The CD4⁺ T cell immunodominant Anaplasma marginale major surface protein 2 stimulates γδ T cell clones that express unique T cell receptors

Abstract

Major surface protein 2 (MSP2) of the bovine rickettsial pathogen Anaplasma marginale is an abundant, serologically immunodominant outer membrane protein. Immunodominance partially results from numerous CD4⁺ T cell epitopes in highly conserved amino and carboxy regions and the central hypervariable region of MSP2. However, in long-term cultures of lymphocytes stimulated with A. marginale, workshop cluster 1 (WC1)⁺ γδ T cells and CD4⁺ αβ T cells proliferated, leading to a predominance of γδ T cells. As γδ T cells proliferate in A. marginale-stimulated lymphocyte cultures, this study hypothesized that γδ T cells respond to the abundant, immunodominant MSP2. To test this hypothesis, γδ T cell clones were isolated from MSP2 vaccinates and assessed for antigen-specific proliferation and interferon-γ secretion. Seven WC1⁺ γδ T cell clones responded to A. marginale and MSP2, and three of these proliferated to overlapping peptides from the conserved carboxy region. The γδ T cell response was not major histocompatibility complex-restricted, although it required antigen-presenting cells and was blocked by addition of antibody specific for the T cell receptor (TCR). Sequence analysis of TCR-γ and -δ chains of peripheral blood lymphocytes identified two novel TCR-γ chain constant (C_γ) regions. It is important that all seven MSP2-specific γδ T cell clones used the same one of these novel C_γ regions. The TCR complementarity-determining region 3 was less conserved than those of MSP2-specific CD4⁺ αβ T cell clones. Together, these data indicate that WC1⁺ γδ T cells recognize A. marginale MSP2 through the TCR and contribute to the immunodominant response to this protein.

MSP2, γδ TCR, bovine

INTRODUCTION

Anaplasma marginale is a tick-borne rickettsial pathogen of cattle that invades and replicates exclusively within erythrocytes. Control of acute infection and maintenance of low-level rickettsemia characteristic of persistent infection with this organism are believed to result from an adaptive immune response directed against outer membrane proteins [1–3]. CD4⁺ T cells and interferon-γ (IFN-γ) are important for control of closely related pathogens [4–6], and in cattle immunized with A. marginale outer membranes, complete protection against challenge correlated with CD4⁺ T cell-mediated IFN-γ production [2].

The immune response to the A. marginale surface is directed predominantly to a subset of outer membrane proteins, including major surface protein 2 (MSP2), which is an immunodominant 36–44 kDa protein that has highly conserved amino and carboxy regions flanking a central hypervariable region (HVR) [3, 7–9]. Conserved regions and HVR of MSP2 contain numerous major histocompatibility complex (MHC) class II-restricted CD4⁺ T cell epitopes, consistent with the immunodominant nature of this surface protein [10, 11]. The CD4⁺ T cell epitopes recognized by a large number of immune animals expressing many different class II haplotypes are clustered in the HVR [amino acids (aa) 171–229] and conserved regions (aa 101–170 and aa 272–361), whereas linear B cell epitopes are found predominantly in the HVR, suggesting the structure of the protein influences T and B cell recognition [12].

γδ T lymphocytes have been implicated in immunity against many viral, bacterial, and protozoal diseases of mice and humans [13]. In cattle γδ, T cells may also play a role in immunity to A. marginale infection. First, young calves, which have high circulating levels of workshop cluster 1 (WC1)⁺ γδ T lymphocytes, are more resistant than adults to Anaplasma infection [14, 15]. Second, the number of peripheral γδ T cells decreases significantly late in acute A. marginale infection [16]. The reason for this is unknown, but it could be the result of sequestration of γδ T cells out of the peripheral blood into other tissues such as the spleen, which is involved in controlling the infection. Finally, WC1⁺ γδ T cells from cattle immunized with A. marginale outer membranes expand in cultures of peripheral blood mononuclear cells (PBMC) stimulated with bacterial antigen [17]. As PBMC cultured with A. marginale develop into long-term cell lines containing more than 90% WC1⁺ γδ T cells [17], and MSP2 is an abundant outer-membrane protein that contains numerous CD4⁺ T cell epitopes, we hypothesized that γδ T cells respond to MSP2. To test this hypothesis, seven γδ T cell clones were isolated from two MSP2-immunized cattle and evaluated for proliferative and cytokine responses to antigen. We describe the responses of these clones to A. marginale, MSP2, and an epitope present in aa 292–311. Furthermore, we identify the T cell receptor (TCR)-γ and -δ chain sequences used by the MSP2-specific γδ T cell clones and compare the TCR-γ and -δ chain complementarity determining region-3 (CDR3) sequences with those of TCR-α and -β chain CDR3 sequences, used by CD4⁺ T cell clones shown previously to respond to a different epitope (aa 280–291) within the same cluster [18].

MATERIALS AND METHODS

Preparation of A. marginale and MSP2 antigens

A. marginale (Florida strain) organisms were isolated from thawed, infected bovine erythrocytes by sonication and differential centrifugation as described previously [19]. To prepare antigen for in vitro assays, the sonicate was diluted in phosphate-buffered saline (PBS; pH 7.4) containing 25 μg/ml of the protease inhibitors, antipain, and E-64 (Boehringer Mannheim, Indianapolis, IN) and 300 μg/ml phenylmethylsulfonyl fluoride (Sigma Chemical Co., St. Louis, MO). Native MSP2 was purified essentially as described [20, 21]. Briefly, pelleted A. marginale were resuspended in electrophoresis sample buffer and separated on multiple preparative sodium dodecyl sulfate gradient (10–20%) polyacrylamide gels. One lane of the gel with molecular markers was cut, transferred, and blotted with MSP2-specific monoclonal antibody (mAb) AnaF19E2 to orient the MSP2 on the preparative gels. The MSP2 bands were excised, and the protein was electroeluted from the gel fragments, concentrated, dialyzed against PBS, and purified a second time on preparative gels [20]. MSP2 was verified by immunoblotting to be reactive with MSP2-specific mAb. MSP2 peptides were prepared as described previously [10, 11]. All peptides were diluted in PBS and filter-sterilized prior to assay. A. marginale sonicate, MSP2, and peptides P10 and P18 were tested at 10 μg per ml and 30 μg per ml (peptides P10 and P18) for endotoxin by the Limulus ameobocyte lysate assay according to the manufacturer's (BioWhittaker, Walkersville, MD) instructions. All samples contained <0.06 endotoxin units per ml of endotoxin, which is the limit of sensitivity of the assay.

Cattle and immunization with MSP2

Two Holstein steers (98B61 and 01B71), 4–5 months old, weighing 150–200 kg at the start of separate experiments, were verified to be serologically negative for A. marginale by competitive inhibition enzyme-linked immunosorbent assay (ELISA) [20, 22, 23]. The calves were immunized six times with MSP2 and interleukin (IL)-12 as described in detail [20, 23]. Animal C97 is a Brahman X Angus cow that was used as a donor of MHC-mismatched antigen-presenting cells (APC) in some assays. The MHC class II DRB3, DQA, and DQB alleles were determined by sequencing using the protocol described previously [24]. The DRB3 and DQ haplotypes for the animals used in this study are listed in Table 1. As a result of linkage disequilibrium, cattle with unrelated class II alleles will also have different class I alleles [25].

TABLE 1

MHC Class II Haplotypes of Animals Used in This Study

Animal	DRB3	MHC class II allelea
		DQA	DQB
98B61	*1201	*12011	*1201
		*2201	*10051
	*1201	*12011	*1201
		*2201	*10051
C97	*4501	*0301	*0402
	*3001	AY730727b	AY730728
01B71	*1201	*12011	*1201
		*2201	*10051
	*0101	*0101	*0101

a

Alleles were determined by sequencing cDNA obtained from PBMC, and nomenclature is described in the web site http://www2.ri.bbsrc.ac.uk/bola/.

b

The class II DQA and DQB sequences deposited in Genbank (accession numbers are shown) were not described previously and await nomenclature designation.

TABLE 2

Production of IFN-γ by γδ T Cells Stimulated with A. marginale or MSP2

Antigen	IFN-γ (U per ml) produced by γδ T cell clonea
	61.1G9	61.1G10	61.1G11
None	<0.2	<0.2	<0.2
URBC	<0.2	<0.2	<0.2
A. marginale	15.8 ± 1.8	1.2 ± 1.0	34.1 ± 3.6
MSP2	1.5 ± 1.4	<0.2	4.0 ± 0.6

a

γδ T cells were cultured for 3 days with 10 μg per ml medium or antigen, and supernatants pooled from two wells were tested for IFN-γ by ELISA in duplicate wells. Results are presented as the mean ± 1 sd. The limit of detection in this assay was ∼0.2 U IFN-γ per ml.

Isolation of γδ T cell clones from MSP2 vaccinates

T lymphocyte lines were established from animals 98B61 and 01B71 by using 4 × 10⁶ PBMC per well of 24-well plates (Costar, Cambridge, MA) in 1.5 ml complete RPMI-1640 medium [26] with 10 μg per ml A. marginale sonicate, 5 μg per ml MSP2, or 5 μg per ml peptide P10 from the conserved C-terminal region of MSP2 [10]. The lines were restimulated weekly for 2 weeks by subculturing to a density of 7 × 10⁵ cells per well with antigen and 2 × 10⁶ irradiated (3000 rads), autologous PBMC as a source of APC. Cells were cloned by limiting dilution by plating 10, 3, 1, or 0.3 cell per well in the presence of 5 × 10⁴ APC, antigen, and 10% bovine T cell growth factor (TCGF) [27] in 96-well round-bottomed plates (Sarstedt, Newton, NC). Cells were restimulated weekly by replacing 100 μl of the supernatant with 100 μl medium containing antigen, 2–3 × 10⁴ APC, and 20% TCGF. If the cloning frequency were less than 30%, cells were expanded to 48- and 24-well plates. After expansion, cell-surface phenotypes were determined by flow cytometry, and antigen responsiveness was analyzed with proliferation assays. WC1⁺ γδ T cell clones 61.1G9, 61.1G10, and 61.1G11 and CD4⁺ αβ T cell clones 61.1C8, 61.1E8, and 61.1F12 were obtained from animal 98B61 after stimulating PBMC for 1 week with 10 μg per ml MSP2 peptide P12 and for 1 week with 10 μg per ml peptide P10 and cloning with peptide P10 [18]. WC1⁺ γδ T cell clones 61.1A4 and 61.2D7 and CD4⁺ αβ T cell clone 61.2G4 were obtained from animal 98B61 and after stimulating PBMC for 1 week with 5 μg per ml peptide P12, 1 week with 5 μg per ml peptide P10 and cloning with peptide P10, which corresponds to aa 272–301; P12 corresponds to aa 312–341 in the sequence predicted from the Florida strain msp2 11.2 genomic DNA clone [8] (see Table 3). γδ T cell clones 71.1G7 and 71.2B1 were obtained from animal 01B71 after stimulating PBMC for 2 weeks with 5 μg per ml A. marginale (Florida strain) sonicate and cloning with A. marginale. As controls for assays measuring MSP2 peptide stimulation of MSP2-responsive γδ T cell clones, WC1⁺ γδ T cell clone G1.2A9 obtained from cow G1, which was infected with Fasciola hepatica [28], and clones C15.1A7 and C15.2F12 obtained from cow C15, which was infected with Babesia bovis [29], were also used. Clone G1.2A9 was obtained by limiting dilution cloning of a cell line cultured with 25 μg per ml F. hepatica adult worm antigen as described [28], and clones C15.1A7 and C15.2F12 were obtained by cloning a cell line cultured with 20 μg per ml recombinant B. bovis spherical body protein-1-glutathione-S-transferase fusion protein as described [29].

TABLE 3

Proliferation of γδ T Cell Clones to Peptides P10 and P18 of A. marginale MSP2

		Proliferation (SI) of the following clonesa
Antigen	Concentration (μg per ml)	61.1G9	61.1G10	61.1G11	G1.2A9b	C15.1A7b	C15.2F12b
TCGF	10%	87.7	50.3	57.2	31.1	3.3	7.5
A. marginale	10	44.0	9.3	24.3	0.6	1.0	0.7
MSP2	10	10.5	3.6	6.7	0.4	1.0	0.5
P10	10	3.1	4.0	3.0	0.8	1.1	1.7
P12	10	2.7	0.7	1.1	NDc	ND	ND
P13	10	1.5	0.9	1.3	ND	ND	ND
P18	30	6.3	18.3	50.0	0.5	0.8	0.6
	10	4.2	3.2	13.4	0.6	0.6	0.8
	3	2.2	1.4	3.7	ND	ND	ND
P10d (aa 272–301)		VAGAFARAVEGAEVIEVRAIGSTSVMLNAC
P18 (aa 292–311)		GSTSVMLNACYDLLTDGIGV
P12 (aa 312–341)		VPYACAGIGGNFVSVVDGHINPKFAYRVKA
P13 (aa 332–361)		NPKFAYRVKAGLSYALTPEISAFAGAFYHK

a

T cells were stimulated with 10% bovine TCGF or 3–30 μg per ml antigen for 3 days in the presence of APC, and results are presented as SI, determined by dividing the mean cpm of duplicate cultures of cells cultured with antigen by the mean cpm of duplicate cultures of cells with medium. Significant responses compared with medium were determined by the Student's one-tailed t-test (boldface).

b

Clone G1.2A9 was obtained from a F. hepatica-infected cow, and clones C15.1A7 and C15.2F12 were obtained from a B. bovis-infected cow.

c

Not determined.

d

The aa sequences of the peptides used are shown. Overlapping aa are in boldface. The aa position number is based on the position number relative to the sequence predicted from the msp 11.2 genomic clone [8].

Cell-surface phenotypic analysis

Differentiation markers on T cell lines and clones were analyzed by indirect immunofluorescence and flow cytometry as described previously [30]. The mAb used were specific for bovine CD2 (mAb MUC2A), CD3 (mAb MM1A), CD4 (mAb CACT 138A), CD8 (mAb CACT 80C and BAT 82B), the δ-chain of the γδ TCR (mAb CACT 61A or GB21A), γδ TCR subset markers N6 (mAb CACTB6A) and N7 (mAb CACTB81A), the WC1 marker (mAb BAQ 4A or IL-A29), and subsets of WC1 N3 (mAb CACTB15A) and N4 (mAb BAQ89A). mAb IL-A29 was obtained from the International Laboratory for Research on Animal Diseases (Nairobi, Kenya), and all other mAb were purchased from the Monoclonal Antibody Center [Washington State University (WSU), Pullman]. mAb GD3.1, GD3.5, and GD3.8, which recognize other γδ T cell-surface markers and delineate cell subsets [31], were kindly provided by Dr. Mark Jutila (Montana State University, Bozeman).

Lymphocyte proliferation assays

Proliferation assays were carried out in replicate wells of round-bottomed 96-well plates for 3–4 days, essentially as described [3]. T cell clones were assayed 7 days after the last stimulation with antigen and APC. T cells (3×10⁴ per well) were cultured in duplicate wells in a total volume of 100 μl complete medium containing 2 × 10⁵ APC and antigen. Antigens consisted of 1–30 μg per ml A. marginale sonicate, native MSP2, synthetic MSP2 peptides, and as a control, membranes prepared from uninfected bovine red blood cells (URBC) or B. bovis [26]. Protein concentrations in all antigen preparations were determined by the Bradford assay. To measure proliferation, cells were radiolabeled for the last 18 h of culture with 0.25 μCi [³H]thymidine (Dupont New England Nuclear, Boston, MA), radiolabeled nucleic acids were harvested onto glass filters, and radionucleotide incorporation was measured with a Betaplate 1205 liquid scintillation counter (Wallac, Gaithersburg, MD). In some experiments, 50 μg per ml protein G-affinity-purified immunoglobulin G (IgG)2b mAb GB21A (antibovine δ TCR chain) [32] was added to proliferation assays. An irrelevant, isotype-matched control mAb PIq45A2, specific for bovine IgM (Monoclonal Antibody Center, WSU), was used to as a negative control. Cells were cultured for 3–4 days in duplicate or triplicate wells. All experiments were performed at least twice. Results are presented as the mean number of counts per minute (cpm) of replicate cultures ± 1 sd or the stimulation index (SI), which represents the mean cpm of replicate cultures of cells plus antigen/the mean cpm of replicate cultures of cells plus medium or URBC.

Detection of IFN-γ in supernatants of T cell clones

γδ T cell clones (3×10⁵ cells per ml) were cultured for 3 days with APC (2×10⁶ cells per ml) and 10 μg per ml URBC, MSP2, or A. marginale antigen, and supernatants were tested for IFN-γ using a commercial ELISA kit (Bovigam, CSL Ltd., Parkville, Victoria, Australia) according to the manufacturer's protocol. The IFN-γ activity in culture supernatants, diluted 1:4–1:20, was determined by comparison with a standard curve obtained with a supernatant from a Mycobacterium bovis-purified protein derivative-specific T cell clone, which contained 440 U IFN-γ per ml (previously determined by the neutralization of vesicular stomatitis virus [30]). In our assay, 0.6 U corresponds to 1 ng IFN-γ [33]. The results are presented as units of IFN-γ per ml supernatant.

Statistical analysis

Significance in proliferation and IFN-γ production in response to antigen compared with medium or URBC was determined by the Student's one-tailed t-test. Comparisons of proliferation in the presence of TCR-blocking mAb versus control mAb or autologous versus MHC-mismatched APC were analyzed using the two-tailed Student's t-test. P < 0.05 is considered significant.

Sequencing the γδ TCR of MSP2-responsive T cell clones

γδ T cell clones were washed after 7 days of culture with antigen and APC and were then cultured for 20 h with 10% TCGF without antigen or APC. RNA was collected using the TRIzol reagent (Gibco-BRL, Gaithersburg, MD) as described by the manufacturer. RNA was reverse-transcribed, and cDNA was amplified using the switching mechanism at the 5′ end of RNA transcript rapid amplification of cDNA ends (SMART RACE) cDNA amplification kit (BD Biosciences, Clontech, Palo Alto, CA). TCR cDNA was specifically amplified using the SMART RACE protocol and the Advantage 2 polymerase chain reaction (PCR) enzyme system (BD Biosciences, Clontech). The sequence was obtained using SMART RACE and primers designed from previous reports [34, 35]. The primers used for the RACE reverse primers were the δ-chain constant (C_δ) region 5′-GCAGGTCACTGGAGCTTCAGCTT-3′ (GenBank accession no. D90419) and known bovine C_γ regions 5′-ATGGTCAGCCAGCTGAACTTCATGTAGG-3′ (GenBank accession no. X63684), 5′-ACGGTCAGCCAGCTGAACTTCATGTATG-3′ (GenBank accession no. AY644517), 5′-ACGGTCAGCCAGCTTAACTTCATGTATG-3′ (GenBank accession no. D90412), and 5′-ACGGTCAGCCAGCTGAGCTTCATGTATG-3′ (GenBank accession no. X63680). All sequences were confirmed by PCR using primers based on sequence obtained in the SMART RACE protocol. The forward primer used for confirmation of PCR for the δ-chain was based on RACE sequence and was 5′-CATTTGTGCAGGAAAATCCATGCCTC-3′. Two new reverse primers made for two previously undescribed C_γ regions, based on preliminary data from this study and M. S. Abrahamsen's laboratory, were 5′-TAATTGAAGGAAGAAAAATTGTGGGTTTTG-3′ and 5′-TGCAGGCATGTGTAGCGACCTCTTT-3′.

Using sequence data obtained by performing SMART RACE with PBMC, forward primers were identified that amplified sequence from the two new C_γ regions. The forward primer used to amplify the γ-chain, including the C_γ region similar to the fourth sheep constant region, is 5′-CATGATCTTTGGTGAAGGAACAAAAG-3′. The forward primer used to amplify the γ-chain, including the C_γ region similar to the fifth sheep constant region, is 5′-GAGACTTCCTTCCAACAGACCTTGCCT-3′. The GenBank accession numbers for the fourth and fifth sheep C_γ regions are Z12967 and Z13986, respectively.

5′ RACE was also used for sequencing the α- and β-chains of MSP2 peptide P10-specific αβ T cell clones 61.1C8, 61.1E8, 61.1F12, and 61.2G4. The primers used were as follows for the α-chain, 5′-CCGCAGCGTCATGAGCAGAT-3′ (GenBank accession no. AY227782.2) and 5′-CCATGTTGAGCACGGTGCTG-3′ (GenBank accession no. D10394.1), and for the β-chain, 5′-ACAGCGTACAGGGTGGCCTT-3′ (GenBank accession no. D90140.1) and 5′-CCGTGGAACTGGACTTGGCA-3′ (GenBank accession no. D90139.1).

RESULTS

A. marginale antigen-induced response of WC1⁺ γδ T cells

Previous studies demonstrated that A. marginale-stimulated WC1⁺ γδ T cells expanded in cultures of PBMC from A. marginale outer membrane-immunized cattle [17]. In the present study, γδ T cells were similarly found to expand in cultures of PBMC from MSP2-immunized cattle. Seven antigen-responsive γδ T cell clones were isolated from two animals. All clones were γδ TCR⁺, WC1⁺, CD2⁻, CD3⁺, CD4⁻, CD8⁻, GD3.1⁺, GD3.5⁺, and GD3.8⁺ when examined by flow cytometry. In addition, the clones were positive for TCR markers N6 and N7 as well as N3, a surface marker corresponding to WC1.2 [36]. The expression of the same surface phenotype by all γδ T cell clones suggests that these clones represent a subset of γδ T cells.

All clones were tested in proliferation assays with MSP2 and A. marginale, and all responded significantly to both antigens (Fig. 1). IFN-γ levels in the supernatants of three T cell clones stimulated with antigen were analyzed by ELISA. γδ T cell clones 61.1G9 and 61.1G11 produced significant IFN-γ in response to A. marginale and less IFN-γ in response to MSP2, whereas clone 61.1G10 produced a small amount of IFN-γ in response to A. marginale only (Table 2). IFN-γ levels correlated with proliferation (SI) by these clones (r²=0.71; P<0.05).

Fig. 1

Proliferative response of γδ T cell clones to A. marginale and MSP2. γδ T cell clones were cultured for three days with APC and 10 μg per ml A. marginale, MSP2, or negative control URBC antigen. Results are reported as SI, determined as the mean cpm of cells cultured with antigen/mean cpm of cells cultured with medium. Responses to A. marginale and MSP2 are statistically significant (P≤0.05) when compared with the negative control antigen URBC, as indicated by asterisks.

Clones that proliferated strongly to MSP2 were also tested for their ability to respond to peptides (Table 3) spanning the conserved C region of MSP2, which stimulated significant proliferation of αβ CD4⁺ T cells derived from the same animal [10, 11, 18] and constituted an epitope cluster when a large number of cattle were tested [12]. It is interesting that the response by three γδ T cell clones from animal 61 derived by cloning in the presence of peptide P10 responded weakly to peptide P10 but not to peptide P12 or P13 (Table 3). Additional experiments were then performed with peptide P18, which overlaps peptide P10 by 10 aa, and peptide P18 elicited significant dose-dependent proliferation of these clones in three independent experiments (representative data are presented in Table 3). We found that for the majority of clones tested, when equal amounts of antigen were used in the same experiment, the response to A. marginale was greater than the response to MSP2, which was in turn greater than the response to peptide (Fig. 1 and Tables 2 and 3). Several of the other γδ T cell clones described in this study were tested but did not respond to any peptide (data not shown). In addition, three WC1⁺ γδ T cell clones derived from cattle infected with the helminth parasite F. hepatica (clone G1.2A9) or the protozoan parasite B. bovis (clones C15.1A7 and C15.2F12) did not respond to A. marginale, MSP2, or peptides P10 and P18 (Table 3). Incidentally, these γδ clones did not proliferate to the specific antigens used to generate the cell lines and clones (data not shown).

To determine whether the antigen response by γδ T cell clones was MHC-restricted, proliferation of clones 61.1G9 and 61.1G10 was compared using autologous APC and APC from animal C97, which are mismatched at MHC alleles (Table 1). The γδ T cells responded similarly to A. marginale in the presence of autologous APC and APC expressing different MHC molecules, whereas the response to B. bovis was not significant when either source of APC was used (Fig. 2a and 2b). However, APC were required for the response to A. marginale, as both clones failed to proliferate in response to this antigen when APC were absent (Fig. 2c).

Fig. 2

The proliferative response of γδ T cell clones to A. marginale is not MHC-restricted. Clones 61.1G9 (a) and 61.1G10 (b) were stimulated in duplicate with 10 μg per ml A. marginale antigen or control antigen B. bovis membranes with autologous and mismatched APC from animal C97. In a separate experiment, the same clones were stimulated with 10 μg per ml A. marginale without or with autologous APC (c). The response to A. marginale is significant (P<0.05) with autologous or MHC-mismatched APC, as indicated by asterisks.

To determine if the γδ TCR was required for proliferation to antigen, mAb GB21A specific for the δ-chain of the TCR was added during proliferation assays. mAb GB21A effectively blocked proliferation to A. marginale, whereas the isotype control mAb had no statistically significant effect (Fig. 3).

Fig. 3

Antibody specific for the γδ TCR blocks A. marginale-induced proliferation of γδ T cell clones, where clones 61.1G11 and 61.1A4 were cultured in duplicate with autologous APC and 10 μg per ml A. marginale antigen in the presence of 50 μg per ml mAb GB21A specific for the TCR-δ chain or isotype control mAb PIq45A2. Responses were significantly decreased (P≤0.05) with mAb GB21A when compared with A. marginale alone or with the isotype control mAb, as indicated by asterisks. FL, A. marginale, Florida strain.

Sequence analysis of TCR-γ and -δ chains

Analysis of the γδ TCR sequences was performed to identify any areas of sequence homology that might clarify the mechanism of antigen recognition. To determine primers for sequencing the TCR-γ and -δ chains, RACE was performed on RNA isolated from PBMC from three animals, including animals 98B61 and 01B71, which were the source of the γδ T cell clones. We identified two novel C_γ regions not described previously in cattle that are similar to the fourth and fifth C_γ regions in sheep [37], which we have designated C_γ6 (GenBank accession no. AY35450) and C_γ5 (GenBank accession no. AY35449), respectively (Table 4). The bovine C_γ5 region has 91% identity with the nucleotide sequence of the ovine C_γ5 region (Fig. 4a), and the bovine C_γ6 region has 86% identity with the ovine C_γ4 region (Fig. 4b). The bovine variable (V)_γ and joining (J)_γ regions also have considerable homology with previously described ovine sequences [37].

Fig. 4

Comparison of two novel bovine C_γ and ovine Cγ5 and Cγ4 nucleic acid sequences. Nucleotide positions are listed above the sequence. (a) Comparison of bovine and ovine Cγ5 sequences. (b) Comparison of bovine Cγ6 and ovine Cγ4 sequences.

TABLE 4

Nomenclature and Relationship of Ovine and Bovine C_δ Regions

Ovine Cγ	Bovine Cγ
Cγ1	Cγ4
Cγ2	Cγ1,Cγ2
Cγ3	Cγ3
Cγ4	Cγ6a
Cγ5	Cγ5b

a

Bovine C_γ region identified in this study, which corresponds to the ovine C_γ4.

b

Bovine C_γ region identified in this study, which corresponds to the ovine C_γ5 used by all γδ T cell clones responding to A. marginale MSP2.

After developing the SMART RACE and PCR protocols for sequencing the γδ TCRs from PBMC, TCR sequences were determined for each of the γδ T cell clones. Clones 61.1G9 and 61.1G11 have identical V_δ regions, whereas all others are different. There seems to be no association between animals and use of a V_δ region. There are two J_δ regions used, but again, there is no association between the animal and the J_δ region used. Sequences were confirmed by PCR based on the RACE sequence (Fig. 5a). Overall, there is 91.4% aa similarity and 62.4% identity between the δ-chains of the seven clones. The δ-chain CDR3 varies greatly in length from 13 to 22 aa and has only two conserved residues, as based on the ImMunoGeneTics database (http://imgt.cines.fr) [38]. The lack of conservation of the δ-chain CDR3 is similar to what was reported for human δ-chains [39].

Fig. 5

Comparison of TCR-δ and TCR-γ chain sequences of A. marginale MSP2-responsive γδ T cell clones. The aa sequence alignment of TCR-δ chains (a) and TCR-γ chains (b) from the indicated clones is shown. Position for aa is shown above the sequence. Identical aa are shaded black. Conserved substitution aa are gray. Lines indicate leader (L), variable (V), joining (J), constant (C), and CDR3. Asterisks indicate conserved cysteine residues. The diversity (D)_δ region is not indicated because of the ambiguous boundary. Boundaries of L, V, J, and C regions and CDR3 are predicted based on data from Lefranc [38].

The γ-chain sequences for all T cell clones were also determined. All clones used the same previously undescribed bovine C_γ region that corresponds to the fifth sheep C_γ region. In addition, all clones have an identical J_γ region. The translated γ-chain sequences obtained from all seven clones are presented in Figure 5b and have 97.7 and 72.9% similarity and identity, respectively. The γ-chain sequences from clones 61.1G9 and 61.1G11 are identical. Despite the conservation of the C_γ and J_γ regions in γδ T cell clones from the two animals, the V_γ region, which is identical in T cell clones derived from one individual, differed between individuals. Also noticeable is the conservation of the J_γ region use among the clones of both individuals. The conservation of use of the V, J, and C regions of the γ-chain is striking compared with that of the δ-chain, suggesting that the conserved V_γ, J_γ, and C_γ regions of the TCR may be important for antigen recognition. Despite the conservation of the V_γ, J_γ, and C_γ regions, the CDR3 is not highly conserved. There are 3 conserved aa in the CDR3, but the length varies from 5 to 11 aa.

Sequence analysis of TCR-α and -β chains

The specific requirements for interaction between the TCR and antigen could not be clearly determined, as many blocks of sequences in the γ- and δ-chains were conserved between γδ T cell clones. As the requirements for antigen specificity, including conservation in CDR3 regions, are more clearly defined in αβ T cells, it was of interest to compare the γδ TCR sequences with the αβ TCR sequence of four CD4⁺ T cell clones, which were also isolated from animal 98B61 and also responded to peptide P10. All CD4⁺ T cell clones responded to aa 272–291 (VAGAFARAVEGAEVIEVRAI), and the epitope within peptide P10 was further mapped for clones 61.1E8, 61.1C8, and 61.2G4 to aa VEGAEVIEVRAI (ref. [18] and unpublished observations). When compared, the TCR-α chains have 100% similarity and 93.3% identity, and the CDR3 region only varies by 1 aa in length and has 3 aa, which are conserved within the region (Fig. 6a). The TCR-β chains have 99% similarity and 82.9% identity, and the CDR3 are uniform in length and contain 5 identical aa (Fig. 6b). It is interesting that clones 61.2G4 and 61.1C8 have the identical TCR sequence, although they were isolated in separate cloning experiments. Thus, the TCR-γ and -δ chains of the γδ T cell clones are more diverse than the TCR-α and -β chains of the CD4⁺ T cell clones.

Fig. 6

Comparison of TCR-α and TCR-β chain sequences of A. marginale MSP2 peptide P10-specific αβ T cell clones. The aa sequence alignment of TCR-α chains (a) and TCR-β chains (b) from the indicated clones is shown. Position for aa is shown above the sequence. Identical aa are shaded black. Conserved substitution aa are gray. Lines indicate the L, V, J, C, and CDR3 regions. Asterisks indicate conserved cysteine residues. The D_β region is not indicated because of the ambiguous boundary. Boundaries of L, V, J, and C regions and CDR3 are predicted based on data from Lefranc [38].

DISCUSSION

This study tested the hypothesis that γδ T cells respond to the A. marginale immunodominant outer membrane protein MSP2 known to stimulate CD4⁺ T cells [10, 11]. We demonstrate that WC1⁺ γδ T cell clones do proliferate and secrete IFN-γ following stimulation with APC and A. marginale or MSP2 but not with APC and medium alone. Furthermore, control antigens, which included membranes prepared from uninfected bovine erythrocytes or B. bovis merozoites isolated from parasite-infected bovine erythrocyte cultures, did not stimulate the MSP2-responsive γδ T cell clones. Together, these controls suggest that irradiated monocytes within the APC [40] are not stimulating the γδ T cell clones to proliferate in the absence of A. marginale or MSP2.

Using peptides from the conserved C region of MSP2, which were known to stimulate CD4⁺ T cells, the 30-mer peptide P10 (VAGAFARAVEGAEVIEVRAIGSTSVMLNAC), present in the epitope-rich region of MSP2 consisting of aa 272–361 [12], also stimulated significant proliferation of three MSP2-responsive γδ T cell clones. Upon further analysis, the clones were shown to respond more strongly to peptide P18 (GSTSVMLNACYDLLTDGIGV), which overlaps peptide P10 by 10 N-terminal aa. Peptide P18 was not recognized by peptide P10-specific CD4⁺ T cell clones obtained from this same animal, and in fact, the epitope (VEGAEVIEVRAI) recognized by these αβ CD4⁺ T cell clones does not overlap peptide P18. Thus, peptide P10 apparently contains at least two nonoverlapping T cell epitopes: aa VEGAEVIEVRAI, recognized by CD4⁺ αβ T cells [18], and contiguous aa GSTSVMLNAC, also present in peptide P18, which stimulates γδ T cells from animal 61. The novel finding that γδ T cells as well as CD4⁺ T cells respond to MSP2 may additionally explain the immunodominance of this surface protein.

All seven γδ T cell clones identified in this study used one of two newly identified Cγ regions. The novel C_γ regions are similar to those described in sheep and correspond to the ovine fourth and fifth C_γ regions. As predicted by Hein and Dudler [37], this sequence information indicates that the bovine genome contains all of the C_γ regions found in sheep and an apparent duplication of the second ovine C_γ region. We have designated the new bovine C_γ sequences as C_γ5, most closely related to ovine C_γ5 sequence, and C_γ6, which is most closely related to the ovine C_γ4 sequence (Table 3). In other species, different C_γ regions have been associated with different tissue distributions and function [13].

The TCR CDR3 is important for antigen recognition by αβ T cells [39, 41]. However, with the exception of clones 61.1G9 and 61.1G11, sequences of the CDR3 of the TCR-γ and -δ chains of the MSP2-specific γδ T cell clones in this study do not show a high degree of similarity in number of conserved aa or length of the CDR3. Overall, the γ- and δ-chain CDR3 are less conserved than the corresponding CDR3 in the MSP2 peptide P10-specific αβ T cell clones. Conversely, there is significant homology of other portions of the γδ TCR, including the consistent use of previously undescribed J_γ and C_γ regions by all clones and conservation as well in V_γ, V_δ, and J_δ sequences. These findings indicate that unlike αβ T cells, the interaction of γδ T cells with antigen is not likely through the CDR3. This is supported by the finding that the CDR3 of clones 61.1G9/61.1G11 and 61.1G10, which respond to MSP2 peptide P18, differ by 10 (δ-chain) and 7 (γ-chain) aa. However, blocking the TCR by addition of mAb specific for the δ-chain led to a significant decrease in the proliferative response to antigen, suggesting that the TCR is involved in antigen recognition.

The response to antigen by the γδ T cell clones in this study required irradiated PBMC as a source of APC, although APC, expressing class I and class II alleles distinct from those expressed by the clones, were competent. The requirement for APC and ability of MHC-mismatched APC to stimulate γδ Tcell clones in the presence of A. marginale antigen are consistent with surface presentation of antigen by molecules other than conventional MHC. Alternatively, noncognate interactions of APC with γδ T cells through production of soluble growth factors in response to antigen may contribute to their activation [42, 43]. However, this possibility is less likely, as other γδ T cell clones isolated from cattle infected with unrelated pathogens did not proliferate in response to A. marginale or MSP2 in the presence of APC.

Murine and human γδ T cells can directly recognize nonconventional MHC class I-related molecules such as MHC class I-related chain A/B and T10/T22 [44, 45]. Furthermore, in cattle, γδ cells proliferate strongly to self-antigens expressed on irradiated monocytes [40]. Thus, if nonconventional MHC class I-like molecules were up-regulated on APC, stressed or activated by A. marginale antigen, γδ T cells with limited TCR diversity could respond to multiple stressors through recognition of the nonclassical MHC molecule. In this model, A. marginale sonicate could potentially induce more stress or stimulation to APC than a peptide, which would in turn lead to a more robust up-regulation of the stress molecules and a more intense stimulation of the γδ T cells. Our data are consistent with such a model, as there was a general hierarchy of responsiveness of A. marginale > MSP2 > peptide.

In summary, bovine WC1⁺ γδ T cells respond to the immunodominant A. marginale outer membrane protein MSP2 by proliferating and producing IFN-γ. The MSP2 epitope recognized by three γδ T cell clones is located within a CD4⁺ T cell epitope cluster in the conserved C region of the protein. All MSP2-responsive γδ T cell clones express a novel TCR C_γ region sequence. Further study is needed to elucidate the mechanisms of antigen presentation and γδ T cell activation by A. marginale, MSP2, and specific peptide. Possible mechanisms include presentation of A. marginale antigen by nonclassical MHC molecules, antigen-induced up-regulation of a molecule on APC, which is directly recognized by the γδ T cell clones, response to cytokines secreted by the APC in response to antigen, or direct recognition of microbial products by γδ T cells via Toll-like receptor 2 [46], with required accessory molecules provided by APC. Studies are planned to determine which of these mechanisms is important for the γδ T cell response to MSP2.

ACKNOWLEDGMENTS

This research was supported by USDA National Research Initiative Competitive Grants Program Grant 00-52100-9612 and NIH National Institute of Allergy and Infectious Diseases Grants R01-AI44005 and K08-AI53594. We thank Kim Kegerreis, Daming Zhu, and Bev Hunter for their technical support.