Rapid and Long-Term Disappearance of CD4+ T Lymphocyte Responses Specific for Anaplasma Marginale Major Surface Protein-2 (MSP2) in MSP2 Vac
http://www.100md.com
免疫学杂志 2005年第11期
Abstract
In humans and ruminants infected with Anaplasma, the major surface protein 2 (MSP2) is immunodominant. Numerous CD4+ T cell epitopes in the hypervariable and conserved regions of MSP2 contribute to this immunodominance. Antigenic variation in MSP2 occurs throughout acute and persistent infection, and sequentially emerging variants are thought to be controlled by variant-specific Ab. This study tested the hypothesis that challenge of cattle with Anaplasma marginale expressing MSP2 variants to which the animals had been immunized, would stimulate variant epitope-specific recall CD4+ T cell and IgG responses and organism clearance. MSP2-specific T lymphocyte responses, determined by IFN- ELISPOT and proliferation assays, were strong before and for 3 wk postchallenge. Surprisingly, these responses became undetectable by the peak of rickettsemia, composed predominantly of organisms expressing the same MSP2 variants used for immunization. Immune responsiveness remained insignificant during subsequent persistent A. marginale infection up to 1 year. The suppressed response was specific for A. marginale, as responses to Clostridium vaccine Ag were consistently observed. CD4+CD25+ T cells and cytokines IL-10 and TGF-1 did not increase after challenge. Furthermore, a suppressive effect of nonresponding cells was not observed. Lymphocyte proliferation and viability were lost in vitro in the presence of physiologically relevant numbers of A. marginale organisms. These results suggest that loss of memory T cell responses following A. marginale infection is due to a mechanism other than induction of T regulatory cells, such as peripheral deletion of MSP2-specific T cells.
Introduction
Pathogens in the genus Anaplasma express immunodominant outer membrane proteins with defined conserved and variable domains (1, 2, 3, 4, 5, 6). Antigenic variation in Anaplasma marginale major surface protein 2 (MSP2)3 and in the orthologous MSP2/p44 protein of Anaplasma phagocytophilum, results in evasion of the immune response and has been postulated to be responsible at least in part, for persistent infection in mammalian reservoir hosts (6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20). Although the acute phase of A. marginale infection peaks with levels of 109 rickettsiae per milliliter of blood, persistent infection is characterized by recurrent subclinical cycles of rickettsemia that range from 103 to 107 organisms per milliliter (6, 7, 21, 22, 23). Each cycle of rickettsemia reflects the emergence of organisms that express antigenically variant MSP2 (7). Antigenic variation in MSP2, and in a related surface protein, MSP3, occurs by gene conversion of whole pseudogenes and small segments of pseudogenes into single expression sites, providing an efficient mechanism to generate the large number of variants seen during sequential cycles of persistent infection (8, 12, 13, 24).
The control of the sequential rickettsemic cycles during persistent infection is associated with development of a variant-specific IgG response and, in particular, IgG2 (7). In addition, MSP2 contains numerous MHC class II-restricted CD4+ T cell epitopes in both the highly conserved N- and C-terminal domains as well as in the variant-specific central hypervariable region (HVR) (10, 11, 25, 26). This rich source of epitopes may serve to induce T cell help for generation of variant-specific Ab and control of rickettsemic cycles during persistent infection. In recent studies, we have used cattle immunized with purified MSP2 to define both the T and B lymphocyte epitopes in a specific set of MSP2 variants (10, 26). This allowed us to control the variants used for challenge in the context of the continual generation of new variants that occurs during actual infection. T cell epitopes were also recently mapped in 16 MSP2 vaccinates representing 10 different MHC class II DRB3 alleles. IgG Ab responses were directed against epitopes predominantly located within the HVR, whereas CD4+ T cell proliferative and IFN- responses were directed against multiple epitopes evenly distributed in the highly conserved and hypervariable regions (25). In the present study, we address stimulation and maintenance of anamnestic responses by specific MSP2 variants following infection. This study tested the hypothesis that challenge of cattle with A. marginale expressing MSP2 variants to which the animals had been immunized, would stimulate variant epitope-specific recall CD4+ T cell and IgG responses and variant-specific organism clearance. In this paper, data are presented that support rejection of this hypothesis and, in contrast, demonstrate a newly discovered immune modulation whereby Ag-specific T cell responsiveness is lost upon rickettsial challenge.
Materials and Methods
Data analysis and statistics
All statistical tests were done with Number Cruncher Statistical Software (NCSS2001), version 2.00.0185. The proportion of the msp2 A variant in the organisms used to prepare MSP2 for immunization and for challenge was compared by the Fisher Exact test. Clinical parameters of the immunized groups were compared by Kruskal-Wallis rank sum analysis with the control group ( = 0.05). Correlation of IgG titers with clinical parameters was determined using multivariate linear regression and Spearman rank correlation. One-way ANOVA with Bonferonni correction for multiple comparisons ( = 0.05) was used to determine significant ELISPOT and proliferation responses as compared with medium, and to determine significant differences in CD25+ T lymphocyte populations. Paired two-tailed t tests were used to determine significance differences in cytokine levels between responding and nonresponding PBMC.
Results
Comparison of MSP2 variants in the MSP2 immunogen and challenge organisms
The relationship between msp2 variants in the A. marginale-infected blood used as a source of MSP2 immunogen and the msp2 variants expressed by the organisms in the blood used for challenge was determined by sequencing the msp2 single expression site in these organisms (Fig. 1). In the organisms used as a source of preparing MSP2, 10 of the 19 msp2 clones sequenced were variant A, 4 were variant C, and 5 additional minor variants were each represented by a single clone (Table I). The A. marginale used for challenge was also composed predominantly of the msp2 A variant (19 of 28 clones) and contained 6 additional minor variants (Table I). Although amplification of DNA by PCR to determine the relative frequency of msp2 variants may introduce bias, previous studies have shown a positive correlation in relative levels of msp2 variants amplified by PCR with levels of msp2 expression site genomic DNA, msp2 mRNA, and MSP2 protein (12, 36). In the present study, no statistically significant difference was found in the frequency of the msp2 A variant in the organisms used to prepare native MSP2 for immunization and those used for challenge (p = 0.365). The predicted amino acid sequences of the HVRs of all msp2 variants are shown in Fig. 1, A–C. Variants A, B, C, D, and E have been previously reported (GenBank accession nos. AY138954–AY138958) (10). Two minor msp2 variants (H and O) were present in the challenge organisms that were not found in the MSP2 immunogen (Fig. 1A), and two minor variants (C and E) were present in the immunogen but not detected in the challenge organisms (C).
Challenge of MSP2-immunized cattle with live A. marginale
Following challenge with 103 live A. marginale organisms, all animals had microscopically measurable rickettsemia by 25 DPI (Fig. 2). Ten of the 12 animals had levels of rickettsemia exceeding 108 infected erythrocytes per milliliter of blood at the peak of infection, 5–6 wk postchallenge. The remaining two animals, one from each vaccination group, had >107 infected erythrocytes per milliliter of blood. Clinical parameters including peak rickettsemia levels, days to peak rickettsemia, days to detection of rickettsemia, days to 108 infected erythrocytes per milliliter, and the duration of the rickettsemia peak varied from animal to animal, but there were no significant differences in these parameters of infection between immunized and control groups (Fig. 2).
Comparison of IFN- ELISPOT and proliferation assays to MSP2 and to Clostridium vaccine Ag
We recently showed a significant correlation between proliferation and IFN- ELISPOT responses by PBMC from these MSP2 vaccinates to A. marginale, MSP2, and MSP2-derived peptides (25). Nevertheless, to address the possibility that following A. marginale infection, the response changed from a predominant Th1-like response to a predominant Th2-like response that could not be detected by the IFN- ELISPOT assay, proliferation and IFN- ELISPOT assays were conducted simultaneously with the same aliquots of cells. Consistent with the results using the ELISPOT assay, proliferative responses to A. marginale, MSP2, and all MSP2-derived peptides were severely decreased in all eight vaccinates at the peak of infection and thereafter, whereas the response to the PHA, IL-12, and IL-18 was always observed (representative data for MSP2-immunized animal 71 are shown in Fig. 4, D–F).
To further determine whether the impaired T cell response to A. marginale and MSP2 was Ag specific or reflected a generalized immune suppression, Clostridium spp. Ag, which was used to vaccinate the calves before MSP2 immunization, was also included in the assays. Unlike the responses to A. marginale and MSP2, the response to Clostridium Ag was significant at all time points (Fig. 4). In addition, no significant responses to MSP2-derived peptides were seen (Fig. 4, E and F). Similar results were obtained for all immunized animals tested at various time points up to 7–12 mo postinfection (data not presented), and suggest that the impaired response to MSP2 is long-lived and does not reflect a generalized immune suppression.
Quantitation of CD25+CD4+ T cells by FACS
Regulatory T cells have not been defined in cattle. Nevertheless, to evaluate the potential role of CD25+CD4+ T regulatory cells in the loss of MSP2-specific responses following challenge, CD25+CD4+ T cells were analyzed in PBMC from immunized and control animals by two-color flow cytometry (Table IV). Although the differences in the percentage of CD25+CD4+ T cells in PBMC varied between individual animals, when data for all animals were compared, no significant differences in the percentage of CD25+CD4+ T cells, expressed as either a percentage of total CD4+ T cells or of total CD25+ T cells, were observed during the course of infection (data not shown).
Positively selected CD4+ T cells from peak rickettsemia do not respond to Ag cultured with APC from naive donors
To address the possibility that the inability to detect a CD4+ T cell response following challenge was caused by dysfunctional APC, CD4+ T cells were positively selected from PBMC of two animals (71 and 76) cryopreserved at time points when the cells responded or at peak rickettsemia when no response was detected. PBMC were stimulated with Ag in the presence of MHC class II DRB3 homozygous and half-matched APC from A. marginale naive donor cattle. CD4+ T cells obtained at time points where responses to Ag were previously observed had strong and significant proliferative responses to A. marginale and MSP2 in the presence of APC from naive donors (Table VII). In contrast, CD4+ T cells obtained at the peak of infection had undetectable proliferative responses to Ag, but did proliferate to TCGF, as observed when autologous APC were used to present Ag.
Discussion
The results of this study do not support our hypothesis that challenge with A. marginale expressing MSP2 variants to which the animals had been previously exposed would stimulate strong anamnestic CD4+ T cell and IgG responses directed against conserved and variant-specific epitopes that would be associated with variant-specific organism clearance. In contrast, analysis of the immune response upon challenge demonstrates a newly discovered modulation whereby Ag-specific T cell responsiveness induced by vaccination is lost upon rickettsial challenge.
Sequencing of msp2 transcripts from the blood of cattle obtained during ascending and peak rickettsemia ruled out the possibility that acute rickettsemia resulted from selective expansion either of organisms expressing variants of MSP2 that constituted a minor population in the challenge inoculum, or of organisms expressing novel msp2 sequences. This indicates that either the MSP2-specific immune response induced by immunization or the recall response elicited by the challenge was insufficient to effect clearance.
The lack of strong recall T cell responses to MSP2 following A. marginale infection may be related to the uniformly dramatic loss of MSP2-specific CD4+ T cell responses that occurred in all animals concurrently with development of measurable rickettsemia. However, the sustained Th cell response for the first 3 wk following challenge was apparently sufficient to stimulate a boost in IgG production. This infection-mediated immune modulation of a strong CD4+ T cell response directed against multiple antigenic epitopes (25) has not been previously described for any rickettsial pathogen. However, in other persistent pathogen infection models, anergy induced by altered peptide ligand antagonism (41, 42, 43), or by T regulatory cells producing either TGF-1 or IL-10 (37, 44, 45, 46, 47, 48, 49), has been shown to play a role in down-regulating T cell responses. Although antigenically variant MSP2 epitopes could potentially act as antagonistic peptides, previous studies did not show the ability of naturally occurring variant epitopes to cause anergy of MSP2-specific T cell lines or clones specific for the agonistic MSP2 variant (10). Furthermore, the disappearance of T cell responses to conserved MSP2 epitopes, as well as variable MSP2 epitopes, argues against antigenic variation in MSP2 as a reason for the abrupt loss of T cell responsiveness.
To address the possibility that A. marginale infection induced a T regulatory cell response, experiments were performed to determine changes in the percentage of CD25+CD4+ T cells during the course of infection, to examine IL-10 and TGF-1 production by responding and nonresponding cells, to detect the presence of a population of suppressive cells in peripheral blood by mixing responding and nonresponding cells, and to test positively selected CD4+ T cells. Although our results do not support the role of T regulatory cells in the dramatic loss of MSP2-specific CD4+ T cell immune responses, their role cannot be definitively ruled out, because these cells have not been phenotypically characterized in cattle. We were similarly unable to demonstrate a shift from a dominant IFN- Th1 response (28) response to an IL-4 dominant response following challenge.
To test the possibility that infection impaired APC to present A. marginale Ag to CD4+ T cells, positively selected CD4+ T cells were cultured with Ag in the presence of class II-compatible APC from noninfected donors. However, T cells obtained at the peak of infection were still unable to respond to Ag, ruling out dysfunctional APC as the reason for the sudden loss of response.
Sheep and dogs infected with Anaplasma phagocytophilum develop a transient immunosuppression defined by leukopenia (reduced numbers of T lymphocytes and neutrophils) and an increased susceptibility to other infectious organisms (50, 51, 52). A. phagocytophilum infects neutrophils and alters neutrophil function (53, 54, 55, 56, 57), which may explain the transient generalized immune suppression. However, a similar mechanism of generalized immune suppression by A. marginale is unlikely for the following reasons: 1) this pathogen infects erythrocytes and not neutrophils, 2) the response to unrelated clostridial Ags was not severely impaired during acute infection, and 3) increased susceptibility to unrelated or opportunistic infections has not been reported for cattle with anaplasmosis.
The unsubstantiated role of T regulatory cells in the disappearance of the MSP2-specific memory T cell response, the lack of evidence for altered Ag presentation, and the Ag-specific nature of the immune suppression suggest an alternative mechanism for the loss of T cell responsiveness. One potential mechanism is peripheral T cell deletion that could occur via activation-induced cell death (AICD) following organism challenge (58). During primary HIV infection, naturally induced HIV-specific CD8+ T cell clones with defined TCR V usage were shown to rapidly disappear, independent of changes in the viral epitopes recognized (59). An unrelated study reported in vivo elimination of Ag-specific Th1 cells, obtained from TCR transgenic mice that were adoptively transferred to normal mice, following i.v. challenge with the Ag cytochrome c 1–2 mo later (60). The Ag-specific memory T cells became rapidly activated in vivo upon Ag administration, but by day 8 following Ag challenge, declined to barely detectable numbers and remained depressed or anergic for 3 mo. The authors concluded that Ag challenge of resting Th1 CD4+ T cells led to transient activation followed by cell depletion. In our studies, A. marginale was administered i.v. and the infection took 5 wk to reach peak levels in peripheral blood. Thus, for the first 3 wk following challenge, recall T cell responses remained at prechallenge levels, but were completely undetectable at the peak of infection, 2 wk later. We therefore examined CD4+ T cell IFN- ELISPOT responses in six immunized cattle at 1 wk before the peak of infection was reached (29 or 31 DPI), and observed weakly positive responses in two animals and undetectable responses in four animals (data not shown). These results are consistent with Ag-induced AICD. Furthermore, A. marginale inhibited, in a dose-dependent manner, proliferation of lymphocytes that paralleled a loss in cell viability. A reduction in the response to TCGF from >50 to 100% was observed at organism concentrations equivalent to those observed at peak levels of rickettsemia in vivo following challenge, which ranged from 1 x 107 to 8 x 108 organisms per milliliter of blood (Fig. 2). However, the significance of these in vitro results to the in vivo infection is not clear, because A. marginale is generally intraerythrocytic and the nature of the interaction of infected erythrocytes with T cells is unknown. Nevertheless, these results are also consistent with a loss in immune responsiveness as a consequence of increasing Ag dose in vivo, and a mechanism of AICD.
In conclusion, we hypothesize that MSP2-specific memory T cells were deleted or decreased to undetectable numbers in animals following infection with A. marginale. Our data indicate a newly discovered immune modulation whereby Ag-specific T cell responsiveness is lost upon rickettsial challenge. MSP-2-specific T cells may be deleted as a consequence of high levels of Ag occurring during ascending rickettsemia, and the number of MSP2-specific T cells may remain depressed as a result of the chronic low antigenic exposure during persistent infection. Consistent with this is our inability to detect CD4+ T cell responses in nonimmunized cattle following either i.v. or tick-transmitted A. marginale challenge for up to 3 mo postinfection (this study and our unpublished observations). Additional experiments using MHC class II tetramers to track the fate of epitope-specific T cells in immunized and control cattle during acute and chronic anaplasmosis should clarify the mechanism for the loss of Ag-specific T cell responses following A. marginale infection and determine whether a similar immune modulation occurs during infection of nonvaccinated cattle.
Acknowledgments
We are grateful to Bev Hunter, Emma Karel, and Shelley Whidbee for excellent technical assistance, and Kelly Brayton and Travis McGuire for helpful discussions and assistance.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work is supported by National Institutes of Health Grants AI44005 and AI49276, and U.S. Department of Agriculture National Research Initiative Competitive Grants Program Grant 02-35204-12352.
2 Address correspondence and reprint requests to Dr. Wendy C. Brown, Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA 99164. E-mail address: wbrown{at}vetmed.wsu.edu
3 Abbreviations used in this paper: MSP2, major surface protein 2; HVR, hypervariable region; ODN, oligodeoxynucleotide; DPI, days postinfection; PCV, packed cell volume; URBC, uninfected bovine erythrocyte membrane; SFC, spot-forming cell; TCGF, T cell growth factor; AICD, activation-induced cell death.
Received for publication December 10, 2004. Accepted for publication March 17, 2005.
References
Zhi, N., N. Ohashi, Y. Rikihisa, H. W. Horowitz, G. P. Wormser, K. Hechemy. 1998. Cloning and expression of the 44-kilodalton major outer membrane protein gene of the human granulocytic ehrlichiosis agent and application of the recombinant protein to serodiagnosis. J. Clin. Microbiol. 36: 1666-1673.
Shkap, V., T. Molad, K. A. Brayton, W. C. Brown, G. H. Palmer. 2002. Expression of major surface protein 2 variants with conserved T-cell epitopes in Anaplasma centrale vaccinates. Infect. Immun. 70: 642-648.[
Palmer, G. H., J. R. Abbott, D. M. French, T. F. McElwain. 1998. Persistence of Anaplasma ovis infection and conservation of the msp-2 and msp-3 multigene families within the genus Anaplasma. Infect. Immun. 66: 6035-6039.[
Murphy, C. I., J. R. Storey, J. Recchia, L. A. Doros-Richert, C. Gingrich-Baker, K. Munroe, J. S. Bakken, R. T. Coughlin, G. A. Beltz. 1998. Major antigenic proteins of the agent of human granulocytic ehrlichiosis are encoded by members of a multigene family. Infect. Immun. 66: 3711-3718.
Ijdo, J. W., W. Sun, Y. Zhang, L. A. Magnarelli, E. Fikrig. 1998. Cloning of the gene encoding the 44-kilodalton antigen of the agent of human granulocytic ehrlichiosis and characterization of the humoral response. Infect. Immun. 66: 3264-3269.
French, D. M., T. F. McElwain, T. C. McGuire, G. H. Palmer. 1998. Expression of Anaplasma marginale major surface protein 2 variants during persistent cyclic rickettsemia. Infect. Immun. 66: 1200-1207.(Jeffrey R. Abbott*, Guy H)
In humans and ruminants infected with Anaplasma, the major surface protein 2 (MSP2) is immunodominant. Numerous CD4+ T cell epitopes in the hypervariable and conserved regions of MSP2 contribute to this immunodominance. Antigenic variation in MSP2 occurs throughout acute and persistent infection, and sequentially emerging variants are thought to be controlled by variant-specific Ab. This study tested the hypothesis that challenge of cattle with Anaplasma marginale expressing MSP2 variants to which the animals had been immunized, would stimulate variant epitope-specific recall CD4+ T cell and IgG responses and organism clearance. MSP2-specific T lymphocyte responses, determined by IFN- ELISPOT and proliferation assays, were strong before and for 3 wk postchallenge. Surprisingly, these responses became undetectable by the peak of rickettsemia, composed predominantly of organisms expressing the same MSP2 variants used for immunization. Immune responsiveness remained insignificant during subsequent persistent A. marginale infection up to 1 year. The suppressed response was specific for A. marginale, as responses to Clostridium vaccine Ag were consistently observed. CD4+CD25+ T cells and cytokines IL-10 and TGF-1 did not increase after challenge. Furthermore, a suppressive effect of nonresponding cells was not observed. Lymphocyte proliferation and viability were lost in vitro in the presence of physiologically relevant numbers of A. marginale organisms. These results suggest that loss of memory T cell responses following A. marginale infection is due to a mechanism other than induction of T regulatory cells, such as peripheral deletion of MSP2-specific T cells.
Introduction
Pathogens in the genus Anaplasma express immunodominant outer membrane proteins with defined conserved and variable domains (1, 2, 3, 4, 5, 6). Antigenic variation in Anaplasma marginale major surface protein 2 (MSP2)3 and in the orthologous MSP2/p44 protein of Anaplasma phagocytophilum, results in evasion of the immune response and has been postulated to be responsible at least in part, for persistent infection in mammalian reservoir hosts (6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20). Although the acute phase of A. marginale infection peaks with levels of 109 rickettsiae per milliliter of blood, persistent infection is characterized by recurrent subclinical cycles of rickettsemia that range from 103 to 107 organisms per milliliter (6, 7, 21, 22, 23). Each cycle of rickettsemia reflects the emergence of organisms that express antigenically variant MSP2 (7). Antigenic variation in MSP2, and in a related surface protein, MSP3, occurs by gene conversion of whole pseudogenes and small segments of pseudogenes into single expression sites, providing an efficient mechanism to generate the large number of variants seen during sequential cycles of persistent infection (8, 12, 13, 24).
The control of the sequential rickettsemic cycles during persistent infection is associated with development of a variant-specific IgG response and, in particular, IgG2 (7). In addition, MSP2 contains numerous MHC class II-restricted CD4+ T cell epitopes in both the highly conserved N- and C-terminal domains as well as in the variant-specific central hypervariable region (HVR) (10, 11, 25, 26). This rich source of epitopes may serve to induce T cell help for generation of variant-specific Ab and control of rickettsemic cycles during persistent infection. In recent studies, we have used cattle immunized with purified MSP2 to define both the T and B lymphocyte epitopes in a specific set of MSP2 variants (10, 26). This allowed us to control the variants used for challenge in the context of the continual generation of new variants that occurs during actual infection. T cell epitopes were also recently mapped in 16 MSP2 vaccinates representing 10 different MHC class II DRB3 alleles. IgG Ab responses were directed against epitopes predominantly located within the HVR, whereas CD4+ T cell proliferative and IFN- responses were directed against multiple epitopes evenly distributed in the highly conserved and hypervariable regions (25). In the present study, we address stimulation and maintenance of anamnestic responses by specific MSP2 variants following infection. This study tested the hypothesis that challenge of cattle with A. marginale expressing MSP2 variants to which the animals had been immunized, would stimulate variant epitope-specific recall CD4+ T cell and IgG responses and variant-specific organism clearance. In this paper, data are presented that support rejection of this hypothesis and, in contrast, demonstrate a newly discovered immune modulation whereby Ag-specific T cell responsiveness is lost upon rickettsial challenge.
Materials and Methods
Data analysis and statistics
All statistical tests were done with Number Cruncher Statistical Software (NCSS2001), version 2.00.0185. The proportion of the msp2 A variant in the organisms used to prepare MSP2 for immunization and for challenge was compared by the Fisher Exact test. Clinical parameters of the immunized groups were compared by Kruskal-Wallis rank sum analysis with the control group ( = 0.05). Correlation of IgG titers with clinical parameters was determined using multivariate linear regression and Spearman rank correlation. One-way ANOVA with Bonferonni correction for multiple comparisons ( = 0.05) was used to determine significant ELISPOT and proliferation responses as compared with medium, and to determine significant differences in CD25+ T lymphocyte populations. Paired two-tailed t tests were used to determine significance differences in cytokine levels between responding and nonresponding PBMC.
Results
Comparison of MSP2 variants in the MSP2 immunogen and challenge organisms
The relationship between msp2 variants in the A. marginale-infected blood used as a source of MSP2 immunogen and the msp2 variants expressed by the organisms in the blood used for challenge was determined by sequencing the msp2 single expression site in these organisms (Fig. 1). In the organisms used as a source of preparing MSP2, 10 of the 19 msp2 clones sequenced were variant A, 4 were variant C, and 5 additional minor variants were each represented by a single clone (Table I). The A. marginale used for challenge was also composed predominantly of the msp2 A variant (19 of 28 clones) and contained 6 additional minor variants (Table I). Although amplification of DNA by PCR to determine the relative frequency of msp2 variants may introduce bias, previous studies have shown a positive correlation in relative levels of msp2 variants amplified by PCR with levels of msp2 expression site genomic DNA, msp2 mRNA, and MSP2 protein (12, 36). In the present study, no statistically significant difference was found in the frequency of the msp2 A variant in the organisms used to prepare native MSP2 for immunization and those used for challenge (p = 0.365). The predicted amino acid sequences of the HVRs of all msp2 variants are shown in Fig. 1, A–C. Variants A, B, C, D, and E have been previously reported (GenBank accession nos. AY138954–AY138958) (10). Two minor msp2 variants (H and O) were present in the challenge organisms that were not found in the MSP2 immunogen (Fig. 1A), and two minor variants (C and E) were present in the immunogen but not detected in the challenge organisms (C).
Challenge of MSP2-immunized cattle with live A. marginale
Following challenge with 103 live A. marginale organisms, all animals had microscopically measurable rickettsemia by 25 DPI (Fig. 2). Ten of the 12 animals had levels of rickettsemia exceeding 108 infected erythrocytes per milliliter of blood at the peak of infection, 5–6 wk postchallenge. The remaining two animals, one from each vaccination group, had >107 infected erythrocytes per milliliter of blood. Clinical parameters including peak rickettsemia levels, days to peak rickettsemia, days to detection of rickettsemia, days to 108 infected erythrocytes per milliliter, and the duration of the rickettsemia peak varied from animal to animal, but there were no significant differences in these parameters of infection between immunized and control groups (Fig. 2).
Comparison of IFN- ELISPOT and proliferation assays to MSP2 and to Clostridium vaccine Ag
We recently showed a significant correlation between proliferation and IFN- ELISPOT responses by PBMC from these MSP2 vaccinates to A. marginale, MSP2, and MSP2-derived peptides (25). Nevertheless, to address the possibility that following A. marginale infection, the response changed from a predominant Th1-like response to a predominant Th2-like response that could not be detected by the IFN- ELISPOT assay, proliferation and IFN- ELISPOT assays were conducted simultaneously with the same aliquots of cells. Consistent with the results using the ELISPOT assay, proliferative responses to A. marginale, MSP2, and all MSP2-derived peptides were severely decreased in all eight vaccinates at the peak of infection and thereafter, whereas the response to the PHA, IL-12, and IL-18 was always observed (representative data for MSP2-immunized animal 71 are shown in Fig. 4, D–F).
To further determine whether the impaired T cell response to A. marginale and MSP2 was Ag specific or reflected a generalized immune suppression, Clostridium spp. Ag, which was used to vaccinate the calves before MSP2 immunization, was also included in the assays. Unlike the responses to A. marginale and MSP2, the response to Clostridium Ag was significant at all time points (Fig. 4). In addition, no significant responses to MSP2-derived peptides were seen (Fig. 4, E and F). Similar results were obtained for all immunized animals tested at various time points up to 7–12 mo postinfection (data not presented), and suggest that the impaired response to MSP2 is long-lived and does not reflect a generalized immune suppression.
Quantitation of CD25+CD4+ T cells by FACS
Regulatory T cells have not been defined in cattle. Nevertheless, to evaluate the potential role of CD25+CD4+ T regulatory cells in the loss of MSP2-specific responses following challenge, CD25+CD4+ T cells were analyzed in PBMC from immunized and control animals by two-color flow cytometry (Table IV). Although the differences in the percentage of CD25+CD4+ T cells in PBMC varied between individual animals, when data for all animals were compared, no significant differences in the percentage of CD25+CD4+ T cells, expressed as either a percentage of total CD4+ T cells or of total CD25+ T cells, were observed during the course of infection (data not shown).
Positively selected CD4+ T cells from peak rickettsemia do not respond to Ag cultured with APC from naive donors
To address the possibility that the inability to detect a CD4+ T cell response following challenge was caused by dysfunctional APC, CD4+ T cells were positively selected from PBMC of two animals (71 and 76) cryopreserved at time points when the cells responded or at peak rickettsemia when no response was detected. PBMC were stimulated with Ag in the presence of MHC class II DRB3 homozygous and half-matched APC from A. marginale naive donor cattle. CD4+ T cells obtained at time points where responses to Ag were previously observed had strong and significant proliferative responses to A. marginale and MSP2 in the presence of APC from naive donors (Table VII). In contrast, CD4+ T cells obtained at the peak of infection had undetectable proliferative responses to Ag, but did proliferate to TCGF, as observed when autologous APC were used to present Ag.
Discussion
The results of this study do not support our hypothesis that challenge with A. marginale expressing MSP2 variants to which the animals had been previously exposed would stimulate strong anamnestic CD4+ T cell and IgG responses directed against conserved and variant-specific epitopes that would be associated with variant-specific organism clearance. In contrast, analysis of the immune response upon challenge demonstrates a newly discovered modulation whereby Ag-specific T cell responsiveness induced by vaccination is lost upon rickettsial challenge.
Sequencing of msp2 transcripts from the blood of cattle obtained during ascending and peak rickettsemia ruled out the possibility that acute rickettsemia resulted from selective expansion either of organisms expressing variants of MSP2 that constituted a minor population in the challenge inoculum, or of organisms expressing novel msp2 sequences. This indicates that either the MSP2-specific immune response induced by immunization or the recall response elicited by the challenge was insufficient to effect clearance.
The lack of strong recall T cell responses to MSP2 following A. marginale infection may be related to the uniformly dramatic loss of MSP2-specific CD4+ T cell responses that occurred in all animals concurrently with development of measurable rickettsemia. However, the sustained Th cell response for the first 3 wk following challenge was apparently sufficient to stimulate a boost in IgG production. This infection-mediated immune modulation of a strong CD4+ T cell response directed against multiple antigenic epitopes (25) has not been previously described for any rickettsial pathogen. However, in other persistent pathogen infection models, anergy induced by altered peptide ligand antagonism (41, 42, 43), or by T regulatory cells producing either TGF-1 or IL-10 (37, 44, 45, 46, 47, 48, 49), has been shown to play a role in down-regulating T cell responses. Although antigenically variant MSP2 epitopes could potentially act as antagonistic peptides, previous studies did not show the ability of naturally occurring variant epitopes to cause anergy of MSP2-specific T cell lines or clones specific for the agonistic MSP2 variant (10). Furthermore, the disappearance of T cell responses to conserved MSP2 epitopes, as well as variable MSP2 epitopes, argues against antigenic variation in MSP2 as a reason for the abrupt loss of T cell responsiveness.
To address the possibility that A. marginale infection induced a T regulatory cell response, experiments were performed to determine changes in the percentage of CD25+CD4+ T cells during the course of infection, to examine IL-10 and TGF-1 production by responding and nonresponding cells, to detect the presence of a population of suppressive cells in peripheral blood by mixing responding and nonresponding cells, and to test positively selected CD4+ T cells. Although our results do not support the role of T regulatory cells in the dramatic loss of MSP2-specific CD4+ T cell immune responses, their role cannot be definitively ruled out, because these cells have not been phenotypically characterized in cattle. We were similarly unable to demonstrate a shift from a dominant IFN- Th1 response (28) response to an IL-4 dominant response following challenge.
To test the possibility that infection impaired APC to present A. marginale Ag to CD4+ T cells, positively selected CD4+ T cells were cultured with Ag in the presence of class II-compatible APC from noninfected donors. However, T cells obtained at the peak of infection were still unable to respond to Ag, ruling out dysfunctional APC as the reason for the sudden loss of response.
Sheep and dogs infected with Anaplasma phagocytophilum develop a transient immunosuppression defined by leukopenia (reduced numbers of T lymphocytes and neutrophils) and an increased susceptibility to other infectious organisms (50, 51, 52). A. phagocytophilum infects neutrophils and alters neutrophil function (53, 54, 55, 56, 57), which may explain the transient generalized immune suppression. However, a similar mechanism of generalized immune suppression by A. marginale is unlikely for the following reasons: 1) this pathogen infects erythrocytes and not neutrophils, 2) the response to unrelated clostridial Ags was not severely impaired during acute infection, and 3) increased susceptibility to unrelated or opportunistic infections has not been reported for cattle with anaplasmosis.
The unsubstantiated role of T regulatory cells in the disappearance of the MSP2-specific memory T cell response, the lack of evidence for altered Ag presentation, and the Ag-specific nature of the immune suppression suggest an alternative mechanism for the loss of T cell responsiveness. One potential mechanism is peripheral T cell deletion that could occur via activation-induced cell death (AICD) following organism challenge (58). During primary HIV infection, naturally induced HIV-specific CD8+ T cell clones with defined TCR V usage were shown to rapidly disappear, independent of changes in the viral epitopes recognized (59). An unrelated study reported in vivo elimination of Ag-specific Th1 cells, obtained from TCR transgenic mice that were adoptively transferred to normal mice, following i.v. challenge with the Ag cytochrome c 1–2 mo later (60). The Ag-specific memory T cells became rapidly activated in vivo upon Ag administration, but by day 8 following Ag challenge, declined to barely detectable numbers and remained depressed or anergic for 3 mo. The authors concluded that Ag challenge of resting Th1 CD4+ T cells led to transient activation followed by cell depletion. In our studies, A. marginale was administered i.v. and the infection took 5 wk to reach peak levels in peripheral blood. Thus, for the first 3 wk following challenge, recall T cell responses remained at prechallenge levels, but were completely undetectable at the peak of infection, 2 wk later. We therefore examined CD4+ T cell IFN- ELISPOT responses in six immunized cattle at 1 wk before the peak of infection was reached (29 or 31 DPI), and observed weakly positive responses in two animals and undetectable responses in four animals (data not shown). These results are consistent with Ag-induced AICD. Furthermore, A. marginale inhibited, in a dose-dependent manner, proliferation of lymphocytes that paralleled a loss in cell viability. A reduction in the response to TCGF from >50 to 100% was observed at organism concentrations equivalent to those observed at peak levels of rickettsemia in vivo following challenge, which ranged from 1 x 107 to 8 x 108 organisms per milliliter of blood (Fig. 2). However, the significance of these in vitro results to the in vivo infection is not clear, because A. marginale is generally intraerythrocytic and the nature of the interaction of infected erythrocytes with T cells is unknown. Nevertheless, these results are also consistent with a loss in immune responsiveness as a consequence of increasing Ag dose in vivo, and a mechanism of AICD.
In conclusion, we hypothesize that MSP2-specific memory T cells were deleted or decreased to undetectable numbers in animals following infection with A. marginale. Our data indicate a newly discovered immune modulation whereby Ag-specific T cell responsiveness is lost upon rickettsial challenge. MSP-2-specific T cells may be deleted as a consequence of high levels of Ag occurring during ascending rickettsemia, and the number of MSP2-specific T cells may remain depressed as a result of the chronic low antigenic exposure during persistent infection. Consistent with this is our inability to detect CD4+ T cell responses in nonimmunized cattle following either i.v. or tick-transmitted A. marginale challenge for up to 3 mo postinfection (this study and our unpublished observations). Additional experiments using MHC class II tetramers to track the fate of epitope-specific T cells in immunized and control cattle during acute and chronic anaplasmosis should clarify the mechanism for the loss of Ag-specific T cell responses following A. marginale infection and determine whether a similar immune modulation occurs during infection of nonvaccinated cattle.
Acknowledgments
We are grateful to Bev Hunter, Emma Karel, and Shelley Whidbee for excellent technical assistance, and Kelly Brayton and Travis McGuire for helpful discussions and assistance.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work is supported by National Institutes of Health Grants AI44005 and AI49276, and U.S. Department of Agriculture National Research Initiative Competitive Grants Program Grant 02-35204-12352.
2 Address correspondence and reprint requests to Dr. Wendy C. Brown, Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA 99164. E-mail address: wbrown{at}vetmed.wsu.edu
3 Abbreviations used in this paper: MSP2, major surface protein 2; HVR, hypervariable region; ODN, oligodeoxynucleotide; DPI, days postinfection; PCV, packed cell volume; URBC, uninfected bovine erythrocyte membrane; SFC, spot-forming cell; TCGF, T cell growth factor; AICD, activation-induced cell death.
Received for publication December 10, 2004. Accepted for publication March 17, 2005.
References
Zhi, N., N. Ohashi, Y. Rikihisa, H. W. Horowitz, G. P. Wormser, K. Hechemy. 1998. Cloning and expression of the 44-kilodalton major outer membrane protein gene of the human granulocytic ehrlichiosis agent and application of the recombinant protein to serodiagnosis. J. Clin. Microbiol. 36: 1666-1673.
Shkap, V., T. Molad, K. A. Brayton, W. C. Brown, G. H. Palmer. 2002. Expression of major surface protein 2 variants with conserved T-cell epitopes in Anaplasma centrale vaccinates. Infect. Immun. 70: 642-648.[
Palmer, G. H., J. R. Abbott, D. M. French, T. F. McElwain. 1998. Persistence of Anaplasma ovis infection and conservation of the msp-2 and msp-3 multigene families within the genus Anaplasma. Infect. Immun. 66: 6035-6039.[
Murphy, C. I., J. R. Storey, J. Recchia, L. A. Doros-Richert, C. Gingrich-Baker, K. Munroe, J. S. Bakken, R. T. Coughlin, G. A. Beltz. 1998. Major antigenic proteins of the agent of human granulocytic ehrlichiosis are encoded by members of a multigene family. Infect. Immun. 66: 3711-3718.
Ijdo, J. W., W. Sun, Y. Zhang, L. A. Magnarelli, E. Fikrig. 1998. Cloning of the gene encoding the 44-kilodalton antigen of the agent of human granulocytic ehrlichiosis and characterization of the humoral response. Infect. Immun. 66: 3264-3269.
French, D. M., T. F. McElwain, T. C. McGuire, G. H. Palmer. 1998. Expression of Anaplasma marginale major surface protein 2 variants during persistent cyclic rickettsemia. Infect. Immun. 66: 1200-1207.(Jeffrey R. Abbott*, Guy H)