INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Cowdria ruminantium is a tick-borne intracellular rickettsial agent that causes heartwater, an economically important infectious disease of ruminants in sub-Saharan Africa and in certain Caribbean islands (32). Transmitted primarily by Amblyomma hebraeum and A. variegatum ticks, the organism preferentially infects neutrophils, vascular endothelial cells, and monocytes. Animals that recover from the disease are protected against subsequent homologous challenge (32) by as yet undefined mechanisms of the immune system. Current vaccination strategies are based on the initiation of infections with blood- or culture-derived stabilates followed by treatment with tetracyclines (33). This method is hampered by its dependence on a requirement for effective freezing facilities during storage and transportation as well as by difficulties associated with subsequent control of the vaccine infection. The development of a more practical second-generation vaccine against cowdriosis based on subunit components of the agent is therefore considered important for the future control of the disease. A clear understanding of the immune mechanisms responsible for protection of ruminants against the agent is an important prerequisite for the achievement of this goal.

Available information on immunity to C. ruminantium is derived largely from studies of mouse models (8, 9, 12, 13) and does not provide a clear definition of the basis of protection in ruminants. Specific antibody responses are detected in ruminants as well as in mice following recovery from infection, but the results of serum transfer experiments suggest that these play a minor role, if any, in protection (12), although such sera can neutralize C. ruminantium infection in vitro (8). Cell-mediated immune mechanisms are important for protection against other rickettsial infections, with both cytotoxic T cells (10) and T-cell-derived cytokines, particularly gamma interferon (IFN-) (17), being implicated. Studies in the mouse and in ruminants have indicated that cell-mediated immunity is important in protection against C. ruminantium. Transfer of immune T cells protects mice against challenge (13). Additionally, athymic mice fail to develop immunity once vaccinated by the infection-and-treatment method, while in vivo depletion of T cells by intravenous inoculation with anti-Thy 1.2 monoclonal antibody (MAb) abrogated the immunity to C. ruminantium in mice (9). Independent groups have demonstrated that leukocyte-derived factors from cattle inhibit the growth of the agent in vitro (23, 29). One of these inhibitory factors has been identified as IFN- (26, 29). These observations support a role for T cells in ruminant immunity to C. ruminantium, but no information is currently available on the kinetics of these T-cell responses during infection.

Endothelial cells have been shown to be capable of presenting antigens to immune T cells in other systems (34). Infection of bovine brain endothelial cells with C. ruminantium also induces cytokine production (4). Because they have the potential of being major antigen-presenting cell populations in C. ruminantium-infected cattle, we have used autologous infected endothelial cells and monocytes to investigate the immunological mechanisms, in particular, the role of T-cell responses, in animals immune to C. ruminantium infection. We report here that T lymphocytes from immune cattle proliferate when cocultured with these cells, suggesting that they can be used to identify protective antigens of C. ruminantium for further exploitation in the development of a subunit vaccine for heartwater.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals and autologous endothelial cell lines. Male Ayrshire calves aged 8 to 10 months were used. The calves were reared in a heartwater-free area and were seronegative for C. ruminantium-specific antibodies by immunoblot analysis (22) at the outset of the study. Bovine testicular vein endothelial cell lines (EC) were established from each of the experimental animals as described previously (7) with some modifications. Briefly, the testicular vein was collected from each animal after unilateral castration and placed in Ca²⁺- and Mg²⁺-free Dulbecco's phosphate-buffered saline (pH 7.2) (PBS) supplemented with 400 IU of penicillin per ml, 400 µg of streptomycin per ml, and 2.5 µg of amphotericin B (Fungizone; Squibb, Islando, Tvl, South Africa) per ml (rinse buffer). The vessel was then slit longitudinally and washed twice before being cut into 1-cm² pieces, which were placed lumen side down on a drop of collagenase (1 mg/ml in rinse buffer) and incubated for 1 h at 37°C. The resulting cell suspension was centrifuged for 5 min at 200 × g and resuspended in 24 ml of Dulbecco's minimal essential medium (Gibco BRL, Grand Island, N.Y.) supplemented with 10% heat-inactivated fetal bovine serum (Hyclone Laboratories Inc.), 200 IU of penicillin per ml, 150 µg of streptomycin per ml, 2.5 mg of Fungizone per ml, 300 µg of endothelial cell growth supplement (no. E2759 Sigma, St. Louis, Mo.) per ml, and 2 mM L-glutamine. A 1-ml volume of the cell suspension was then seeded in each well of a 24-well tissue culture plate (Costar, Cambridge, Mass.). Once confluent, monolayers in each well were detached by treatment with 0.25% trypsin-EDTA solution containing 2.5 mg of trypsin per ml and 0.2 mg of EDTA per ml in Hanks balanced salt solution (HBSS; Sigma) and passaged into 25-cm² tissue culture flasks. The cells were maintained in complete Dulbecco's minimal essential medium and used for C. ruminantium infection between passages 4 and 10.

C. ruminantium culture and antigen preparation. Two strains of C. ruminantium isolated from Plumtree and Mbizi in Zimbabwe (23) were used for the study. The strains were stored as culture-derived stabilates in liquid nitrogen and propagated in vitro in monolayers of an endothelial cell line (BPA 593) derived from a bovine pulmonary artery grown in Glasgow minimal essential medium (GIBCO) supplemented with 10% tryptose phosphate broth, 10% heat-inactivated fetal bovine serum, 200 IU of penicillin per ml, 150 µg of streptomycin per ml, 2 mM L-glutamine, and 20 mM HEPES buffer (pH 7.2) (7). For the preparation of whole C. ruminantium antigen, elementary bodies (EBs) were harvested from terminally infected cultures of bovine EC. Host cell debris was removed by centrifugation at 400 × g for 5 min at 4°C. The supernatants were then centrifuged at 30,000 × g (Beckman model J-21B, rotor type JA20) for 30 min at 4°C, and the pellet was resuspended, washed three times in PBS, and frozen at 20°C as C. ruminantium EB antigen. The protein content of the antigen preparation was determined by the Bradford method (Pierce, Rockford, Ill.).

Immunization of animals. Four calves were infected by intravenous inoculation of 5 ml of culture supernatant containing 10⁸ viable EBs of the Plumtree isolate. Immunity in these calves was confirmed by rechallenge with 10⁸ homologous culture-derived EBs at least 4 weeks after the initial exposure. An additional calf was immunized by application of 20 A. hebraeum ticks infected with C. ruminantium Mbizi, produced in the laboratory by feeding uninfected nymphs on an infected sheep as described previously (25). This animal was treated on the days 2 and 3 of fever by intravenous injection of 10 mg of oxytetracycline per kg of body weight, and immunity was confirmed by rechallenge with infected ticks 6 weeks later. During rechallenge, naive control animals were also included to prove that the challenge was viable.

Isolation of PBMC and monocytes. Peripheral blood mononuclear cells (PBMC) were prepared by flotation of jugular venous blood collected in Alsever's solution on Ficoll-Paque (Pharmacia, Uppsala, Sweden). Monocytes were separated from PBMC by adherence to polystyrene as specified in published protocols (14). Briefly, 15 ml of a suspension of PBMC in culture medium containing 5 × 10⁶ cells/ml was placed in a 75-cm² culture flask (Costar) and incubated for 2 h at 37°C. After incubation, the flasks were shaken gently and nonadherent cells were removed by pipetting and rinsing with warm (37°C) RPMI 1640 medium (GIBCO). Adherent cells were removed with PBS containing 0.02% EDTA and washed twice in medium by centrifugation for 10 min at 200 × g. All centrifugations were carried out at 4°C in 10-ml polycarbonate tubes. Monocytes were infected with culture-derived C. ruminantium in the same manner as for EC.

Production and use of TCGF. T-cell growth factors (TCGF) were supplied as supernatants of bovine PBMC cultured for 18 h in the presence of 2.5 µg of concanavalin A (ConA) per ml. Uninfected bovine EC were cultured in medium containing 10% TCGF with 0.1 M methyl--mannoside (Sigma) and incubated for 48 h to induce class II major histocompatibility complex (MHC) antigens (29). EC monolayers were rinsed three times in PBS to remove residual TCGF prior to C. ruminantium infection. Monocytes were not pretreated with TCGF since they constitutively express class II MHC.

Preparation of monocytes and endothelial cells as antigen-presenting cells. Infected and uninfected autologous EC and monocytes were fixed in 0.1% glutaraldehyde as described previously (27). This was done to overcome nonspecific inhibition of proliferative responses observed when unfixed irradiated infected EC were used as antigen-presenting cells. Briefly, the cells were washed three times in cold HBSS and resuspended in 100 µl of the same buffer before addition of an equal volume of HBSS containing 0.1% glutaraldehyde. After approximately 30 s on ice, the fixative was quenched by addition of 200 µl of 0.2 M lysine in HBSS. Fixed cells were sedimented by centrifugation at 200 × g for 5 min, washed twice in culture medium, and resuspended in complete medium.

Generation of C. ruminantium-specific T-cell lines. T-cell lines specific for C. ruminantium-infected monocytes or EC were established from PBMC of immunized cattle. Briefly, 5 × 10⁶ PBMC were stimulated in 24-well plates (Costar) by 2.5 × 10⁵ autologous infected EC or monocytes. After 5 days, viable cells were isolated by flotation on Ficoll-Paque and restimulated under the same conditions but in the presence of 2.5 × 10⁵ irradiated (3,000 rads) autologous PBMC per well as fillers. Cell lines were maintained through weekly restimulations and were periodically evaluated for antigen specificity.

Lymphocyte proliferation assays. C. ruminantium-specific proliferation of PBMC from naive and immunized cattle was assayed in a total volume of 200 µl containing 5 × 10⁵ cells seeded in flat-bottom 96-well plates (Costar). After isolation, PBMC were washed three times and resuspended in HEPES-free RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 50 µM 2-mercaptoethanol, 200 IU of penicillin per ml, and 150 µg of streptomycin per ml. Where T-cell lines were assayed, 5 × 10⁴ cells were added to each well. Cultures of PBMC or T-cell lines received 2.5 × 10⁴ and 2.5 × 10³ monocytes or EC respectively, as stimulators. Anti-class II MHC MAb J11 was added to some cultures at a final dilution of 1:500 to determine the MHC restriction of the responding T-cell populations. MAb J11 (immunoglobulin G1 [IgG1]) recognizes a monomorphic determinant on bovine class II MHC (2). C. ruminantium EB antigen was added to a final concentration of 10 µg/ml. Cultures were incubated at 37°C in a humidified atmosphere of 5% CO₂ in air for 5 days in cultures containing EB antigen and when EC were used as stimulators and 3 days when monocytes were were used as stimulators. Proliferation was assessed by the addition of 0.5 µCi of [¹²⁵I]iododeoxyuridine (Amersham International, Little Chalfont, United Kingdom) to each well for the last 8 h of the assay and measuring the incorporated radioactivity with a gamma counter. Results are expressed as mean counts per minute (cpm) of triplicate cultures.

Phenotypic analysis. The cell surface phenotypes of responding cell populations in PBMC and T-cell line proliferation assays were determined by indirect immunofluorescence staining and analysis with a FACscan (Becton-Dickinson, Sunnyvale, Calif.) as described previously (19). MAbs defining bovine leukocyte populations comprised MM1A (CD3) (11), IL-A12 (CD4) (1), IL-A51 (CD8) (20), IL-A30 (bovine IgM) (35) and GB21A ( T-cell receptor) (21). Fluorescein isothiocyanate-conjugated swine anti-mouse Ig (Southern Biotechnology Associates, Inc.) was used as the secondary reagent.

Cytokine mRNA and IFN- protein assay. Total cellular RNA was prepared from 5 × 10⁶ cells of each responding PBMC or T-cell line after 3, 6, 24, and 48 h of antigen stimulation. The cells were pelleted by centrifugation at 300 × g for 10 min and washed once in PBS. Cell pellets were lysed with 1 ml of RNAzol (Biogenesis, Poole, England), and RNA was extracted from the lysate as described by the manufacturers. As a positive control, RNA was prepared from bovine PBMC stimulated with ConA (Sigma) for 6 and 48 h. The resultant RNA was free of genomic DNA as ascertained by analysis on agarose gels. First-strand cDNA was synthesized from 1 µg of total cellular RNA in 20-µl reaction mixtures with avian myeloblastosis virus reverse transcriptase in the presence of RNase inhibitor, using the Promega reverse transcription (RT) system as specified by the manufacturer. Expression of interleukin-1 (IL-1), IL-2, IL-4, IL-10, IFN-, tumor necrosis factor alpha (TNF-), TNF-, and the  chain of the IL-2 receptor (IL-2R) was detected by PCR amplification of cDNA with a panel of oligonucleotide primers (Table 1) in multiplex reactions. Expression of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (G3PDH) by each cell population was also evaluated to ensure mRNA integrity. Amplifications were performed on 1/10 of the RT reaction product in a final volume of 100 µl of PCR buffer (50 mM KCl, 10 mM Tris-HCl [pH 9.0] at 25°C, 1.5 mM MgCl₂, 10 IU of Taq polymerase [Promega], 50 ng of each primer, 0.1% [wt/vol] Triton X-100, deoxynucleoside triphosphates at 2.5 µM each). The mixtures were amplified for 30 cycles of incubation at 93°C for 1 min, 55°C for 1.5 min, and 72°C for 2 min followed by a final extension at 72°C for 5 min in a PTC-100 programmable thermal controller (MJ Research, Inc.). The conditions selected resulted in amplification of products in a linear range. After electrophoresis in 2% agarose gels (100 ng of cDNA per lane), the PCR products were visualized by staining with ethidium bromide.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Clinical responses of cattle during immunization and challenge with C. ruminantium. All animals infected with the culture-derived Plumtree strain of C. ruminantium developed a febrile reaction on day 7 after infection and then recovered without treatment. The animal infected by application of ticks infected with the Mbizi strain of C. ruminantium developed febrile reactions 24 days later and was treated on days 2 and 3 of fever (Table 2). None of the immunized cattle developed febrile reactions following subsequent rechallenge. In contrast, all naive challenge control animals became febrile on day 8 after culture-derived infection or day 18 after tick-mediated infection (data not shown).

Lymphocyte proliferative responses of immunized cattle to C. ruminantium antigen and to autologous infected endothelial cells. PBMC from five immune animals showed limited proliferation in response to whole C. ruminantium antigen (Fig. 1A). In addition, preliminary experiments indicated that irradiated infected autologous EC were poor stimulators of immune PBMC (Fig. 1B). To enhance the stimulatory capacity, autologous EC were treated with 10% TCGF to induce surface expression of class II MHC molecules. PBMC from immune but not naive animals proliferated in the presence of autologous EC infected with C. ruminantium after this treatment and fixed with glutaraldehyde (Fig. 1C). This response varied in magnitude among animals and was partially blocked by inclusion of the anti-class II MHC MAb, J11. The strongest responses were observed in animals BM 308, BM 307, and BM 309, which were immunized by needle infection with the Plumtree strain of C. ruminantium. The response in primary and secondary cultures was characterized by an increase in the proportions of CD4⁺ T cells and T cells. The proportions of B cells increased only marginally during the primary response and then gradually decreased with subsequent restimulations. After multiple stimulations, the cultures became dominated by T cells (Table 3).

View larger version (16K): [in this window] [in a new window]   FIG. 1.   In vitro response of bovine PBMC to autologous C. ruminantium-infected endothelial cells. (A) Responses to whole C. ruminantium EB antigen. MED, medium. (B) Responses to irradiated TCGF-treated C. ruminantium-infected endothelial cells. (C) Responses to glutaraldehyde-fixed TCGF-treated C. ruminantium-infected endothelial cells. Values are mean counts per minute (± standard error) of triplicate cultures.

Proliferative responses of PBMC from immunized cattle to autologous infected monocytes. Significant proliferative responses were also observed when PBMC from immune animals were cocultured with glutaraldehyde-fixed autologous infected monocytes (Fig. 2). These primary responses were partially blocked by the addition of MAb J11 and were phenotypically characterized by the presence of CD4⁺ T cells, T cells, and B cells. However, after two or three restimulations, cultures contained mainly CD4⁺ and T cells, with the latter predominating in subsequent restimulations (Table 3).

View larger version (33K): [in this window] [in a new window]   FIG. 2.   In vitro response of bovine PBMC to autologous glutaraldehyde-fixed C. ruminantium-infected monocytes. Values are mean counts per minute (± standard error) of triplicate cultures.

Proliferative responses of C. ruminantium-specific T-cell lines. T-cell lines generated by multiple stimulation with either autologous infected monocytes or EC were tested on a number of occasions for antigen-specific proliferation. The lines proliferated in response to autologous infected monocytes or EC, and proliferation was only marginally reduced by inclusion of the anti-class II MHC MAb J11 (Table 4).

Cytokine expression by immune PBMC and T-cell lines. RT-PCR analysis of cytokine gene expression in responding PBMC and T-cell lines generated with these antigen presentation systems revealed weak expression of IFN-, IL-2, and IL-4 mRNA at 3, 6, and 24 h after stimulation (Fig. 3). However, T-cell lines stimulated for 48 h showed strong expression of IFN-, TNF-, TNF-, and IL-2R, weak expression of IL-2 in one line, and no expression of IL-4 or IL-10 mRNA (Fig. 4). Analysis of culture supernatants by the bovine IFN- immunoassay confirmed that these cell lines produced significant amounts of the protein (Table 5).

View larger version (30K): [in this window] [in a new window]   FIG. 3.   Kinetic analysis of cytokine profiles of immune PBMC and T-cell lines by RT-PCR. Total RNA was prepared from PBMC stimulated with either autologous C. ruminantium-infected EC or monocytes (M) for 3 and 6 h. Similarly, T-cell lines (TC) stimulated with either autologous C. ruminantium-infected EC or monocytes for 6 and 24 h were used for RNA preparation to evaluate IFN-, IL-4, and IL-2 expression. Normal PBMC stimulated with ConA for 6 h (ConA Blasts) served as a positive control. Marker sizes are in base pairs.

View larger version (57K): [in this window] [in a new window]   FIG. 4.   RT-PCR analysis of cytokine mRNA in T-cell lines stimulated for 48 h. (A) Stimulation with autologous C. ruminantium-infected monocytes (M). Lanes 2, 3, and 4 contain primers for G3PDH, IL-1, and IL-4, while lanes 5, 6, and 7 contain primers for G3PDH, IL-2R, IL-10, IL-2, IFN-, TNF-, and TNF-. (B) Stimulation with autologous C. ruminantium-infected EC. F100 ConA blasts and PCR mixture with no template were used as positive and negative controls, respectively. Marker sizes (M) are in base pairs.

View this table: [in this window] [in a new window]   TABLE 5.   Production of IFN- by C. ruminantium-specific T-cell lines cultured in the presence of autologous infected endothelial cells or monocytes

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We have demonstrated that immunization of cattle against heartwater by infection and treatment results in the generation of T lymphocytes that recognize infected EC and monocytes but in only low-grade proliferation in response to whole C. ruminantium EB antigen. Recognition of infected EC is dependent on induction of class II MHC expression by treatment with TCGF before infection and fixation with glutaraldehyde before inclusion in proliferative assays. These studies have established essential parameters for examining the nature of the cellular immunity engendered in cattle by infection with C. ruminantium. Our failure to detect marked T-cell responses to C. ruminantium EB antigen in vitro contrasts with the report that such responses are observed in animals immunized with EB lysate formulated in complete Freund's adjuvant (30). This suggests that antigen processing and presentation by infected cells may be essential for the induction of T-cell responses to the agent during infection. The partial blockage of these responses by anti-class II MHC MAb J11, together with the phenotypic profiles of the responding PBMC, indicates that this response is partly mediated by class II MHC-restricted CD4⁺ T cells. A role for CD4⁺ T cells in protection of cattle immunized against heartwater with inactivated C. ruminantium has been suggested previously (24, 30). This population has traditionally been considered to be prominent in the regulation of cellular immune responses to other intracellular pathogens (18). However, T-cell responses to intracellular pathogens such as Mycobacterium, Listeria, and Salmonella are characterized by the activation of both (mainly CD4⁺) and T cells (31). The results of the present study suggest that both these populations are induced in cattle undergoing C. ruminantium infection.

Evidence is accumulating that T cells can contribute to protection in infections with some intracellular organisms, including Toxoplasma gondii and Mycobacterium tuberculosis (3). It has been argued that these cells represent a more primitive T-cell population that acts as a first line of defense against certain pathogens (16). In the mouse model of T. gondii, a T-cell response can be detected in vitro only if infected cells are used as stimulators in proliferation assays (28). Murine T-cell responses to T. gondii are characterized by the production of cytokines (IFN-, IL-2, and TNF-) and by cytotoxicity for infected cells and are believed to confer protection against the parasite (28). In our studies of C. ruminantium-immune cattle, multiple restimulation of T-cell cultures gave rise to an enrichment of T cells; indeed, most of the T-cell lines generated in the study were dominated by these cells. This is consistent with our failure to block the response by T-cell cultures by using anti-class II MAb J11. A similar scenario of T-cell outgrowth has been observed in analysis of bovine T-cell responses to Babesia bovis and Fasciola hepatica (5, 6), although it is not clear in these studies whether these cells have a role in protective immunity. Although it is possible that cultured T cells respond to an antigen(s) expressed by the C. ruminantium-infected cell, the observed outgrowth of these cells in T-cell lines upon continual restimulation could be attributed to the effect of TCGF. This would suggest that the T-cell response detected in these cultures may have no direct relevance to the in vivo situation. Nonetheless, the cytokine profiles of these cells (TNF-, IFN-, and TNF-) resemble those observed in T-cell responses to T. gondii infections and in tuberculosis, which may be consistent with a role in immunity or in the pathogenesis of the disease.

Intracellular survival and replication is a strategy adopted by many infectious agents for evading the immune response of the vertebrate host. C. ruminantium replicates primarily in EC and can also be demonstrated in monocytes and neutrophils. The inaccessibility of the organism to serum antibody during much of the infection period suggests that humoral responses play only a small role in protection. Indeed, it is possible that even EBs that become opsonized during the rickettsemic phases can survive after uptake by phagocytes. Our results suggest that the major cellular immune responses of cattle to C. ruminantium infection are in the CD4⁺ and T-cell compartments. It is not yet clear how these populations mediate their effects, but it is possible that it is through the elaboration of cytokines. Previous studies have shown that ConA lymphoblast supernatants inhibit the growth of C. ruminantium in bovine EC in vitro (23). The active cytokine in these supernatants was later identified as IFN- (26). Other workers have shown that recombinant IFN- can also inhibit Cowdria growth in vitro (29). IFN- has been reported to lyse rickettsial agents or cells infected with them (15), and this mechanism may also be responsible for its inhibition of C. ruminantium growth in vitro (23).

These observations suggest that new vaccine strategies for heartwater should focus on the immune responses that enhance production of Cowdria-inhibitory cytokines such as IFN- (26, 29). We have observed the induction of Cowdria-specific T-cell responses in immune cattle that respond to autologous infected monocytes and EC. In addition, responding cells express the Cowdria-inhibitory cytokine IFN-, in response to specific stimulation in vitro. It is therefore likely that protection against heartwater after immunization by infection and treatment or after spontaneous recovery is mediated, at least in part, by generation of C. ruminantium-specific T-cell responses resulting in the secretion of IFN-. Our results highlight the need to identify Cowdria antigens expressed on or secreted by EBs that are presented at the surface of infected monocytes or EC. The responding PBMC and T-cell lines generated in this study will be useful in identifying such antigens to allow their evaluation as potential candidates for subunit vaccines against cowdriosis.