INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Lyme disease is the most common vector-borne disease in the United States (28). The disease is transmitted to humans by the bite of an infected Ixodes tick harboring either Borrelia burgdorferi sensu stricto, B. afzelii, or B. garinii, related spirochetes collectively termed B. burgdorferi sensu lato. In the majority of patients, Lyme disease is characterized by the initial appearance of a rash-like skin lesion termed erythema migrans (EM) (39, 42). Early and late clinical manifestations include arthritis, neurological manifestations, lymphadenopathy, and carditis, which reflects dissemination to visceral targets (1, 16, 24, 29, 30, 34, 38, 40-43). Although Lyme disease is rarely fatal, it can be quite debilitating.

Implicit in the development of an effective vaccine against Lyme disease is a thorough understanding of the pathogenetic mechanisms as well as the host immune response operative during host-spirochete interaction. In the murine model of Lyme disease, the histopathology is characterized by a lymphocytic and plasma cell infiltration, occasional perivascular cuffing, and few or no detectable macrophages (8, 13, 18, 27, 32, 37). BALB/c and C3H mice have been used in elucidating the immune response of the host following infection with B. burgdorferi. Humoral immunity has been shown to play a major role in resistance to Lyme disease, although cellular immunity has also been demonstrated to influence the outcome of arthritis severity and infection (4-6, 20-23, 25, 26, 33, 47, 48). However, a direct relationship between the spirochetes and host immune cells has not been demonstrated. In addition, the chronic nature of the disease in the murine model precludes determining the immune mechanisms involved in clearance of organisms and resolution of the disease.

The rabbit model of Lyme disease has unique features relevant to the immunobiology of B. burgdorferi infection. It is the only animal model besides the rhesus monkey (31) that reproducibly results in EM that is indistinguishable from that of human disease after intradermal inoculation of B. burgdorferi sensu stricto (15). Most importantly, untreated skin and visceral infection is ultimately cleared, resulting in complete immunity against reinfection with up to 2 × 10⁷ challenge organisms (15). Thus, in the rabbit model of Lyme disease, it is possible to examine the effective host immune factors, humoral and cellular, associated with acquired resistance against infection with B. burgdorferi.

In the current study, we directed our efforts toward elucidating the outcome of host-spirochete interaction following the intradermal inoculation of naive and infection-derived immune rabbits with B. burgdorferi B31 from the time of intradermal inoculation to the time of EM development and through development of challenge immunity. We specifically addressed the location and persistence of spirochetes in relationship to the presence and distribution of polymorphonuclear cells (PMNs), macrophages, T cells, and B cells in an effort to determine the potential cellular mechanisms responsible for pathogenesis and host immunity.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Adult, male, New Zealand White rabbits 6 to 9 months of age (Irish Farms, Norco, Calif.) were housed individually in a temperature-controlled environment ranging from 18 to 21°C. Prior to intradermal (i.d.) inoculation with B. burgdorferi, the backs of the rabbits were clipped closely with an electric clipper fitted with a size 40 blade to expose the skin (Oster Professional Products, McMinnville, Tenn.).

Bacterial strains and preparation of inocula for infection and challenge. Virulent B. burgdorferi sensu stricto, strain B31, was isolated from infected rabbit tissue, grown in BSK II medium (7) to maximum density, and then passaged twice more in fresh BSK II medium. After the final passage (passaged two times), the organisms were centrifuged gently at 8,000 × g for 10 min and washed three times in heat-inactivated (56°C for 30 min) normal rabbit serum (NRS), diluted 1:1 with phosphate-buffered saline (PBS; pH 7.2) (NRS-PBS), in order to remove foreign protein components responsible for nonspecific reactions in the rabbit (15, 36). Inocula for infection and challenge were resuspended in NRS-PBS after the final wash to a final concentration containing 10⁷ or 10⁸ B. burgdorferi per ml depending on the experiment. As in previously published experiments (15, 36), the spirochetes in the final suspension were motile and virulent.

Generation and challenge of infection-derived immune rabbits. Based on previous studies in which a high degree of infection-derived immunity to challenge was established (15), rabbits were infected i.d. with doses ranging from 6 × 10⁶ to 8 × 10⁷ B. burgdorferi depending on the experiment. Inocula were administered in 0.1-ml doses at several sites. Six months after infection, at a time when EM lesions had healed, dermal infection was absent and, on the basis of our previously described studies, visceral dissemination was no longer demonstrable (15), the rabbits were challenged i.d. with doses ranging from 10⁷ to 10⁸ B. burgdorferi depending on the experiment. In addition, naive rabbits were inoculated in the same manner to serve as controls.

Skin biopsy and culture. Skin punch biopsies were obtained from all experimental and control rabbits at various time points postinfection (p.i.) and postchallenge (p.c.). Rabbits were anesthetized by intramuscular injection with 45 mg of Ketaset (Fort Dodge Laboratories, Fort Dodge, Iowa) and 8.8 mg of Xylazine (Lloyd Laboratories, Shenandoah, Iowa) per kg of body weight. A 4- to 5-mm sterile punch biopsy (Baker and Cummings, Lakewood, N.J.) was taken adjacent to the inoculation site from the clipped back of each rabbit. Each biopsy specimen was immediately minced and cultured in 5 ml of BSK II medium containing 100 µg of phosphomycin per ml and 50 µg of rifampin per ml (Sigma, St. Louis, Mo.).

Preparation of polyclonal mouse anti-B. burgdorferi immunoglobulin G (IgG) for immunohistochemical detection of B. burgdorferi in skin sections. Virulent B. burgdorferi was grown as described above to passage 3. Organisms were washed three times by suspension in PBS followed by centrifugation at 8,000 × g for 10 min. After the final wash, organisms were resuspended in PBS to a final volume containing 10¹⁰ B. burgdorferi per ml. The organisms were then disrupted by ultrasonic treatment and stored at 80°C until ready for use.

Immunohistochemical staining of skin biopsies. Skin punch biopsies of 4 to 5 mm were obtained from all experimental and control rabbits at various time points p.i. and p.c. The punch biopsies were embedded in O.C.T. compound (Sakura Finetek, U.S.A., Inc., Torrance, Calif.) and quick-frozen in a dry-ice-2-methylbutane bath. Skin punch biopsies from naive rabbits were also obtained and frozen in O.C.T. compound to serve as negative controls for immunohistochemical staining. Frozen O.C.T.-embedded skin specimens were kept at 80°C until ready for use.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Detection of B. burgdorferi and characterization of the cellular infiltrate during EM in the skin of naive rabbits following infection. To identify and characterize the cellular immune response in the development and resolution of EM, rabbits were infected i.d. with 10⁷ B. burgdorferi, and representative skin biopsies were obtained at the time of EM development (day 7 p.i) and EM resolution (day 14 p.i.). Frozen serial sections of skin biopsies were analyzed for the presence of B. burgdorferi, B cell, T cells, and CD11b⁺ PMNs or macrophages by immunohistochemical staining. In addition, skin biopsy samples were analyzed for the presence of B. burgdorferi by culture in BSK II medium. As presented in Table 1, all rabbits developed culture-positive EM lesions at day 7 p.i. At day 14 p.i., EM lesions had resolved, but all animals remained skin culture positive. Histological examination showed the presence of intact spirochetes at day 7 which increased in relative numbers by day 14 p.i. (Fig. 1). At the time of the appearance of EM lesions that contained intact spirochetes (day 7 p.i.), there was predominantly a lymphocytic B-cell infiltration with fewer T cells (Fig. 2). By comparison, CD11b⁺ cells representing PMNs were only occasionally detected in these sections (Table 1). The type and relative quantity of the cellular infiltrate did not differ at day 14 (data not shown) from that of day 7 even though the number of B. burgdorferi was greater at day 14. 

View larger version (143K): [in this window] [in a new window]   FIG. 1.   Immunohistochemical staining for the detection of B. burgdorferi in skin biopsies from infected naive rabbits at the time of EM development (7 days p.i.) (a) and at the time of EM resolution (14 days p.i.) (b). Magnification, ×78.

View larger version (89K): [in this window] [in a new window]   FIG. 2.   Immunohistochemical staining of skin biopsies from infected naive rabbits at the time of EM development (7 days p.i.). Serial skin sections were stained for B. burgdorferi (a), B-cell lymphocytes (b), and T-cell lymphocytes (c). Magnification, ×125.

Skin histopathology in rabbits during initial infection and through the development of immunity. Based on previous studies, rabbits infected i.d. 5 months earlier with B. burgdoreri B31 develop complete immunity to challenge reinfection (15). To investigate the role of cellular immunity in the progression and resolution of disease, rabbits were infected i.d. with 6 × 10⁶ to 9 × 10⁶ B. burgdorferi, and representative skin biopsies were obtained at 3 and 13 weeks p.i. and again at 24 weeks p.i. when immunity was established. Frozen serial sections of skin biopsies were analyzed by immunohistochemical staining, as described above, and skin biopsy samples were cultured in BSK II medium for detection of viable B. burgdorferi. At 3 weeks p.i., as shown in Table 2, 80% of rabbits had culture-positive skin. Immunohistochemical detection showed that, compared to day 14 p.i. (Table 1), the quantity of spirochetes, B cells, and T cells had decreased. By 13 weeks p.i. as well as by 24 weeks p.i., spirochetes were no longer histochemically detectable nor could they be cultured from the skin (Table 2). In addition, lymphocytes were also not detected from skin sections; only rarely were macrophages detected at week 13 p.i. These data are consistent with our previous observations demonstrating that by 5 to 6 months p.i., skin and visceral infection in untreated rabbits is cleared (15).

Comparative histopathology in the skin immediately following challenge of naive and infection-immune rabbits. To examine early cell mediated immune events following challenge of immune animals, infection-derived immune rabbits previously infected with 8 × 10⁷ B. burgdorferi were challenged i.d. with 8 × 10⁷ organisms, and representative skin biopsies were obtained at 2, 12, 24, 36, 48, 72, and 96 h p.c. To compare these early events to those which occur in the nonimmune animal and to more thoroughly investigate immune mechanisms immediately following infection, naive rabbits were inoculated in the same manner, and representative skin biopsies obtained at the same time points as above. Serial sections of skin biopsies were analyzed by immunohistochemical staining and for the presence of viable B. burgdorferi by culture in BSK II medium as before.

View larger version (150K): [in this window] [in a new window]   FIG. 3.   Immunohistochemical staining of skin biopsies from infected naive rabbits 36 h p.i. Serial skin sections were stained by H & E (a and b), for B. burgdorferi (c and d), and for the CD11b PMN/macrophage marker (e and f). Magnification is at ×31 (a, c, and e) and ×125 (b, d, and f) from an outlined area of panels a, c, and e, respectively.

View larger version (169K): [in this window] [in a new window]   FIG. 4.   Immunohistochemical staining of skin biopsies from infection-immune rabbits 24 h after i.d. challenge with B. burgdorferi. Serial skin sections were stained by H & E (a and b), for B. burgdorferi (c and d), and for the CD11b PMN/macrophage marker (e and f). Magnification is at ×31 (a, c, and e) and ×125 (b, d, and f) from an outlined area of a, c, and e, respectively. Note that similar results were obtained from serial skin sections of biopsies taken at 2 and 12 h p.c. (data not shown).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study we compared the cellular events, including the morphological detection of spirochetes, in the skin of naive and infection-derived immune rabbits following i.d. challenge with B. burgdorferi B31. The observations made encompass the time from inoculation to the development of EM and through the development of challenge immunity. In the early events following the infection of naive animals, morphologically identifiable spirochetes were readily detected at 2 h and through 3 weeks p.i., which paralleled the demonstration of culture-positive skin infection during this time. We have recently reported, using quantitative PCR, that over 10⁸ borreliae are present per cm² of EM biopsied skin (36), which indicates that the rabbit is a very permissive host during the early course of infection. At 13 weeks p.i., however, skin cultures were consistently negative and no intact spirochetes could be detected by immunohistochemical staining. By comparison, skin samples from infection-immune animals following challenge never showed intact spirochetes; only B. burgdorferi antigen was detected at 2, 12, and 24 h p.c., which could no longer be detected at 36 h. In addition, culture-positive samples were no longer demonstrable from these immune rabbits after the 36-h time point. These findings demonstrate that the infection-derived immunity that develops in experimental rabbit Lyme disease rapidly destroys spirochetes at the site of challenge, as evidenced by the loss of spirochetal morphology and viability.

In order to determine whether specific cell-mediated immune mechanisms contribute to the immunity against challenge that develops in rabbits, immunohistochemical analysis to identify infiltrating PMNs, macrophages, T cells, and B cells was performed. In parallel, the same analysis of naive rabbits following challenge was performed for comparison to that of the immune animals and to characterize the cellular response that occurs from the start of infection to the development of immunity. It was observed that immediately following challenge of either naive or infection-immune rabbits an influx of PMNs occurred and were most prominently present at 12 and 24 h. This PMN infiltration was subsequently replaced by only a sparse infiltrate of macrophages, which in the naive animals persisted for up to 13 weeks p.i. In immune animals, the macrophages were absent by 72 h, a finding concordant with the disappearance of spirochetal antigen. T and B lymphocytes were detected during the time when EM lesions appeared in infected naive animals (day 7 p.i.) but also following the time that EM lesions had resolved (day 14 p.i.). After 13 weeks p.i., T and B lymphoctytes were no longer detectable. By comparison, T and B lymphocytes were never detected in the skin of challenged infection-immune challenged animals. Taken together, these findings show that in naive animals, the resolution of local skin infection correlates with the infiltration of T and B lymphocytes as well as macrophages, suggesting a primary role for cell-mediated immune mechanisms at the time of borrelial clearance. The early inflammatory response of PMNs and macrophages does not appear to affect the initial establishment and persistence of spirochetal infection. By comparison, no correlation was seen between the disappearance of viable spirochetes and a cellular infiltrate of any particular immune cell type in the immune rabbits following challenge, suggesting that humoral immunity plays the primary role in the persistence of acquired resistance. This is consistent with our recent study showing that passive immunization with serum from infection-immune rabbits can completely protect naive rabbits against challenge with large numbers of virulent B. burgdorferi (12). Thus, both cell-mediated and humoral immune mechanisms would appear to be important in the development and persistence of acquired resistance that results during the course of experimental rabbit Lyme disease.

In contrast to the rabbit model of Lyme disease, where infected animals ultimately clear the infection and develop immunity to challenge reinfection, infection in mice and humans is chronic. Thus, a comparison of the immune response that occurs following infection in mice with those in rabbits may provide insights into potential differences that account for the development of immunity versus that of chronic infection. In the mouse model, the host-immune response to B. burgdorferi infection has been studied primarily in BALB/c and C3H strains, the latter noted to be more permissive to B. burgdorferi infection. The humoral arm of the murine immune response has been thought to play a major role in the relative resistance of strains to Lyme disease while the cellular arm has been demonstrated to influence the outcome of disease, including the severity of arthritis (23). Following infection, BALB/c mice, the strain relatively more resistant to Lyme disease infection, are known to express an early predominant Th2 response (humoral immunity), whereas the more susceptible C3H mouse strain has an early predominant Th1 response (cell-mediated immunity) (20, 21, 22). B. burgdorferi-infected C3H mice, compared to infected BALB/c mice, harbor an increased spirochete burden in their joints and concomitantly show a more severe level of arthritis. It has been suggested that differential cytokine production may contribute to the mouse strain-related differences in susceptibility to Lyme disease (22). In this regard, it has been shown that the ablation of the endogenous cytokine interleukin-12 (IL-12), which helps to promote the development of a Th1 response, results in a reduced level of arthritis. Conversely, the ablation of endogenous IL-6, which helps to promote the Th2 response, results in an increased level of arthritis (4). These findings suggest that, in the murine model, humoral mechanisms are primarily responsible for the relative resistance to infection, while a predominant cell-mediated immune response is associated with increased disease severity. In contrast to these studies, Zeidner and colleagues (47) have shown that when susceptible mice are infected via the natural tick vector (Ixodes scapularis), exogenously administered Th1 cytokines reduce the severity of infection. Thus, the difference between chronic infection in mice and complete immunity in rabbits may lie with differences in the specific B. burgdorferi molecules recognized by the humoral and cell-mediated arms of immune response during infection.

We have recently shown that immune rabbits, but not rabbits hyperimmunized with purified outer membrane from in vitro-cultivated borreliae, are completely protected following challenge using organisms acquired from infected rabbit skin (36). We demonstrated that these organisms no longer express OspA and upregulate proteins that have been associated with host adaptation, such as OspC. These findings suggest that upregulated and/or host-adapted borrelia specific molecules are primarily responsible for the high degree of immunity that results following rabbit infection (36). Several proteins have been identified previously in the murine model as being upregulated during infection, including DbpA (10), OspC (35), and OspE and OspF (44). Proteins identified as uniquely expressed during infection include EppA (11), p35 and p37 (14), the OspE-F homologue p21 (45), the OspF homologues bbk2.1 (3), pG (46), and the proteins encoded on operon 2.9-7lpB (2). Many of these proteins, however, have not shown great potential as protective immunogens in the mouse model as a result of either incomplete or no protection (14, 17) or due to antigenic heterogeneity between strains which does not result in heterologous cross-immunity (9, 19). Similarly, we have found in the rabbit model no correlation between challenge immunity and antibody detected against some of these proteins, including DbpA and OspC (12, 36). However, the finding in this study that rabbit skin infection results in large numbers of structurally intact borrelia, which we have found to be host-adapted (36), provides a rich source for identifying potentially new candidate protective immunogens. Indeed, recent antigenic analysis of extracted infected rabbit skin tissue has revealed several novel and strongly antigenic molecules not detected using cultivated borreliae (J. N. Miller and M. A. Lovett, unpublished data). We are currently studying the role these antigens may play in the clearance of spirochetes during skin infection and the development of protective immunity in the rabbit model of Lyme disease.