|Home | About | Journals | Submit | Contact Us | Français|
Development of a high level of sustained borreliacidal antibody is paramount for maintaining protection against infection with Borrelia burgdorferi. We show that production of borreliacidal antibody can be enhanced by preventing the effects of gamma interferon (IFN-γ). When lymph node cells capable of producing borreliacidal antibody were cultured with anti-murine IFN-γ, an eightfold increase in borreliacidal antibody production was obtained. However, anti-IFN-γ treatment of these cells also enhanced their ability to adaptively induce arthritis. When anti-IFN-γ-treated lymph node cells producing borreliacidal antibody were infused into C3H/HeJ mice and the mice were then challenged with B. burgdorferi, the mice developed severe destructive Lyme arthritis. Additional studies are needed to delineate the immune response responsible for the induction of arthritis and production of borreliacidal antibody. These studies are needed to ensure an effective and safe vaccine against infection with B. burgdorferi.
Vaccination with Borrelia burgdorferi or its components can induce antibodies that prevent infection with the Lyme disease spirochete (1, 6, 8-10, 13, 27, 34). These vaccine-induced antibodies kill B. burgdorferi in the presence of complement (17, 24). Specifically, it has been shown that antisera raised against recombinant outer surface protein OspA can prevent borrelial infection by passive immunization (8, 17, 31, 34), kill B. burgdorferi in vitro (13, 17, 18, 20, 24, 27, 29, 30), and sterilize B. burgdorferi-infected ticks (11). However, anti-OspA borreliacidal activity wanes rapidly after active immunization. Padilla et al. (24) reported that vaccination of experimental animals with recombinant OspA induced a borreliacidal antibody response that peaked 6 weeks after vaccination and declined rapidly. Moreover, human volunteers administered several doses of recombinant OspA produced borreliacidal antibody that rapidly decreased (24). Only one vaccinee had antibody 6 months after vaccination. The poor OspA-borreliacidal antibody response may have been a factor in prompting withdrawal of the vaccine for use in humans.
Very little is currently known about the immunologic events that might promote and maintain a sustained borreliacidal antibody response after vaccination. Therefore, we developed an in vitro system to determine the effects of immunological mediators on production of borreliacidal activity. Several cytokines have been evaluated, including interleukin (IL)-4, a known B-lymphocyte stimulator (26), and its antagonist, gamma interferon (IFN-γ) (25). However, these cytokines failed to induce high levels of borreliacidal antibody (21, 22).
Here, we report that treatment of lymph node cells containing borreliacidal antibody-producing cells with anti-murine IFN-γ augments the production of anti-B. burgdorferi antibody. Unfortunately, treatment of these cells with anti-murine IFN-γ also enhances their ability to induce arthritis.
Inbred C3H/HeJ mice 8 to 12 weeks old were obtained from our breeding colony located at the Wisconsin State Laboratory of Hygiene. Mice weighing 20 to 30 g were housed four per cage at an ambient temperature of 21°C. Food and acidified water were provided ad libitum.
B. burgdorferi sensu stricto isolate 297 was originally isolated from human cerebrospinal fluid (35). Low-passage (<6) organisms were cultured once in modified Barbour-Stoenner-Kelly (BSK) medium (2) containing screened lots of bovine serum albumin (5) to a concentration of 5 × 107 spirochetes per ml. Five-hundred-microliter samples were then dispensed into 1.5-ml screw-cap tubes (Sarstedt, Newton, N.C.) containing 500 μl of BSK supplemented with 10% glycerol (Sigma Chemical Co., St. Louis, Mo.), sealed, and stored at −70°C. When necessary, a frozen suspension of spirochetes was thawed and used to inoculate fresh BSK medium. Spirochetes were viewed by dark-field microscopy and enumerated with a Petroff-Hausser counting chamber.
B. burgdorferi organisms were grown in 1 liter of BSK medium for 6 days, pelleted by centrifugation (10,000 × g, 15°C, 10 min), and washed three times with phosphate-buffered saline (PBS; pH 7.4). The washed pellet was resuspended in 1% formalin, incubated at 32°C for 30 min with periodic mixing, washed three times by centrifugation with PBS (12,000 × g, 10°C, 15 min), and resuspended in PBS. Subsequently, the formalin-inactivated spirochetes were mixed in a volume of a 1% suspension of aluminum hydroxide (Reheis, Berkeley Heights, N.J.) to yield 4 × 106 spirochetes/ml.
Forty mice were mildly anesthetized with methoxyflurane contained in a mouth-and-nose cup and vaccinated subcutaneously in the inguinal region with 0.25 ml (≈106 B. burgdorferi) of the formalin-inactivated vaccine preparation. The suspension contained approximately 100 μg of borrelial protein. Nonvaccinated mice were injected with BSK medium or aluminum hydroxide alone.
Three to five nonvaccinated mice per experimental protocol were mildly anesthetized with methoxyflurane contained in a mouth-and-nose cup and injected intraperitoneally with 2 ml of 3% thioglycolate in PBS. Four days after injection, mice were euthanized by CO2 asphyxiation, and 8 ml of cold Hanks' balanced salt solution (Sigma) was injected intraperitoneally. The peritoneal cavity was massaged for ≈1 min, and the exudate cells were recovered by aspiration with a syringe. The suspension of peritoneal exudate cells was centrifuged at 1,500 rpm for 10 min at 4°C. The supernatant was decanted, and the cells were resuspended in Dulbecco's modified Eagle's medium (DMEM; Sigma) that was free of antimicrobial agents but supplemented with 10% heat-inactivated (56°C, 45 min) fetal bovine serum (HyClone Laboratories, Logan, Utah), 5 × 10−5 M 2-mercaptoethanol (Sigma), and l-glutamine (2.92 mg/ml; Sigma). Aliquots of the cell suspension were then poured over polystyrene tissue culture dishes (100 by 20 mm; Corning Glass Works, Corning, N.Y.) and incubated at 37°C in a humidified atmosphere of 5.0% CO2 for 4 to 6 h. After incubation, nonadherent cells were aspirated from the tissue culture dishes.
The dishes were rinsed twice with 8-ml portions of warm Hanks' balanced salt solution to further eliminate nonadherent cells. Five milliliters of cold, nonenzymatic cell lifter (Sigma) was then added to each tissue culture dish and incubated at 4°C for 30 min. Macrophages were detached from the surface of the dishes by vigorously tapping and gently scraping the inside of the tissue culture dishes with a sterile rubber policeman. Suspensions of macrophages from several tissue culture dishes were aspirated, pooled, and centrifuged at 1,500 rpm for 10 min at 4°C. The supernatant was decanted, and the pellet was resuspended in 1 ml of DMEM. Cell viability was determined by trypan blue exclusion. The preparations of macrophages obtained by this method were 98% free of lymphocyte contamination, as determined by flow cytometry.
Twelve to fifteen mice were euthanized by CO2 inhalation 17 days after vaccination with formalin-inactivated B. burgdorferi (22). Inguinal lymph nodes were removed from vaccinated and nonvaccinated mice and placed separately into cold DMEM. Single-cell suspensions of lymph node cells were prepared by teasing apart the lymph nodes with forceps and gently pressing them through a sterile stainless steel 60-mesh screen into antimicrobial-free cold DMEM supplemented with 10% heat-inactivated fetal bovine serum, l-glutamine, and 2-mercaptoethanol. Lymph node cells were washed twice by centrifugation (1,500 rpm, 4°C, 10 min) with DMEM. Supernatants were decanted, and pellets were resuspended in 1 ml of cold DMEM. Cell viability of 90% was assessed by trypan blue exclusion.
Sterile six-well flat-bottomed tissue culture dishes (Becton Dickinson, Lincoln Park, N.J.) were inoculated with lymph node cells (5 × 106) obtained from vaccinated or nonvaccinated mice, macrophages from nonvaccinated mice (1 × 105), and 106 live B. burgdorferi. DMEM was added to the suspensions of cells to bring the final volume to 3 ml. Cells were cultivated at 37°C in the presence of 5.0% CO2. In some experiments, 10 μg of recombinant IFN-γ (rIFN-γ) or 75 μg of rat anti-murine IFN-γ (R&D Systems, Minneapolis, Minn.) was added to cultures of immune lymph node cells, macrophages, and B. burgdorferi at 10 min of incubation. These concentrations of rIFN-γ and anti-murine IFN-γ yielded maximum responses. In similar fashion, control cultures were incubated with a rat isotype-nonspecific antibody. One set of cultures was inoculated for determination of borreliacidal activity, while another set was inoculated for subsequent infusion into recipient mice and flow cytometric analysis.
Twenty-four hours after cultivation, nonimmune or immune lymph node cells treated with rIFN-γ, anti-murine IFN-γ, or an isotype-nonspecific antibody were aspirated and dispensed into separate centrifuge tubes. The cells were washed four times by centrifugation with warm DMEM (1,000 rpm, 22°C, 10 min). The cells were then resuspended in ≈250 μl of warm DMEM, and viability was enumerated by trypan blue exclusion. Subsequently, the concentrations of cells were adjusted with warm DMEM to yield 107 lymphocytes/ml. Four groups of three mice each were mildly anesthetized with methoxyflurane with a mouth-and-nose cup and injected subcutaneously in the hind paw with 0.1 ml of the lymphocyte suspensions. A fifth group of mice were administered 0.1 ml of warm DMEM.
Twenty-four hours after in vitro cultivation, 5 × 105 cells were transferred to chilled centrifuge tubes containing 500 μl of cold DMEM. Cells were then incubated (15 min, 4°C, dark conditions) with 5 μl of phycoerythrin-conjugated rat anti-murine CD4 or rat anti-murine CD8 (PharMingen, San Diego, Calif.). Separate populations of cells were also incubated with phycoerythrin-conjugated rat IgG2a (PharMingen) as an isotype control. Cells were then washed two times by centrifugation with PBS containing 0.1% bovine serum albumin (1,500 rpm, 4°C, 10 min), fixed with 1% paraformaldehyde, and kept in the dark until flow cytometric analysis.
Data were acquired on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, Calif.) with CellQuest acquisition and analysis software (Becton Dickinson). Cells were detected by forward scatter, side scatter, and phycoerythrin fluorescence. Data from 10,000 events were analyzed by histogram profiles of phycoerythrin fluorescence. Gates were established with samples stained with the isotype control antibody. The percentage of CD4+ and CD8+ cells was determined by the shift in phycoerythrin fluorescence of the stained cells.
At day 9 of cultivation at 37°C in the presence of 5.0% CO2, 1.0-ml samples of the supernatants were removed after gentle agitation and replaced with an equal volume of warm DMEM. Supernatants were collected after centrifugation at 13,000 rpm for 8 min to remove spirochetes and other cellular debris. Supernatants were stored at −70°C until used.
Swelling of the left hind paws of mice was used to evaluate the inflammatory response. Before experimentation, mice were chosen randomly and their hind paws were measured before they were assigned to cages. After infusion of cells or administration of DMEM, the left hind paws of recipient mice were measured periodically for 21 days with a dial-type vernier caliper (Fisher Scientific, Pittsburgh, Pa/) graduated in 0.1-mm increments. Severe swelling was assessed by mildly anesthetizing each mouse, carefully measuring the width and thickness of each left hind paw, combining these values, and determining mean dimensions within a given group. The standard error of the mean was also calculated for each mean caliper value.
Mice were anesthetized by inhalation of methoxyflurane contained in a mouth-and-nose cup and bled by intracardiac puncture. Recipient mice were bled 21 days after infusion of cell culture aspirates. The blood was allowed to clot, and serum was separated by centrifugation (1,500 rpm, 4°C, 10 min), divided into 0.20-ml aliquots, dispensed into 1.5-ml screw-cap tubes (Sarstedt), and frozen at −70°C until used.
Twenty-one days after administration, mice were euthanized by CO2 asphyxiation, and the left hind legs were amputated at the midfemur and fixed in 10% neutral buffered zinc-formalin. The legs were then placed in decalcifying solution (Fisher) for 24 h, followed by addition of fresh decalcifying solution for an additional 24 h. Following decalcification, legs were washed three times in double-distilled H2O, bisected longitudinally, placed in embedding cassettes, embedded in paraffin, cut into 6-μm sections, placed on glass slides, and stained with hematoxylin and eosin. Sections were cryptically coded for unbiased histopathological examination by a certified pathologist.
Frozen supernatants or sera were thawed, heat-inactivated (56°C, 30 min), sterilized with a 0.22-μm filter (Acrodisk; Gelman Sciences, Ann Arbor, Mich.), and serially twofold diluted (neat to 1:8,192 for supernatants; 1:10 to 1:1,280 for sera) with fresh BSK medium. One-hundred-microliter aliquots of each dilution were transferred to 1.5-ml screw-cap tubes (Sarstedt), and 100 μl of BSK containing 104 B. burgdorferi organisms per ml was added along with 20 μl of sterile guinea pig complement (Sigma). The tubes were then gently shaken and incubated for 3 days at 32°C. Controls included filter-sterilized supernatants obtained from cultures of nonimmune lymph node cells or macrophages alone and DMEM.
After incubation, 100 μl of each suspension was removed and placed into individual 1.5-ml screw-cap tubes (Sarstedt). Subsequently, 100 μl of a solution of propidium iodide (1.0 mg/ml; Molecular Probes, Eugene, Oreg.) diluted 1:20 in sterile PBS was added. The suspensions were briefly mixed before being incubated at 56°C for 30 min to permit intercalation of propidium iodide into the spirochetes. One hundred microliters of each sample was then filtered through 0.22-μm-pore-size Nuclepore polycarbonate membrane filters (47-mm diameter; Whatman Nuclepore, Clifton, N.J.) under negative pressure with a single-place sterility test manifold (Millipore Corporation, Bedford, Mass.) attached to a vacuum pump. Membrane filters were washed with ≈8 ml of sterile double-distilled H2O, removed from the vacuum apparatus, allowed to dry, and placed onto glass microscope slides. Coverslips were placed on the filters before viewing with a Laborlux S fluorescence microscope (Leitz, Wetzlar, Germany) with a 50x oil immersion objective.
The number of spirochetes on each filter was quantitated by viewing ≈30 fields. The borreliacidal antibody titer was defined as the reciprocal of the dilution preceding the dilution at which the number of spirochetes or clumping was equal to the control. Generally, individual spirochetes with a few clumps were uniformly distributed throughout the fields on filters of the control supernatants.
A t test (36) was used to determine significant differences in the titers of borreliacidal antibody. In addition, differences in edema between treatment groups were tested by two-way analysis of variance, utilizing the Minitab statistical analysis program. The Fisher least-significant-difference test (36) was used to examine pairs of means when a significant F ratio indicated reliable mean differences. The alpha level was set at 0.05 before the experiments were started.
Ten C3H/HeJ mice were vaccinated with formalin-inactivated B. burgdorferi in aluminum hydroxide. Lymph node cells were isolated 17 days after vaccination, cultured with macrophages and B. burgdorferi, and treated with anti-murine IFN-γ, rIFN-γ, or an isotype-nonspecific antibody (Fig. (Fig.1).1). An eightfold increase in borreliacidal antibody was detected in supernatant obtained from cultures of lymph node cells treated with anti-murine IFN-γ (titer 4,096) compared with levels of borreliacidal antibody detected in supernatant obtained from lymph node cells treated with an isotype nonspecific antibody (titer 512). In contrast, treatment of lymph node cells with rIFN-γ inhibited the borreliacidal antibody response (titer 4). When these studies were repeated five times with 50 mice, anti-murine IFN-γ augmented borreliacidal antibody production 8- to 16-fold.
Lymph node cells were obtained from 17-day vaccinated mice and cultured with macrophages and B. burgdorferi in the presence or absence of anti-murine IFN-γ, an isotype-nonspecific antibody, or rIFN-γ for 24 h. Subsequently, each group of treated lymph node cells was injected into the hind paws of normal C3H/HeJ mice (Fig. (Fig.2).2). A fourth group of four mice was injected in the hind paws with lymph node cells cultured with macrophages and B. burgdorferi obtained from nonvaccinated mice. Swelling of the hind paws was detected in all groups because infection with B. burgdorferi alone can induce edema of the hind paws (3, 19, 23, 32). However, the severity of swelling was greater in recipients infused with lymph node cells treated with anti-murine IFN-γ or rIFN-γ. Of these two groups, the hind paws of recipients of lymph node cells treated with anti-murine IFN-γ were more swollen day 18 to 21 after cell transfer. When these experiments were repeated three times with 36 mice, similar results were obtained.
Twenty-one days after cell transfer, destructive arthropathy was detected in the tibiotarsal joints of recipients of lymph node cells treated with anti-murine IFN-γ (Fig. 3B and C). A dense proliferation of synoviocytes and fibroblasts was observed on the periphery of the periosteum and surrounding tendons (Fig. (Fig.3B).3B). Focal erosion of bone was also detected (Fig. (Fig.3C).3C). By contrast, only a mild synovitis was found in recipients infused with lymph node cells treated with rIFN-γ, despite the presence of edema (Fig. (Fig.2).2). Synoviocyte-fibroblast proliferation was present in the periarticular soft tissue and flexor and extensor tendons (Fig. (Fig.3E).3E). However, the synoviocyte-fibroblast proliferation was considerably less than the proliferation detected in recipients of lymph node cells treated with anti-murine IFN-γ. Similar histopathology was detected in recipients of lymph node cells treated with an isotype-nonspecific antibody (Fig. (Fig.3D).3D). The synoviocyte-fibroblast proliferation enveloped the flexor and extensor tendons and was present in the periarticular soft tissue. Recipients injected with lymph node cells from nonvaccinated mice also developed a mild synovitis (Fig. (Fig.3F),3F), while recipients infused with culture medium (Fig. (Fig.3A)3A) had no significant histopathology.
Sera were obtained from recipient mice 21 days after infusion of lymph node cells from nonvaccinated mice and immune lymph node cells treated with anti-murine IFN-γ, rIFN-γ, or an isotype-nonspecific antibody (Table (Table1).1). No significant differences (P > 0.05) in the level of borreliacidal antibody were detected among the groups.
Table Table22 shows that treatment of lymph node cells with anti-murine IFN-γ, rIFN-γ, or an isotype-nonspecific antibody did not greatly alter the T-lymphocyte populations. An increase in CD8+ T lymphocytes (5%) was detected in lymph node cells treated with anti-murine IFN-γ. No differences in the populations of CD4+ T lymphocytes was detected among these groups.
Although vaccination with B. burgdorferi or its components induce protective borreliacidal antibody (1, 20, 27, 28, 30, 33), the duration of protection is short, less than 8 weeks (24). The development and maintenance of a high level of sustained borreliacidal antibody is paramount for prolonged protection against infection with B. burgdorferi. Previously, we showed (21, 22) that treatment of cells capable of producing borreliacidal antibody with exogenous IL-4 or IFN-γ failed to augment borreliacidal activity.
We now show that neutralization of IFN-γ does enhance borreliacidal antibody production. However, when anti-murine IFN-γ treated cells were transferred to normal C3H/HeJ recipients, the mice developed severe destructive arthritis. These results show that cell populations capable of producing enhanced borreliacidal antibody also contain cells capable of inducing severe destructive arthritis. Since we used a whole-cell vaccine, these results might be expected because of the numerous outer surface proteins of B. burgdorferi. However, subunit vaccines may also contain amino acid sequences capable of inducing adverse effects even though they can induce borreliacidal antibody. Our in vitro assay will be valuable for testing both effects.
This study demonstrates for the first time that borreliacidal antibody can be augmented by preventing the effects of IFN-γ. When lymph node cells capable of producing borreliacidal antibody were treated with anti-murine IFN-γ, an eightfold increase in borreliacidal antibody production was obtained. These results suggest that the amount of endogenous IFN-γ present in cultures of borreliacidal antibody producing cells affects the level of production of borreliacidal antibody. In support, we showed previously that addition of exogenous IFN-γ to cells capable of producing borreliacidal antibody abrogated the borreliacidal response (21).
The mechanism by which IFN-γ controls the production of borreliacidal antibody is unknown. It is known that IFN-γ is an antagonist for IL-4, a known B-lymphocyte stimulator. Neutralization of IFN-γ may enhance borreliacidal antibody production by up-regulation of IL-4. When rIL-4 was added to cells capable of producing borreliacidal antibody, it also failed to enhance borreliacidal antibody production (22). Moreover, neutralization of endogenous IL-4 failed to alter production of borreliacidal antibody. Collectively, these results suggest that a cytokine(s) other than IFN-γ and IL-4 is responsible for the modulation of borreliacidal antibody. Additional experiments are needed to determine which cytokine(s) augments borreliacidal antibody production in anti-murine IFN-γ-treated borreliacidal antibody-producing cells.
Although treatment of immune cells with anti-murine IFN-γ enhanced in vitro production of borreliacidal antibody, a similar response was not detected in recipients of these cells. In fact, no significant differences in the levels of borreliacidal activity were detected among groups that received nonimmune cells or cells treated with anti-murine IFN-γ, rIFN-γ, or an isotype-nonspecific control antibody. These results suggest that borreliacidal antibody producing cells can be rapidly downregulated by host cells. In support, Jensen et al. (12) showed that production of borreliacidal antibody could be prevented by incubating antibody producing cells with cells incapable of producing cidal antibody. The mechanism of downregulation may include IL-4 (26) and IFN-γ (25). We showed previously that IL-4 and IFN-γ could abrogate the borreliacidal antibody response (21, 22). Another explanation is that borreliacidal antibody production in vivo was delayed. In support, borreliacidal activity is not detected in vitro until 6 days after incubation (22).
These results are encouraging because enhanced production of borreliacidal antibody was achieved. However, treatment of borreliacidal antibody-producing cells with anti-murine IFN-γ also augmented the ability of these cell populations to induce arthritis. When cells containing borreliacidal antibody-producing cells treated with anti-murine IFN-γ were infused into immunocompetent C3H/HeJ recipients, they developed severe destructive Lyme arthritis. Histopathologic examination showed that severe erosion of bone had occurred. The severity of the arthropathy detected in recipients of anti-IFN-γ-treated cells exceeded the histopathology reported previously in other immunocompetent murine models of Lyme borreliosis (3, 19, 23, 32). These studies (3, 19, 23, 32) and ours showed that mice infused with rIFN-γ-treated borreliacidal antibody-producing cells, antibody-producing cells treated with a isotype-nonspecific control antibody, or nonimmune cells developed mild synovitis without erosion of bone.
Brown and Reiner (4) showed that IFN-γ was not required for the development or resolution of arthritis. When IFN-γ-deficient C3H mice were challenged with B. burgdorferi, they developed arthritis. By contrast, Keane-Myers and Nickell (14, 15) showed in cell depletion experiments that CD8+ T-lymphocyte-derived IFN-γ promoted the development of arthritis while CD4+ T lymphocytes prevented arthropathy. Although we found no difference in CD4+ T lymphocyte populations, we did detect a 5% increase in CD8+ T lymphocytes in lymph node cell cultures treated with anti-murine IFN-γ. It is doubtful, however, that these cells could produce enough IFN-γ in the presence of anti-murine IFN-γ to have an effect in vivo. In support, we showed that rIFN-γ-treated cells failed to induce severe destructive Lyme arthritis. Recipients of rIFN-γ-treated cells developed a mild synovitis similar to the synovitis detected in mice challenged with B. burgdorferi.
The inability of anti-murine IFN-γ treatment of immune cells obtained from B. burgdorferi vaccinated mice to decrease adverse arthritic effects, despite augmenting production of borreliacidal antibody, does not lessen the significance of this finding. Borrelial vaccines could be easily evaluated with this assay for increased production of borreliacidal antibody and for the ability to induce arthritis. Lymph node cells from vaccinated mice would be cultured in the presence of macrophages, borrelial vaccine, and anti-murine IFN-γ. The level of production of borreliacidal antibody could be readily determined in vitro, while infusion of these anti-murine IFN-γ-treated cells into recipients would assess their ability to induce arthritis. If a vaccine induced both borreliacidal antibody and arthritis, selective deletion of epitopes that did not affect the production of borreliacidal antibody could be determined by retesting the modified vaccine. Presently, we are using this system to determine the arthritogenic epitopes on OspA.
In conclusion, neutralization of IFN-γ in cultures of cells from B. burgdorferi-vaccinated mice enhanced both the production of borreliacidal antibody and the development of arthritis. Our in vitro-in vivo system can easily be used to evaluate the potential of borrelial vaccines for induction of borreliacidal antibody and arthritis. Additional studies are needed to delineate the cytokine mechanism(s) that prevents the development of arthritis and yet augments borreliacidal antibody production. These studies are necessary to ensure an effective and safe vaccine against infection with B. burgdorferi.
We thank the Wisconsin State Laboratory of Hygiene, the public health laboratory for the state of Wisconsin, and the Gundersen Medical Foundation for continued support.