ABSTRACT
Previous studies have shown that cells and cytokines associated with interleukin-17 (IL-17)-driven inflammation are involved in the arthritic response to Borrelia burgdorferi infection. Here, we report that IL-17 is a contributing factor in the development of Lyme arthritis and show that its production and histopathological effects are regulated by interleukin-10 (IL-10). Spleen cells obtained from B. burgdorferi-infected, “arthritis-resistant” wild-type C57BL/6 mice produced low levels of IL-17 following stimulation with the spirochete. In contrast, spleen cells obtained from infected, IL-10-deficient C57BL/6 mice produced a significant amount of IL-17 following stimulation with B. burgdorferi. These mice developed significant arthritis, including erosion of the bones in the ankle joints. We further show that treatment with antibody to IL-17 partially inhibited the significant hind paw swelling and histopathological changes observed in B. burgdorferi-infected, IL-10-deficient mice. Taken together, these findings provide additional evidence of a role for IL-17 in Lyme arthritis and reveal an additional regulatory target of IL-10 following borrelial infection.
INTRODUCTION
Borrelia burgdorferi, the causative agent of Lyme borreliosis, stimulates a complex series of inflammatory events that functions to eliminate the spirochete following infection but which also frequently leads to arthritis in humans (1). In addition, B. burgdorferi induces production of the anti-inflammatory cytokine interleukin-10 (IL-10) early after stimulation of host cells (2–4), suggesting a possible mechanism of facilitating infection in the host. IL-10 also plays a central role in regulating the immune response during sustained infection with B. burgdorferi. Notably, a deficiency of IL-10 in otherwise “arthritis-resistant” C57BL/6 mice predisposes these animals to developing remarkable joint pathology following B. burgdorferi infection (3). Despite exhibiting more severe arthritis, these B. burgdorferi-infected IL-10-deficient mice are more effective in eliminating the pathogen than are wild-type counterparts (3). Therefore, the dual outcomes associated with the inflammatory response to borrelial infection—robust antipathogen and arthritic responses—are inexorably linked to IL-10. As additional immune mechanisms against B. burgdorferi continue to be identified, the extent to which IL-10 participates in these responses must be defined further.
Accumulating evidence supports the contribution of such novel mechanisms in the development of Lyme arthritis. The pathology of disease in humans is driven, in large part, by Th1 cells and their prototypical inflammatory cytokine, gamma interferon (IFN-γ) (5–8). However, studies with multiple animal models have shown that a Th1 response is not absolutely necessary for the development of arthritis after B. burgdorferi infection (9–11), while the inflammatory cytokine interleukin-17 (IL-17) contributes to disease (12–15). Specifically, blocking IL-17 and various Th17-associated cytokines in vivo reduced the severity of arthritis observed in models of Borrelia-associated arthritis in the presence (15–17) or absence (12–14) of IFN-γ. Importantly, Infante-Duarte et al. described the capacity of synovial cells of human Lyme arthritis patients to produce IL-17 when stimulated with B. burgdorferi microbes or lipoproteins (18). Specifically, the borrelial antigen neutrophil-activating protein A (NapA) stimulates innate immune cells to provide a cytokine environment conducive to the development of Th17 cells within the joint (19). In addition, NapA is retained in the synovia of humans with Lyme arthritis and maintains this Th17 response through its chemoattractant activity on various inflammatory cell types, including T cells that produce both IL-17 and IFN-γ (20). Therefore, IL-17 is a likely contributor to the pathogenesis of Lyme arthritis. Since IL-10 is a known modulator of the IL-17 response (21), it is necessary to determine the relationship of these cytokines during B. burgdorferi infection.
In this report, we show that IL-10 inhibits the production of IL-17 from B. burgdorferi-stimulated cells. In addition, we demonstrate that multiple cell types are likely involved in the IL-17 response following borrelial infection, and that the arthritis observed in infected IL-10-deficient mice develops, in part, due to the actions of IL-17. Collectively, these findings provide additional evidence that IL-17 contributes to the arthritis induced by B. burgdorferi and reveal an additional target of regulation by IL-10 in the context of borrelial infection.
MATERIALS AND METHODS
Mice.Six- to 12-week-old, wild-type and IL-10-deficient mice on the C57BL/6 background were obtained from J.-A. Lyons (University of Wisconsin—Milwaukee) and housed at the University of Wisconsin—Milwaukee animal facility. The mice weighed 20 to 30 g each and were housed in a humidity-controlled environment at an ambient temperature of 21°C with a light-and-dark cycle of 12 h. Food and acidified water were provided ad libitum. Experimental protocols were reviewed and approved by the Institutional Animal Care and Use Committee for the University of Wisconsin—Milwaukee.
Organisms and preparation.Virulent B. burgdorferi 297 organisms (isolated from human spinal fluid) in modified Barbour-Stoenner-Kelly (BSK) medium supplemented with 10% glycerol were provided by S. M. Callister (Gundersen Health System, La Crosse, WI). Low-passage (<10) organisms were grown at 32°C until they reached a concentration of 107 microbes/ml, dispensed into 1.5-ml screw-cap tubes (Sarstedt, Newton, NC), and stored at −80°C. Prior to infection of mice, a 1-ml frozen suspension of spirochetes was thawed, added to 4 ml of BSK medium (Sigma-Aldrich, St. Louis, MO), and incubated at 34°C for 6 days. On the day of infection, microbes were visualized by dark-field microscopy for motility and enumerated using a Petroff-Hausser counting chamber.
Infection of mice.Naive mice were anesthetized with isoflurane in a nose-and-mouth cup and then injected subcutaneously in each hind paw with 106 viable B. burgdorferi strain 297 organisms in a volume of 50 μl of BSK medium. Control groups consisted of naive mice that were injected with BSK medium alone. Experimental and control groups with or without infection with B. burgdorferi contained 4 or 5 mice each.
Administration of anti-IL-17 antibody.Neutralizing anti-mouse IL-17A purified antibody (clone eBioMM17F3; IgG1) and mouse IgG1 purified isotype control antibody were purchased from eBioscience (San Diego, CA). The antibodies were resuspended in filter-sterilized (0.22 μm-pore-size filter; Fisher Scientific, Waltham, MA) phosphate-buffered saline (PBS) to achieve concentrations of 50 μg/ml. Groups of B. burgdorferi-infected mice were injected within 3 h after infection with 2.5 μg (50 μl) of anti-IL-17 antibody subcutaneously in each hind paw while anesthetized with isoflurane contained in a nose-and-mouth cup. Mice were then injected daily thereafter with anti-IL-17 antibody for 7 days. Groups of B. burgdorferi-infected control mice were injected with isotype control antibody in a similar manner.
Assessment of swelling.Edematous changes of the hind paws of mice were evaluated in order to provide a measure of inflammation following infection with B. burgdorferi. A digital caliper (Marathon, Richmond Hill, Ontario, Canada) with a sensitivity of 0.01 mm was used to measure the width and thickness of each tibiotarsal joint. The width and thickness measurements were averaged to provide the mean caliper value. Baseline swelling levels of each group were measured prior to infection, and the changes in swelling from these baseline levels were determined at 3, 6, and 8 days after infection.
Cell culture and enzyme-linked immunosorbent assay (ELISA).Spleens were harvested from 4 or 5 mice per group on day 1 or 8 after infection. Single-cell, whole-spleen suspensions from each group were generated by gently passing cells through a nylon mesh screen (BD Falcon; BD Biosciences, San Jose, CA) into cold RPMI 1640 medium supplemented with l-glutamine and 25 mM HEPES (Cellgro; Corning, Manassas, VA). A total of 1 × 106 whole spleen cells were incubated at 37°C in 5% CO2 with or without 1 × 105 viable B. burgdorferi organisms in BSK medium and with or without 5 μg of anti-mouse CD4 antibody (clone GK1.5; eBioscience) in PBS. Supernatants were collected after 6 and 24 h of incubation. Controls consisted of cells incubated with PBS but without antibodies or with BSK medium but without organisms.
Levels of IL-17 present in the supernatants of cell cultures were assessed using the IL-17A ELISA Ready-Set-Go! kit (eBioscience, San Diego, CA) and detected according to the manufacturer's directions. The plates were read at 450 nm, and values were expressed as optical densities. Standard curves were created and used to calculate pg of IL-17A/ml of each sample.
Borreliacidal-antibody assay.Blood was obtained by intracardiac puncture 8 days after infection, and sera were used to determine titers of borreliacidal antibodies. Samples of equal volume were pooled within a group of 4 or 5 mice. Each pooled sample was diluted 1:20 in PBS, passed through a 0.22-μm filter, and then further diluted serially, 1:40 to 1:10,240, with PBS. The diluted specimens were heated at 56°C for 30 min to inactivate complement and then cooled to 35°C. The complement-inactivated specimens were incubated with 104 viable B. burgdorferi strain 297 organisms and 20 μl of sterile guinea pig complement serum (Sigma-Aldrich, St. Louis, MO; 50% hemolytic component ≥ 200 U/ml) in PBS for 24 h. The presence of viable microbes was then observed in 20 fields using dark-field microscopy. The titer of borreliacidal antibody was considered the reciprocal of the final dilution in which motile B. burgdorferi organisms were observed. Control samples contained B. burgdorferi and complement alone, B. burgdorferi and PBS alone, and B. burgdorferi and PBS added to naive, nonimmune serum.
Preparation of tissues for histological examination.At 8 days after infection, the different groups of 4 or 5 mice each were euthanized, and their hind paws were amputated above the knee joint. The paws were cryptically coded, fixed in 10% buffered zinc formalin, and then decalcified. Following decalcification, the paws were placed in tissue-embedding cassettes, embedded in paraffin, and cut into 6-μm sections. The sections were placed on glass slides and stained with hematoxylin and eosin. An unbiased examination was performed by a board-certified pathologist (T.F.W.). The following scale was used to quantify the degree of histopathological changes: 0, no change; 1, mild leukocytic infiltration of the subsynovial tissues; 2, moderate infiltration of the subsynovial tissues, synovium, and synovial space; 3, profuse leukocytic infiltration of subsynovial tissues, synovium, and synovial space; and 4, profuse leukocytic infiltration of tissues with accompanying destruction of bone or cartilage.
Statistics.Data were analyzed by analysis of variance (ANOVA), with the Tukey-Kramer post hoc test used to examine pairs of means when a significant F value indicated reliable mean differences between the control and test groups. Paired data were analyzed using a two-tailed Student t test. The alpha level was set at 0.05 prior to initiation of the experiments. Data are expressed as the means ± standard errors of the means. Figures reflect the results of a single experiment. Replicate experiments showed similar relationships.
RESULTS
B. burgdorferi-induced IL-17 production is increased 8 days after infection in the absence of IL-10.We measured the in vitro production of IL-17 by cells of wild-type and IL-10-deficient mice that were uninfected or infected with B. burgdorferi for 8 days. Cells from wild-type mice, regardless of whether the mice were previously infected, produced negligible amounts of IL-17 after 6 h of incubation with B. burgdorferi (less than 4 pg/ml) (Fig. 1A). Production of IL-17 by these stimulated cells did not differ from that of unstimulated counterparts. In addition, stimulated cells from B. burgdorferi-infected wild-type mice produced slightly more IL-17 than stimulated cells from uninfected wild-type mice, but this increase was not significant. In contrast, B. burgdorferi-stimulated cells from uninfected and infected IL-10-deficient mice produced significantly more IL-17 than stimulated wild-type counterparts (P ≤ 0.05) (Fig. 1A). Production of IL-17 by stimulated IL-10-deficient cells also was significantly greater than that of unstimulated IL-10-deficient cells (P ≤ 0.05) (Fig. 1A). These increases in IL-17 production following in vitro stimulation were observed among cells of both uninfected and B. burgdorferi-infected IL-10-deficient mice. Additionally, stimulated cells from infected IL-10-deficient mice produced significantly more IL-17 than stimulated cells from uninfected IL-10-deficient mice (P ≤ 0.05) (Fig. 1A).
IL-17 production is increased in the absence of IL-10 8 days after infection with B. burgdorferi. Splenic cells obtained from wild-type (black bars) and IL-10-deficient (white bars) mice 8 days after injection of B. burgdorferi or BSK medium were cultured in the presence or absence of the spirochete for 6 (A) and 24 (B) hours. (A) After 6 h of incubation, cells from infected or uninfected wild-type mice produced minimal IL-17, regardless of in vitro stimulation. In contrast, stimulated cells from uninfected and B. burgdorferi-infected IL-10-deficient mice produced significant amounts of IL-17. The IL-17 production by these stimulated IL-10-deficient cells was significantly greater than that of stimulated wild-type counterparts as well as that of unstimulated IL-10-deficient counterparts. Additionally, stimulated cells from infected IL-10-deficient mice produced significantly more IL-17 than stimulated cells from uninfected IL-10-deficient mice. (B) Production of IL-17 by wild-type cells remained low after 24 h of incubation with or without the spirochete. In contrast, stimulated cells from B. burgdorferi-infected IL-10-deficient mice produced a large amount of IL-17. The production of IL-17 by these cells was significantly greater than that of stimulated cells from infected wild-type mice, stimulated cells from uninfected IL-10-deficient mice, and unstimulated cells from infected IL-10-deficient mice. Additionally, unstimulated cells from infected IL-10-deficient mice produced significantly more IL-17 than did unstimulated cells from infected wild-type mice and uninfected IL-10-deficient mice. Data are mean levels of IL-17/ml of culture supernatant ± SEMs. *, P ≤ 0.05; **, significant (P ≤ 0.05) increase compared to unstimulated counterparts. ND, not detected.
After 24 h of in vitro stimulation with B. burgdorferi, cells from uninfected and infected wild-type mice produced low levels of IL-17 (less than 10 pg/ml) (Fig. 1B). Production of IL-17 by wild-type cells stimulated in vitro was slightly, but not significantly, greater than that of unstimulated wild-type cells. In addition, the production of IL-17 by cells of infected wild-type mice did not differ from the production of IL-17 by cells of uninfected wild-type mice. In contrast, B. burgdorferi-stimulated cells of uninfected and infected IL-10-deficient mice produced more IL-17 than did stimulated wild-type counterparts (over 100 pg/ml and 600 pg/ml, respectively) (Fig. 1B); however, only the latter cells exhibited a statistically significant increase (P ≤ 0.05). Unstimulated cells from infected IL-10-deficient mice also produced significantly more IL-17 than their unstimulated counterparts from infected wild-type mice (P ≤ 0.05) (Fig. 1B). In addition, stimulation of cells from infected IL-10-deficient mice caused a significant increase in IL-17 production compared to that obtained without stimulation (P ≤ 0.05) (Fig. 1B). Moreover, cells from infected IL-10-deficient mice produced significantly more IL-17 than cells from uninfected IL-10-deficient mice, regardless of whether the cells were stimulated in vitro (P ≤ 0.05) (Fig. 1B).
Effects of early borrelial exposure on IL-17 production.Splenic cells obtained from wild-type mice infected with B. burgdorferi for 1 day produced a negligible amount of IL-17 (less than 4 pg/ml) when cultured with or without B. burgdorferi for 6 h (Fig. 2A). In contrast, cells from uninfected wild-type mice produced significantly more IL-17 than their counterparts from infected wild-type animals, regardless of whether the cells were stimulated in vitro with the spirochete (P ≤ 0.05) (Fig. 2A). Also, stimulated cells from uninfected wild-type mice produced significantly more IL-17 than stimulated cells from uninfected IL-10-deficient mice (P ≤ 0.05) (Fig. 2A). In vitro stimulation did not cause a significant change in IL-17 production among cells of either uninfected or B. burgdorferi-infected wild-type mice. In contrast, in vitro exposure to B. burgdorferi for 6 h caused a significant decrease in the production of IL-17 by cells of both uninfected and infected IL-10-deficient mice (P ≤ 0.05) (Fig. 2A).
Effects of early B. burgdorferi exposure on production of IL-17. Splenic cells obtained from wild-type (black bars) and IL-10-deficient (white bars) mice 1 day after injection of B. burgdorferi or BSK medium were cultured in the presence or absence of the spirochete for 6 (A) and 24 (B) hours. (A) After 6 h of incubation, cells from uninfected wild-type mice produced significantly more IL-17 than did cells from infected wild-type mice, regardless of stimulation in vitro. Stimulated cells from uninfected wild-type mice also produced significantly more IL-17 than stimulated cells from uninfected IL-10-deficient mice. In addition, in vitro stimulation with B. burgdorferi for 6 h caused a significant decrease in IL-17 production by cells from IL-10-deficient, but not wild-type, mice, regardless of whether the mice were previously infected. (B) After 24 h of incubation, cells from uninfected wild-type mice produced significantly more IL-17 than did cells from infected wild-type mice, regardless of stimulation in vitro. Similarly, these cells produced significantly more IL-17 than those of uninfected IL-10-deficient counterparts. In addition, in vitro stimulation with B. burgdorferi for 24 h caused a significant decrease in IL-17 production by cells from infected wild-type mice, but not by those from uninfected wild-type mice, uninfected IL-10-deficient mice, or B. burgdorferi-infected IL-10-deficient mice. Data are mean levels of IL-17/ml of culture supernatant ± SEMs. *, P ≤ 0.05; **, significant (P ≤ 0.05) decrease compared to unstimulated counterparts.
After 24 h of incubation with or without B. burgdorferi, cells from uninfected wild-type mice produced significantly more IL-17 than did cells from wild-type mice that were infected with the spirochete for 1 day (P ≤ 0.05) (Fig. 2B). Cells from uninfected wild-type mice also produced a significantly greater amount of IL-17 than cells from uninfected IL-10-deficient mice, regardless of in vitro stimulation (P ≤ 0.05) (Fig. 2B). Exposure to B. burgdorferi in vitro for 24 h caused a significant decrease in IL-17 production by cells of infected wild-type mice (P ≤ 0.05) (Fig. 2B) but did not significantly affect cells from uninfected counterparts. In addition, stimulation with the spirochete for 24 h did not affect IL-17 production by cells of uninfected or B. burgdorferi-infected IL-10-deficient mice.
IL-17 production following blockage of CD4 in vitro.Suspensions of splenic cells derived from wild-type and IL-10-deficient mice, either uninfected or infected with B. burgdorferi for 8 days, were stimulated in vitro with spirochetes in the presence or absence of 5 μg/ml of anti-CD4 antibody for 6 and 24 h. After 6 h of incubation, no differences in IL-17 production were observed due to the presence of anti-CD4 antibody (data not shown).
After 24 h of incubation, the presence of anti-CD4 antibody slightly enhanced the production of IL-17 by cells of uninfected wild-type mice (Fig. 3A). However, this increase in IL-17 production was not statistically significant. In contrast, treatment of cells from B. burgdorferi-infected wild-type mice with anti-CD4 antibody significantly reduced IL-17 production compared to that of untreated cells (P ≤ 0.05) (Fig. 3A). Treatment of cells from uninfected or infected IL-10-deficient mice with anti-CD4 antibody caused a decrease in IL-17 production that was not statistically significant (Fig. 3B).
Partial inhibition of IL-17 production by blocking of CD4 in vitro. Wild-type (A) and IL-10-deficient (B) splenic cells obtained from uninfected mice and mice infected with B. burgdorferi for 8 days were stimulated with the spirochete in the presence or absence of 5 μg/ml of anti-CD4 antibody for 24 h. (A) Addition of anti-CD4 antibody to cells of uninfected, wild-type mice did not significantly affect IL-17 production. In contrast, anti-CD4 antibody treatment of cells from B. burgdorferi-infected wild-type mice caused a significant reduction in IL-17 production. (B) Anti-CD4 antibody treatment of cells from uninfected or B. burgdorferi-infected IL-10-deficient mice reduced the production of IL-17. However, these reductions were not statistically significant. Data are mean levels of IL-17/ml of culture supernatant ± SEMs. *, P ≤ 0.05.
Effects of anti-IL-17 antibody treatment on hind paw swelling.Wild-type and IL-10-deficient mice were infected in the hind paws with viable B. burgdorferi strain 297 organisms. One group of mice from each strain was administered anti-IL-17 antibody on the day of infection and then daily for 7 days, while the remaining group of mice received an isotype control antibody. The B. burgdorferi-infected, untreated IL-10-deficient mice had increased swelling of the hind paws compared to those of infected, untreated wild-type mice up to 6 days after infection (P ≤ 0.05) (Fig. 4). The edematous changes in both groups decreased to levels below baseline measurements by day 8 after infection.
Treatment of B. burgdorferi-infected IL-10-deficient mice with anti-IL-17 antibody significantly reduces hind paw swelling. Wild-type (black squares) and IL-10-deficient (white squares) mice were infected with B. burgdorferi and administered anti-IL-17 antibody (dashed lines) or isotype control antibody (solid lines) within 3 h of infection and daily thereafter for 7 days. Hind paw swelling of infected but untreated IL-10-deficient mice was greater than that of infected but untreated wild-type mice. Anti-IL-17 antibody treatment of infected wild-type mice caused a significant reduction in hind paw swelling that was evident on day 6 after infection. Treatment of B. burgdorferi-infected IL-10-deficient mice with anti-IL-17 antibody caused a substantial reduction in hind paw swelling that was evident on day 3 after infection and which lasted through day 6 after infection. Data are mean changes in paw swelling relative to baseline ± SEMs. *, P ≤ 0.05.
Treatment of B. burgdorferi-infected mice with antibody to IL-17 ameliorated the swelling observed in untreated mice. Administration of anti-IL-17 antibody significantly reduced the edematous changes in B. burgdorferi-infected IL-10-deficient mice at 3 and 6 days after infection (P ≤ 0.05) (Fig. 4). By day 8 after infection, there were no significant differences in hind paw swelling between these groups. Similarly, treatment of B. burgdorferi-infected wild-type mice with anti-IL-17 antibody reduced hind-paw swelling; however, the effects of the antibody in these mice were delayed and milder. A significant decrease in paw swelling was observed only at day 6 after infection (P ≤ 0.05) (Fig. 4). Mice that were injected with BSK medium and then injected daily with anti-IL-17 or isotype control antibody did not show changes in hind paw swelling from baseline levels (data not shown).
Effect of IL-17 antibody administration on histopathology.Histopathological analysis of hind paws, including the tibiotarsal joints and knees, was performed on untreated or anti-IL-17 antibody-treated mice infected with B. burgdorferi in order to determine whether the presence of IL-10 regulates IL-17-mediated arthritis induced by the spirochete. B. burgdorferi-infected wild-type mice exhibited minimal histopathological changes on day 8 after infection, with only mild hyperplasia observed in some animals (Table 1). B. burgdorferi-infected wild-type mice that were administered anti-IL-17 antibody displayed no significant histopathological changes. The histopathological changes observed in untreated, infected wild-type mice were too mild to attribute the lack of pathology in treated animals to the anti-IL-17 antibody.
Treatment of B. burgdorferi-infected IL-10-deficient mice with anti-IL-17 antibody reduces the severity of histopathologya
In contrast, the severity of pathology among untreated IL-10-deficient mice, particularly in the knee and perisynovium of the tibiotarsal joints, was significantly greater than that of untreated wild-type mice (P ≤ 0.05) (Table 1). IL-10-deficient mice displayed moderate lymphocytic infiltration, as well as hyperplasia and hypertrophy, in the tibiotarsal tissues and knee 8 days after infection with B. burgdorferi (Table 1). In addition, these mice exhibited mild erosion of bones in the tibiotarsal joint. Treatment of B. burgdorferi-infected, IL-10-deficient mice with antibodies to IL-17 slightly reduced, but did not abrogate, cellular infiltration, hyperplasia, and hypertrophy of the synovium and perisynovium of the tibiotarsal joint (Table 1). However, treatment of infected IL-10-deficient mice with anti-IL-17 antibodies did significantly reduce the histopathological changes of the knee joints observed in infected, but untreated, IL-10-deficient mice (P ≤ 0.05) (Table 1).
Borreliacidal-antibody titers.Borreliacidal-antibody titers were obtained with sera from wild-type and IL-10-deficient mice that were uninfected or infected with B. burgdorferi for 8 days. The borreliacidal-antibody titers were low (≤80) (Fig. 5) in all groups. No significant differences were detected.
IL-10 deficiency does not affect borreliacidal-antibody titers 8 days after infection. Sera from uninfected and B. burgdorferi-infected wild-type (black bars) and IL-10-deficient (white bars) mice were collected on day 8 after infection. Borreliacidal-antibody titers of all groups were low (≤80). No significant differences in titers between groups were detected.
DISCUSSION
In this report, we support and extend recent findings that IL-17 is a contributing factor in the development of Lyme arthritis by showing that its production and histopathological effects are regulated by IL-10. Spleen cells obtained from wild-type C57BL/6 mice, which are resistant to developing arthritis following infection with B. burgdorferi (22), produced low levels of IL-17 following stimulation with the spirochete. In contrast, spleen cells obtained from IL-10-deficient C57BL/6 mice, which develop moderate arthritis after borrelial infection (3), produced a significant amount of IL-17 at the time of arthritic manifestations following infection with B. burgdorferi. We further showed that the significant hind paw swelling and histopathological changes observed in B. burgdorferi-infected, IL-10-deficient mice were partially inhibited by administration of anti-IL-17 antibody. Collectively, these findings support our claims (12–17, 23) and those of others (19, 20) that IL-17 is involved in the development of arthritis following infection with B. burgdorferi. More importantly, they reveal a significant regulatory mechanism for an inflammatory pathway that is induced by borrelial infection (18) and which is known to contribute to several other types of arthritis (24–27).
The finding that IL-10 regulates the arthritic, IL-17-mediated response to B. burgdorferi adds to the known properties of the anti-inflammatory cytokine in controlling the immune response to, and pathological outcomes of, borrelial infection. Early investigations into the differences in pathology observed between C3H and C57BL/6 mice demonstrated that the effect of IL-10 is a key factor in determining disease outcome (3, 28). Until now, much of the investigation pertaining to the inflammatory targets of IL-10 focused on the Th1/IFN-γ response, which has long been known as a major inflammatory pathway of later-stage Lyme arthritis (5, 6). For example, modulation of IL-10 in vitro reciprocally influenced IFN-γ production by stimulated synovial cells of Lyme arthritis patients (29). In addition, differences in IL-10-mediated regulation of IFN-γ accounted for the differences in arthritic severity exhibited by C3H and C57BL/6 mice (3, 28). In support of this, recent findings show that IL-10 plays a particularly important role in controlling the IFN-γ-driven arthritis caused, in large part, by CD4+ T cells in B. burgdorferi-infected mice (30). Our current findings suggest that IL-10 additionally exerts its anti-inflammatory effects during B. burgdorferi infection by suppressing the production of IL-17, even in the presence of IFN-γ.
These findings are important because they provide additional evidence that IL-17 is elicited in response to B. burgdorferi and, ultimately, contributes to the pathology of Lyme arthritis. Consideration of IL-17 as an inflammatory mediator of Lyme arthritis stems from the observation that while Th1-mediated inflammation is a major cause of pathology (5–8), arthritis also develops independently of IFN-γ (9–11). Infante-Duarte et al. first determined that stimulated cells of Lyme arthritis patients produce IL-17 (18). Following this observation, a series of studies showed that various Th17-associated cytokines were associated with pathology in murine models of Borrelia-induced arthritis (12–17), supporting claims of the arthritic potential of IL-17 in Lyme arthritis (23). Supporting this, NapA of B. burgdorferi stimulated innate cells to drive the development of Th17 cells that were present in the synovia of humans with Lyme arthritis (19). NapA is a key antigen in the recruitment of inflammatory cells which, along with NapA itself, stimulate the production of IL-17 from different T cell populations (20). However, Th17 cells were greatly outnumbered by Th1 cells in the synovia of humans with antibiotic-refractory or antibiotic-responsive Lyme arthritis, comprising approximately 5% of all CD4+ T cells (8). As a result, little, if any, contribution to disease was attributed to Th17 cells. Although some studies have demonstrated the reciprocal nature of Th1 and Th17 cells (31), others have shown that these inflammatory pathways can function simultaneously and, occasionally, even be expressed simultaneously by an individual cell (20, 32). Our current findings support the assertion that IL-17 contributes to the pathology of Lyme arthritis.
Our findings also suggest that multiple cell types are responsible for the production of IL-17 in response to B. burgdorferi. Various cell types, including Th17 cells, produce the cytokine (33). We showed that unseparated spleen cells from infected C57BL/6 wild-type mice, when incubated with B. burgdorferi and antibodies to CD4, produced significantly less IL-17 than did stimulated cells not additionally incubated with antibodies. However, the production of IL-17 by the latter cells was initially low. When we blocked CD4 in cultures of stimulated cells obtained from IL-10-deficient mice, we observed decreases in IL-17 production; however, these decreases were not statistically significant. It is possible that the concentration of anti-CD4 antibody we used in this study was insufficient to counter the significant production of IL-17 by these IL-10-deficient cells. However, it is also probable that CD4+ and non-CD4+ cell populations produced IL-17 in this system. We are currently assessing the production of IL-17 by these cells in the presence of multiple doses of anti-CD4 antibody and are characterizing IL-17-producing cell phenotypes during infection.
It is important also to note that the immune pathways associated with regulation of IL-17 are complex. For example, in support of our findings, opposing functions of IL-10 and various IL-17-producing cell types have been reported in other disease models, including infections (34, 35) and, importantly, arthritis (36, 37). However, cells which produce both cytokines exist (32, 38), and IL-10-producing regulatory T (Treg) cells have also been shown to aid in the production of Th17 cells at certain periods of development (39). Moreover, the roles of Th17 cells and IL-17 in Lyme arthritis have been challenged by the finding that humans with a reduced capacity to bind the Th17 cell survival cytokine interleukin-23 have reduced IL-17 production but do not exhibit reduced arthritis (40). In addition, we showed that the Treg cell-associated cytokine interleukin-35 enhances, rather than inhibits, arthritis in murine model of disease and reduces IL-17 production only slightly (41). Taken together, these findings demonstrate the complexity of inflammatory and regulatory pathways involved in Lyme arthritis and necessitate their further investigation.
Our findings suggest that IL-10 regulates B. burgdorferi-mediated pathology in part by limiting the arthritic potential of IL-17. The data we present here are reminiscent of our findings linking IL-17 and putative Treg cells in a model of Borrelia-induced arthritis (13, 42, 43). We showed that blocking IL-17 in IFN-γ-deficient, Borrelia-vaccinated and -infected mice prevented pathology and increased the production of CD25-expressing CD4+ cells (13). Concomitantly injecting these mice with antibody to CD25 reduced this cell population and resulted in massive erosion of the ankle joints (13). Adoptive transfer of CD4+ CD25+ cells, enriched following anti-IL-17 antibody treatment, into infected recipients completely prevented the development of arthritis (42). Importantly, we did not observe regulatory functions among CD4+ CD25+ cells obtained from mice in which IL-17 was not blocked (43). Following these early experiments, Bettelli et al. showed that Th17 cells and CD4+ CD25+ Foxp3+ Treg cells developed via inverse pathways from a common precursor cell (44). In light of our findings from nearly a decade ago, it may not be surprising that we observed a significant IL-17 response in IL-10-deficient, B. burgdorferi-infected mice, as CD4+ CD25+ Foxp3+ Treg cells function, in large part, through their production of IL-10 (21). In addition, it may not be surprising that anti-IL-17 antibody treatment did not completely abrogate arthritis in infected IL-10-deficient mice, since increased IL-17 production is likely just one of many effects of an absence of IL-10. In support of our hypothesis that Treg cells are key mediators of Lyme arthritis (13, 23, 42, 43), more recent findings have shown that patients with persistent, antibiotic-refractory disease produce more Treg cells than those with disease that readily resolves with antimicrobial therapy (8). In addition, longer durations of antibiotic-refractory Lyme arthritis posttreatment are associated with lower frequencies of Treg cells (8, 45). Collectively, these data imply that Lyme arthritis may be modulated by IL-10, which, in the context of borrelial infection, downregulates inflammatory cytokines, including IL-17.
What is the importance of this alternative response to B. burgdorferi? We show that wild-type or IL-10-deficient cells harvested 1 day after infection, when subsequently stimulated with the spirochete, produce an equal or reduced amount of IL-17. Additionally, cells from B. burgdorferi-infected wild-type mice produce less IL-17 than cells from uninfected mice. Collectively, these findings suggest a tendency toward an anti-inflammatory mechanism early after interaction with the spirochete. In support of this, early infection with B. burgdorferi results in an increase in IL-10 (46), suggesting that the resulting suppression of the inflammatory response may allow a successful infection to occur. Our findings suggest that initial exposure to the spirochete may cause an early delay or decrease in IL-17 that not only may assist in successful infection but also may occur independently of IL-10. Suppression of IL-17 production following early exposure to B. burgdorferi may allow some of the initial infectious spirochetes to survive, possibly until the inflammatory response to borrelial antigens becomes more robust. Additional studies, particularly ones using skin samples of tick-infected mice, would be required to test this hypothesis. In contrast, the effects of IL-10 on IL-17 production were more pronounced after 1 week of infection. Borrelial infection in the presence or absence of IL-10 may be associated with an inverse response of various inflammatory cytokines, including not only tumor necrosis factor alpha and IFN-γ (3, 4, 28–30, 46–51) but also IL-17. If a certain threshold of immunosuppression by IL-10 is not achieved, a significant, multifaceted inflammatory response may ensue.
Our histopathological findings support the claim of multiple inflammatory pathways contributing to the development of Lyme arthritis. As expected, B. burgdorferi-infected, C57BL/6 wild-type mice exhibited mild paw swelling and minimal histopathological changes by day 8 after infection. Despite the mild hind paw swelling of these mice, administration of anti-IL-17 antibody significantly decreased the levels of edema by day 6 after infection. In contrast, B. burgdorferi-infected IL-10-deficient mice exhibited moderate swelling and histopathological changes of the hind paws. Treatment with antibody to IL-17 resulted in a significant reduction of swelling that was observed sooner and which was sustained for the duration of the experiment. Importantly, the overall histopathology of these treated mice was decreased, particularly in the knee and perisynovium of the tibiotarsal joint. However, some mice were free of pathology, while others still exhibited a degree of pathology. This supports the idea that IL-17 is a partial contributor to arthritis, and therefore, blocking only this cytokine would likely not completely eliminate disease.
Finally, we found that after 8 days of infection, IL-10 failed to significantly affect the levels of borreliacidal antibodies in the sera of B. burgdorferi-infected mice. A deficiency in IL-10 has been linked to reduced microbial loads and increased levels of B. burgdorferi-specific antibodies (3, 50); however, these findings were observed at least 2 weeks after infection. In addition, a robust innate immune response, rather than the borreliacidal antibody response, was implicated in the reduced levels of B. burgdorferi in the tissues of IL-10-deficient mice (50). Therefore, the lack of significant borreliacidal antibody activity in our IL-10-deficient mice may not be surprising. The apparent lack of an influence of IL-17 on borreliacidal-antibody production in these mice may also not be surprising, considering our previous finding that blocking IL-17 in a model of B. burgdorferi-mediated arthritis failed to affect these titers (14), and the recent finding that antibody production overall may not be affected directly by IL-17 (52). We are further investigating the role of IL-17 in the early antibody response to B. burgdorferi.
In summary, we showed that IL-10 regulates an arthritic IL-17 response following infection with B. burgdorferi. These findings add to the known immunomodulatory functions of IL-10 in the context of borrelial infection, as well as support the growing evidence that IL-17 contributes to disease. These findings were observed using infected IL-10-deficient C57BL/6 mice, which have been described recently as a viable model to investigate elements of adaptive immunity not readily observed in traditional disease models (30). This sentiment echoes our previous claims that the use of models that exhibit the full array of immune events (including T cells) involved in the development of arthritis in humans is vital for a more complete understanding of the inflammatory events associated with Lyme arthritis in humans (16, 17, 23). Here, we showed that this model also allows for investigating the role in Borrelia-mediated arthritis of IL-17, which, while shown to potentially contribute to human disease (19, 20), has not been demonstrated in traditional animal models (16). More importantly, we provide additional evidence for considering an inflammatory pathway that may be targeted for the treatment of Lyme arthritis.
ACKNOWLEDGMENTS
This study was supported by the UWM College of Health Sciences (Stimulus for Extramural Enhancement & Development Award), the UWM Graduate School (Graduate School Research Committee Award), and the Wisconsin State Laboratory of Hygiene, the public health laboratory for the state of Wisconsin.
We thank Steven M. Callister and Dean A. Jobe of the Gundersen Health System in La Crosse, WI, and Jeri-Anne Lyons of the UWM Department of Biomedical Sciences for providing the B. burgdorferi organisms and mice, respectively, used in this study.
FOOTNOTES
- Received 7 September 2013.
- Accepted 9 September 2013.
- Accepted manuscript posted online 16 September 2013.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
