of OspA for T cells that might be responsible for the induction of treatment-resistant Lyme arthritis. Collectively, these findings suggest that OspA is involved in the induction of arthritis in patients infected with B. burgdorferi sensu lato.

In this study, we provided direct evidence that rOspA can induce arthritis. Hamsters vaccinated with rOspA in aluminum hydroxide (alum) developed swelling of the hind paws after infection with B. burgdorferi isolate 297 or C-1-11. Arthritis was detected in the hind paws of all hamsters vaccinated with 30 or 60 µg of rOspA. Histopathologic examination of the swollen hind paws confirmed the development of severe destructive arthritis. In addition, we showed that a canine rOspA vaccine primed (vaccinated) hamsters for induction of arthritis upon challenge with B. burgdorferi isolate 297. Fifty-seven percent of infected, canine rOspA-vaccinated hamsters developed arthritis. Furthermore, when aluminum hydroxide was incorporated into the canine rOspA vaccine, all hamsters developed arthritis after infection with B. burgdorferi isolate 297. These results show that different preparations of rOspA can induce arthritis and that aluminum hydroxide augments the adverse response. The FDA-approved rOspA vaccine for humans contains aluminum hydroxide (30).

In other studies, 50 and 100% of hamsters vaccinated with 120 µg of rOspA developed severe destructive arthritis when challenged with the infectious vaccine-specific isolate of B. burgdorferi or another isolate of B. burgdorferi (C-1-11), respectively. Previously, we showed that humans vaccinated with 30 µg of rOspA and a booster elicited a poor anti-OspA protective borreliacidal antibody response (22) not only against the vaccine-specific agent but also against other isolates of B. burgdorferi sensu lato. In addition, the anti-OspA borreliacidal antibody titer waned rapidly after vaccination. Although Sigal et al. (29) and Steere et al. (30) demonstrated that rOspA was protective in human field trials, neither the level of the anti-OspA borreliacidal antibody response nor its duration of protection against B. burgdorferi isolates was reported. Lim et al. (18) showed that vaccinated hamsters developed severe destructive arthritis before protective borreliacidal antibodies developed and after they waned when challenged with B. burgdorferi or other isolates. Our results and those of Lim et al. (18) and Padilla et al. (22) suggest that rOspA primes subjects for induction of arthritis without inducing sustained high levels of anti-OspA borreliacidal antibodies. In support of this theory, several boosters of rOspA are required over a 2-year period to obtain 68 to 78% protection against infection with B. burgdorferi (29, 30). Patients received a total of 90 µg of rOspA (30). Additional studies are needed in humans to determine the duration of the borreliacidal antibody response against both the vaccine-specific isolate and other isolates of B. burgdorferi. These studies are necessary for defining the composition of the vaccine (number of rOspA molecules) along with the number and schedule of boosters for maintaining high levels of borreliacidal antibody to prevent potential adverse effects upon challenge with homologous or other isolates of B. burgdorferi.

We used a challenge inoculum of approximately 10⁶ viable B. burgdorferi organisms to elicit severe destructive arthritis in rOspA-vaccinated hamsters. The major histopathologic findings of the joint and capsule, as well as the surrounding soft tissue, resulted in swelling, pain, deformity, and selective loss of movement for the hamster. When vaccinated hamsters were challenged with fewer (10² to 10⁴) B. burgdorferi cells, histopathologic responses that resulted in tenosynovitis were detected. This response in hamsters may be similar to the response that occurs in humans. Although vaccine-induced arthritis after natural infection of humans with B. burgdorferi has not been reported (29, 30), this does not rule out the possibility that rOspA is an arthritogenic agent. Repeated vaccinations of humans with rOspA in alum to maintain protection against infection with B. burgdorferi may increase the number of vaccinees reporting symptoms of arthritis. The present phase III clinical trials did not report sufficient numbers of vaccinees challenged with B. burgdorferi to determine whether rOspA induced arthritis. Human subjects afflicted with rOspA-related tenosynovitis before or after challenge with B. burgdorferi should consult a clinician. These numbers of complaints need to be determined.

The immunologic mechanism(s) by which rOspA or whole cells of B. burgdorferi (18) induce arthritis is incompletely understood. We showed previously that both B. burgdorferi-specific CD4⁺ and CD8⁺ T lymphocytes interacted with macrophages to induce severe destructive arthritis (8). In addition, vaccinated hamsters treated with anti-CD4⁺ antibody failed to develop severe destructive arthritis when infected with B. burgdorferi (19). Other investigators (15, 16, 21) have also reported that T cells and their subsets can exert antagonistic influences on the induction of arthritis. Furthermore, rOspA may induce cross-reactive antibodies that initiate an autoimmune response. OspA has been shown to cause polyclonal activation of B cells (33). These findings indicate that components of the anti-OspA response are T-cell dependent and play a key role in the induction of arthritis. Concomitantly, T-cell-independent responses that result in the production of polyreactive antibodies which cross-react with self-components also occur (10, 28). Evidence, therefore, that several different epitopes of OspA are involved with the production of autoantibodies and protective anti-OspA borreliacidal antibodies and the induction of arthritis is accumulating. The epitopes of rOspA responsible for production of autoantibodies and arthritis must be eliminated before rOspA becomes a successful vaccine.

In conclusion, rOspA vaccination induces severe destructive Lyme arthritis. The present rOspA vaccines must be modified to eliminate potential side effects. The production of a nonarthritogenic rOspA vaccine can be readily determined by using the hamster model.