CD4⁺ T Lymphocytes from Anaplasma marginale Major Surface Protein 2 (MSP2) Vaccinees Recognize Naturally Processed Epitopes Conserved in MSP3

Abstract

Major surface protein 2 (MSP2) and MSP3 of the persistent bovine ehrlichial pathogen Anaplasma marginale are immunodominant proteins that undergo antigenic variation. The recently completed sequence of MSP3 revealed blocks of amino acids in the N and C termini that are conserved with MSP2. This study tested the hypothesis that CD4⁺ T cells specific for MSP2 recognize naturally processed epitopes conserved in MSP3. At least one epitope in the N terminus and two in the C terminus of MSP2 were also processed from MSP3 and presented to CD4⁺ T lymphocytes from MSP2-immunized cattle. This T-lymphocyte response to conserved and partially conserved epitopes may contribute to the immunodominance of MSP2 and MSP3.

Anaplasma marginale is a tick-transmitted ehrlichial pathogen of cattle that causes acute anemia and subsequent subclinical persistent infection of erythrocytes. Protective immunity can be achieved by immunizing cattle with a purified outer membrane fraction of A. marginale (8, 19). Several major surface proteins (MSPs), designated MSP1 to MSP5, identified in this fraction have been extensively studied (18). In outer membrane vaccinees completely protected from developing infection after challenge, immunity was characterized by a CD4⁺ T-lymphocyte response against MSP1, MSP2, and MSP3 and an immunoglobulin G2 (IgG2) response directed predominantly against MSP2 (8, 9). Interestingly, the majority of CD4⁺ T-lymphocyte clones from animals vaccinated with outer membranes recognized MSP2 or MSP3, and several clones responded to both proteins (9). These results suggest recognition of epitopes shared by these MSPs.

MSP2 is an ∼36- to 44-kDa immunodominant surface protein that undergoes dynamic antigenic variation in vivo by recombination of pseudogenes or short segments of pseudogenes into a single operon-linked expression site (2-4, 12, 17). The MSP2 variants have conserved N and C regions flanking a central hypervariable region. The importance of MSP2 antigenic variation in vivo as an immune escape mechanism was indicated by the finding that newly emerging variants in persistently infected cattle stimulated a primary variant-specific antibody response (11). Furthermore, MSP2 contains CD4⁺ T-cell epitopes both in the conserved regions flanking the central hypervariable region and in the hypervariable region itself (6, 7). CD4⁺ T-helper (Th)-cell epitopes conserved among MSP2 variants may allow for rapid anamnestic Th-cell responses and efficient IgG production against sequentially emerging variant organisms, controlling the bacteremia to subclinical levels (7). However, Th-cell epitopes in the hypervariable region shown to undergo segmental gene conversion were sufficiently altered to prevent T-cell recognition, suggesting that the variation of Th-cell epitopes is another mechanism of immune evasion by A. marginale (6).

The full-length sequence of the A. marginale strain Florida's ∼2.8-kb msp3 gene was recently determined and shown to undergo a mechanism of genetic recombination similar to that of msp2, resulting in antigenically variant proteins ranging from 70 to 86 kDa with conserved N and C regions flanking a central hypervariable region (5, 14). Comparison of the MSP2 and MSP3 predicted amino acid sequences identified an average total sequence identity of ∼35% between the two proteins, but the exact percent identity is dependent on the variants selected for comparison (5). However, as shown in Fig. 1, blocks of sequence are highly conserved between the two proteins in the N-terminal regions and there is complete amino acid identity of the last 138 amino acids of the C regions (14).

FIG. 1.

Amino acid sequence alignment of MSP3 and MSP2. The predicted sequences and numbering of MSP2 and MSP3 amino acids are derived from the msp2 11.2 genomic DNA clone of strain Florida (17) and the msp3-16D3 allele (14). Areas of amino acid identity have a black background. Hyphens (-) indicate deletions in MSP2 or MSP3 relative to the other protein. An additional sequence in the hypervariable region of MSP3 (MSP3 HVR) is indicated by arrows.

The present study was designed to test the hypothesis that CD4⁺ T lymphocytes specific for MSP2 epitopes that are conserved in MSP3 can respond to these naturally processed epitopes in MSP3. The presence of Th-cell epitopes shared by these proteins could partly explain their immunodominant nature. In the N terminus of MSP2, amino acids (aa) 46 to 69 are 70.8% identical to those of positions 51 to 74 in MSP3, and aa 116 to 144 in MSP2 are 82.7% identical to those in positions 128 to 156 of MSP3 (14). Previous studies had determined that aa 101 to 142 of MSP2 contained at least two Th-cell epitopes recognized by MSP2 vaccinees (animals 59 and 61 [7]), enabling us to test whether CD4⁺ T cells from these cattle would respond to the corresponding sequence in MSP3. To this end, peptides P6 and P7, previously shown to stimulate CD4⁺ T-cell lines from MSP2-immunized animals 61 and 59, respectively (7), and peptides representing the homologous sequence in MSP3 (Table 1) were tested with short-term T-cell lines from these animals. Cell lines from animal 59 had significant proliferative responses to peptides MSP2 P7 and MSP3 P7, although the responses to the latter peptide were consistently weaker (Fig. 2A). These peptides differ by 10 of 30 amino acids. In contrast, the cell lines from animal 61, and to a lesser extent those from animal 59, responded to MSP2 peptide P6, but not to MSP3 peptide P6 (Fig. 2A and B). These peptides differ by 17 amino acids, including 15 at the N terminus (Table 1). These results show that at least one epitope in the N-terminal region of MSP3 is sufficiently conserved with MSP2 to stimulate T cells from cattle immunized with MSP2.

TABLE 1.

A. marginale strain Florida MSP2 peptides in the conserved N terminus with immunostimulatory activity for MSP2-specific CD4⁺ T lymphocytes and corresponding peptides in MSP3

Peptide	Position (aa)a	Sequence
MSP2 P1	(1-30)	MSAVSNRKLPLGGVLMALVAAVAPIHSLLA
MSP2 P6	(101-130)	SAPEPKISFKDSMLTALEGSIGYSIGGARV
MSP3 P6	(113-142)	WSTPLPGLEFGNSTLALGGSIGYRIGGARV
MSP2 P7	(121-150)	IGYSIGGARVEVEVGYERFVIKGGKKSNED
MSP3 P7	(133-162)	IGYRIGGARVEVGIGHERFVIKGGDDTAFL
MSP2 P8	(141-170)	IKGGKKSNEDTASVFLLGKELAYDTARGQV
MSP3 P8	(154-183)	IKGGDDTAFLLGRELALDTARGQLLSSALG
MSP2 P10	(272-301)	VAGAFARAVEGAEVIEVRAIGSTSVMLNAC
MSP3 P10	(806-835)	VAGAFARAVEGAEVIEVRAIGSTSVMLNAC
MSP2 P16	(382-409)	TAGKNKDTGIASFNFAYFGGELGVRFAF
MSP3 P16	(922-949)	TAGKNKDTGIASFNFAYFGGELGVRFAF

^aThe positions and amino acid sequences of previously defined immunostimulatory MSP2 peptides (7) (GenBank accession no. U07862) and the corresponding homologous peptides or sequences in the MSP3 16D3 expression locus clone (14) (GenBank accession no. AY129828) are shown. The positions of the amino acids are based on the msp2 11.2 genomic DNA clone (17) and the msp3-16D3 allele (14). Amino acid differences in MSP3 are underlined.

FIG. 2.

MSP2 and MSP3 share a T-cell epitope (peptide P7) present in the partially conserved region of the N termini. Short-term T-cell lines derived from animals 59 (A) and 61 (B) immunized with MSP2 were tested with MSP2 peptides P6 and P7 and the corresponding peptides from MSP3 shown in Table 1. Positive- and negative-control antigens included A. marginale strain Florida homogenate (FL) and membranes from uninfected red blood cells (URBC), respectively. Results are presented as the mean counts per minute + 1 standard deviation of triplicate cultures of T cells with antigen-presenting cells and 1 or 10 μg of antigen or peptide per ml. Mean background counts per minute of cells cultured in medium was 7,890 ± 5,368 (cell line from animal 59) and 4,329 ± 617 (cell line from animal 61). Asterisks indicate that the response was significantly different from the control response to either medium or URBC (P < 0.05). Results are representative of three experiments.

We next examined whether T lymphocytes specific for epitopes in the completely conserved C-terminal region of MSP2 could also respond to the same epitope in naturally processed MSP3. The C region of MSP2 contains several immunodominant Th-cell epitopes, including at least one present in peptide P10, spanning aa 272 to 301, and one present in peptide P16, spanning the C-terminal aa 382 to 409. These peptides are recognized by MSP2 vaccinees that express different major histocompatibility complex class II haplotypes (7; our unpublished observations). Therefore, CD4⁺ T-cell clones specific for peptides P10 (7) and P16 could be used to assess recognition of the epitope in MSP3. However, because peptide MSP2 P10 and the homologous sequence in MSP3 (Fig. 1; Table 1) are immediately C terminal to the central hypervariable regions of these proteins and amino acids flanking a T-cell epitope can affect recognition (10, 15, 21), it was important to first map the epitope(s) on peptide P10. To do this, short-term T-cell lines and CD4⁺ T-cell clones from MSP2 vaccinees 59, 60, and 61 were tested with peptide P18, which overlaps peptide P10 by 10 amino acids on the C terminus (Table 2). Although all cells responded to peptide P10, none of the cell lines or Th-cell clones responded to peptide P18, suggesting that the epitope was contained within the first 20 amino acids of peptide P10. To define this epitope, a panel of truncated peptides (P10-1 to P10-8 and P25; Table 2) were evaluated for stimulating the peptide P10-specific T cells. Animal 59 did not respond to any of the truncated peptides, indicating that the epitope must contain additional amino acids not represented in these peptides. However, cell lines from animals 60 and 61 did respond to the truncated peptides and the minimal epitope for animal 60 was defined as the 12-mer VEGAEVIEVRAI, whereas the minimal epitope for animal 61 was defined as the 11-mer EGAEVIEVRAI. Analysis of peptide P10-specific CD4⁺ T-cell clones 61.1C8 and 61.1E8 from this animal showed slightly different fine specificities in repeated experiments, with clone 61.1E8 recognizing the 11-mer and clone 61.1C8 recognizing the 12-mer as a minimal epitope (Table 2). However, the response by the latter clone was approximately sevenfold stronger to the 13-mer AVEGAEVIEVRAI, indicating that the additional alanine residue at the N terminus improved the response by this T-cell clone.

TABLE 2.

Mapping the CD4⁺ T-cell epitope in A. marginale MSP2 peptide 10

Peptide	Sequence	Proliferation (stimulation index) of T cells from:a
		Animal no.:			Clone
		59	60	61	61.1C8	61.1E8
P18	GSTSVMLNACYDLLTDGIGV	1.1	2.4	0.9	0.7	0.8
P10	VAGAFARAVEGAEVIEVRAIGSTSVMLNAC	6.0	10.7	10.9	199.1	119.7
P10-1	GAEVIEVRAI	0.7	2.2	2.0	0.9	0.7
P10-2	EGAEVIEVRAI	0.8	2.7	9.2	1.0	185.9
P10-3	VEGAEVIEVRAI	1.3	6.9	10.9	30.7	126.7
P10-4	AVEGAEVIEVRAI	0.8	14.0	11.0	224.8	152.3
P10-5	RAVEGAEVIEVRAI	0.9	11.5	10.2	248.7	138.1
P10-6	ARAVEGAEVIEVRAI	1.1	28.3	11.0	277.2	152.3
P10-7	FARAVEGAEVIEVRAI	1.0	15.7	11.1	264.8	126.9
P10-8	AFARAVEGAEVIEVRAI	1.4	8.3	11.2	269.5	159.4
P25	VAGAFARAVEGAEVIEVRAI	1.0	17.7	11.4	257.0	117.7

^aShort-term T-cell lines or CD4⁺ T-cell clones specific for peptide 10 were tested against 10 μg of peptide per ml, and the results are presented as the stimulation index from a representative experiment that was repeated at least twice. Proliferation was considered significant if the stimulation index was >3.0 (values are in boldface).

To determine whether peptide P10-specific CD4⁺ T cells could respond to the shared epitope in MSP3, CD4⁺ T-cell clone 61.1C8 was used in proliferation assays with native MSP2, native MSP3, peptide P10, and an unrelated peptide, P8 (Fig. 3A). Peptide P8 represents aa 141 to 170 in the N-terminal region of MSP2 that is not conserved in MSP3 (Fig. 1; Table 1). MSP2 was eluted from sodium dodecyl sulfate-polyacrylamide gels as described previously (7, 20), and MSP3 was immunoaffinity purified with MSP3-specific monoclonal antibody AmG75C2 as described previously (13). A control CD4⁺ T-cell clone, 61.4F11 specific for MSP2 peptide P8 (7) was also used (Fig. 3B). The results of this experiment clearly showed that peptide P10-specific clone 61.1C8 responded in a dose-dependent manner to both MSP2 and MSP3, whereas peptide P8-specific clone 61.4F11, which recognizes a conserved MSP2 epitope not present in MSP3, responded only to MSP2. This result also indicated that the response by clone 61.1C8 to MSP3 was not due to contaminating MSP2.

FIG. 3.

MSP2 peptide P10- and peptide P16-specific CD4⁺ T-cell clones respond to epitopes naturally processed in both MSP2 and MSP3. CD4⁺ T-cell clone 61.1C8 specific for MSP2 peptide P10 (A), clone 61.4F11 specific for peptide P8 (B), or clone 61.2A1 specific for peptide P16 (C) were tested with low, medium, and high concentrations of antigen. For clone 61.1C8, these were 3.5, 7.0, and 14.0 μg of MSP2 and MSP3 per ml; 2.5, 5.0, and 10.0 μg of URBC per ml; and 0.1, 1.0, and 10.0 μg of peptides P8 and P10 per ml (A). For clone 61.4F11 these were 3.5, 7.0, and 14.0 μg of all antigens per ml (B). For clone 61.2A1 these were 0.1, 1.0, and 10.0 μg of URBC per ml; 0.4, 2.0, and 10.0 μg of MSP2 per ml; 1.8, 3.5, and 7.0 μg of MSP3 per ml; and 1.0 and 10.0 μg of peptide P16 or peptide P1 per ml. Results are presented as the mean counts per minute + 1 standard deviation of duplicate wells. Mean background counts per minute of T cells cultured with medium were 126 ± 16 cpm (clone 61.1C8), 137 ± 45 cpm (clone 61.4F11), and 7,792 ± 189 cpm (clone 61.2A1). Background counts per minute for clone 61.2A1 were subtracted. Asterisks indicate that the response was significantly different from the response to medium (P < 0.05). Results are representative of at least two separate experiments with each clone.

A second set of clones specific for MSP2 peptide P16 were derived by limiting dilution cloning of T lymphocytes stimulated with the peptide. CD4⁺ T-cell clones 61.2A1 and 61.2G8 were obtained, and the epitope that they recognized mapped to the sequence NFAYFGGELGVRFAF by use of a set of truncated peptides (data not shown). The clones responded to A. marginale homogenate, demonstrating that they recognize naturally processed native antigen (data not shown). These clones were then tested for proliferation against MSP2 and MSP3. Like the peptide P10-specific clone, these clones also responded to both MSP2 and MSP3 but not to peptide P8 (representative data for clone 61.2A1 are shown in Fig. 3C). Thus, epitopes at both ends of the conserved C region shared by MSP2 and MSP3 are recognized by Th-cell clones from MSP2-primed cattle.

MSP2 and MSP3 are surface proteins originally shown to be immunodominant by immunoprecipitation with sera from infected cattle (16). The striking immunodominance of A. marginale MSP2 has been attributed to both protein abundance and the presence of multiple CD4⁺ T-cell epitopes in the hypervariable region and flanking conserved regions (1, 6-8). The data presented here provide a third mechanism for immunodominance: processing of MSP2 T-cell epitopes from native MSP2 and MSP3 with presentation to CD4⁺ T cells. A previous study indicated that T-cell clones from outer membrane vaccinees responded to both proteins (9). The epitope-mapping data presented here for C-terminal peptides P10 and P16 and N-terminal peptide P7 establish that at least three epitopes can be cross-presented from MSP2 and MSP3. Furthermore, the density of T-cell epitopes in the shared regions of MSP2 and MSP3 (7) suggests that additional cross-priming and presentation may occur, further indicating the importance of shared T-cell epitopes in immunodominance. The N and C regions of MSP2 are highly conserved among A. marginale strains, and T-cell epitopes responding to these regions are also conserved (9).

Although MSP3 has been sequenced in only two strains, the N and C regions flanking the central hypervariable region of MSP3 are completely conserved in strains Florida and St. Maries, Idaho (http://www.vetmed.wsu.edu/research_vmp/anagenome/), indicating that the T-cell epitopes shared by MSP2 and MSP3 are a general occurrence among strains. We hypothesize that the immunodominant conserved epitopes in MSP2 and MSP3, which during infection undergo dynamic antigenic variation through gene conversion of pseudogenes into the expression site, may enable a rapid memory Th-cell response for efficient IgG production against newly emerging variants. This response to the conserved regions of MSP2 and MSP3 may contribute to controlling rickettsemia to levels observed in persistent infection.