Gag Protein Epitopes Recognized by CD4⁺ T-Helper Lymphocytes from Equine Infectious Anemia Virus-Infected Carrier Horses

Abstract

Antigen-specific T-helper (Th) lymphocytes are critical for the development of antiviral humoral responses and the expansion of cytotoxic T lymphocytes (CTL). Identification of relevant Th lymphocyte epitopes remains an important step in the development of an efficacious subunit peptide vaccine against equine infectious anemia virus (EIAV), a naturally occurring lentivirus of horses. This study describes Th lymphocyte reactivity in EIAV carrier horses to two proteins, p26 and p15, encoded by the relatively conserved EIAV gag gene. Using partially overlapping peptides, multideterminant and possibly promiscuous epitopes were identified within p26. One peptide was identified which reacted with peripheral blood mononuclear cells (PBMC) from all five EIAV-infected horses, and three other peptides were identified which reacted with PBMC from four of five EIAV-infected horses. Four additional peptides containing both CTL and Th lymphocyte epitopes were also identified. Multiple epitopes were recognized in a region corresponding to the major homology region of the human immunodeficiency virus, a region with significant sequence similarity to other lentiviruses including simian immunodeficiency virus, puma lentivirus, feline immunodeficiency virus, Jembrana disease virus, visna virus, and caprine arthritis encephalitis virus. PBMC reactivity to p15 peptides from EIAV carrier horses also occurred. Multiple p15 peptides were shown to be reactive, but not all infected horses had Th lymphocytes recognizing p15 epitopes. The identification of peptides reactive with PBMC from outbred horses, some of which encoded both CTL and Th lymphocyte epitopes, should contribute to the design of synthetic peptide or recombinant vector vaccines for EIAV.

Significant barriers exist to the development of effective vaccines that induce protection against the lentiviruses that cause severe disease in both humans and animals. Reverse transcription of lentiviral RNA by an error-prone RNA-dependent DNA polymerase results in rapid genotypic variation and the subsequent development of quasispecies within infected individuals (7). Some of the resulting virions differ at immunologically relevant epitopes, thereby eluding existing immune mechanisms, and allowing for further dissemination throughout the host’s tissue (7). Lentiviral DNA integrates into the host genomic DNA and evades immune responses until viral proteins are expressed (9). The development of immune responses capable of recognizing conserved epitopes expressed on virus and viral-infected cells is critical to the control of virus replication and therefore clinical disease.

One lentivirus vaccine strategy is to use subunit preparations, which circumvents some of the problems associated with the use of modified live or inactivated whole-virus preparations (7, 33). Modified live vaccines are efficacious against numerous viral diseases, but concern over reversion to a virulent phenotype or recombination with endogenous or exogenous viruses have limited their use against lentiviruses (8). Inactivated whole-virus preparations are useful against some diseases, but these vaccines generally do not generate a potent major histocompatibility complex (MHC) class I-restricted cytotoxic T-lymphocyte (CTL) response necessary to clear lentiviral-infected cells and are not effective against lentivirus variants (15, 29). Subunit vaccines meet many of the requirements for safety, but improvements in vaccine design and/or delivery are needed in order to induce protective and possibly sterilizing immunity to lentiviruses.

Synthetic peptide vaccines have numerous advantages including the inclusion of only defined epitopes to generate relevant immune responses rather than a broad response that may exacerbate infectivity and disease (4, 23, 29, 48). Exacerbation of lentiviral disease occurs in goats immunized with whole-inactivated caprine arthritis encephalitis virus (23) and horses immunized with equine infectious anemia virus (EIAV) envelope subunit vaccines (29). Antibody directed against the human immunodeficiency virus type 1 (HIV-1) envelope glycoproteins can cause receptor-mediated phagocytosis of antibody-coated virus by a monocytic cell line (46); such reactions in vivo may facilitate infection of target cells. Effective subunit vaccines have been produced which direct immune responses away from disease-potentiating epitopes and instead stimulate responses against epitopes that cause virus neutralization or cytolysis of virus-infected cells (2, 5, 31, 40, 41). These results and advances in peptide vaccine design highlight the need for inclusion of multiple epitopes, especially the requirement for covalent linkage of CTL and Th lymphocyte epitopes for efficient induction of CTL responses (42). Collinear synthesis of peptides containing CTL and Th lymphocyte epitopes has several advantages over conjugating the CTL epitope to a carrier protein and allows for construction of molecules with known immunologic properties (4). Recent studies suggest that multivalent synthetic peptides containing B lymphocyte, CTL, and Th lymphocyte epitopes are functional and indeed can elicit protective immune responses including neutralizing antibody and CTL (1). Given the rapid antigenic variation of lentiviruses, it is likely that multideterminant peptides encoding CTL and Th lymphocyte epitopes, which are conserved among lentivirus isolates, will be necessary to stimulate the requisite cross-protective immune responses necessary to control these viruses.

Towards this end, we have recently identified epitopes recognized by equine leukocyte antigen (ELA)-A MHC class I-restricted CTL from horses infected with EIAV (49). This naturally occurring lentivirus of horses has nucleotide, structural, and antigenic similarities to HIV-1 (14) and infects cells of the macrophage/monocyte lineage (11, 22, 30). These characteristics, and the observation that horses routinely suppress virus replication and control disease in the face of rapid antigenic variation, make EIAV a relevant animal model for investigating control of lentivirus infections. Identification of CTL and Th lymphocyte epitopes, recognized by peripheral blood mononuclear cells (PBMC) from horses controlling EIAV replication, provides information necessary to determine the role of responses to these epitopes in controlling EIAV in carrier horses. This information may be relevant to other lentivirus systems, including HIV-1.

To extend studies which identified EIAV-specific CTL epitopes (49), we used a series of overlapping peptides from the p26 (capsid) and p15 (matrix) proteins of EIAV to identify Th lymphocyte epitopes. PBMC from all five outbred, EIAV-infected horses examined reacted with purified p26 protein. Multiple Th lymphocyte epitopes were identified in the p26 and p15 proteins; however, there was marked variation in PBMC reactivity from horse to horse. Four multideterminant peptides encoding both CTL and Th lymphocyte epitopes were identified within the p26 region. These results extend current knowledge of EIAV by identifying Th lymphocyte epitopes within the p26 and p15 proteins. The identification of these epitopes may contribute to subunit vaccine design by providing the Th lymphocyte stimulus now recognized as being necessary for augmenting CTL responses to synthetic peptides and peptide-expressing recombinant vectors (42, 45).

MATERIALS AND METHODS

Experimental horses.

The horses used in these experiments were adult mixed breed ponies and included five horses infected with EIAV_WSU5 and three that were EIAV-negative. EIAV_WSU5 is a pathogenic tissue culture-adapted strain derived by three sequential back passages of the prototype strain in horses and subsequent biological cloning on equine kidney cell cultures (34). The prototype strain is a tissue culture-adapted strain isolated from the highly virulent Wyoming strain of EIAV (19). All infected horses experienced at least one episode of fever, anemia, and viremia immediately following experimental EIAV inoculation and were infected at least 2 years prior to use in this study (18, 26, 49).

p26 immunoblots.

To confirm the identity and size of the isolated EIAV p26 protein and to demonstrate the absence of other EIAV proteins, immunoblot assays were performed. Briefly, 2 × 10⁶ EIAV-infected equine kidney (EK) cells were lysed with 0.5% Nonidet P-40 in 50 mM of Tris buffer (pH 8.0) containing 5 mM EDTA, 5 mM iodoacetamide, 0.1 mM Nα-p-tosyl-l-lysine chloromethyl ketone and 1 mM PMSF for 30 min on ice and then centrifuged at 12,000 × g for 5 min. Proteins in the supernatants were separated on a 10% Tris-glycine sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and transferred to nitrocellulose. Lysates of JM 109 bacteria transformed with a plasmid expressing EIAV p26 as a fusion protein with glutathione transferase were obtained (24). This recombinant plasmid was made by inserting the p26 gene into the pGEX expression plasmid (24). EIAV p26 was obtained from transformed bacterial lysates by binding the fusion protein to a glutathione-agarose column and cleaving the fusion protein with thrombin. Finally, 20 ng of purified p26 was separated by SDS-PAGE and transferred to nitrocellulose membranes. The nitrocellulose membranes were probed with pre- and post-EIAV-infection serum from horse A1924 (34) or with a monoclonal antibody (MAb) (EIA6A1) recognizing an epitope in the capsid protein p26 (25, 34, 36). The secondary antibodies were a peroxidase-labeled goat anti-horse immunoglobulin G (IgG) (heavy plus light chains [H+L]) used at a 1:2,000 dilution (Kirkegaard and Perry Laboratories, Gaithersburg, Md.) for A1924 serum and a peroxidase-labeled goat anti-mouse IgG (H+L) used at a 1:4,000 dilution (Kirkegaard and Perry Laboratories) for MAb EIA6A1. Probed membranes were incubated in chemiluminescent substrate for horseradish peroxidase (DuPont Company, NEN Life Science Products, Boston, Mass.), and signals were detected by exposure of radiograph film.

Synthetic peptides.

All p26 and p15 specific peptides utilized in this study were synthesized with an automated Applied Biosystems 431 synthesizer (Foster City, Calif.) by the solid-phase method based on Fmoc chemistry (49). In previous experiments, retroviral vectors specific for p15, p26, p11, and p9 were constructed and used to transduce EK cells for use as CTL targets (49). The p15 and p26 proteins contained CTL epitopes, while the p11 and p9 proteins did not. Based on this reactivity, peptides were made and used to map CTL epitopes (49). These same peptides were used in the current study to map Th lymphocyte epitopes of the p15 and p26 proteins (Fig. 1). Sixteen 20-mer peptides, one 22-mer peptide, and one 25-mer peptide were synthesized for the p26 region (Table 1). Nine 20-mer peptides and one 16-mer peptide were synthesized for the p15 region (Table 2). The peptides were sequential and overlapped one another by eight amino acids. Stock solutions of peptides were dissolved in phosphate-buffered saline (PBS) containing 10% dimethyl sulfoxide at a concentration of 10 mg/ml and diluted in RPMI 1640 medium (Gibco-BRL) containing 10% heat-inactivated horse serum–25 mM HEPES–55 μM 2-mercaptoethanol–50-μg/ml gentamicin (complete RPMI medium) prior to use in proliferation assays.

FIG. 1.

Proteins encoded by the gag gene of EIAV. Regions covered by overlapping peptides are in bold, and the major homology region (MHR) is demarcated.

TABLE 1.

Synthetic p26 peptides

Peptide no.a	Amino acid no.b	Peptide sequencec
1	125–144	PIMIDGAGNRNFRPLTPRGY
2	137–156	RPLTPRGYTTWVNTIQTNGL
3	149–168	NTIQTNGLLNEASQNLFGIL
4	161–180	SQNLFGILSVDCTWEEMNAF
5	173–192	TSEEMNAFLDVVPGQAGQKQ
6	185–204	PGQAGQKQILLDAIDKIADD
7	197–216	AIDKIADDWDNRHPLPNAPL
8	209–228	HPLPNAPLVAPPQGPIPMTA
9	221–245	QGPIPMTARFIRGLGVPRERQMEPA
10	242–261	MEPAFDQFRQTYRQWIIEAM
11	254–273	RQWIIEAMSEFIKVMIGKPK
12	266–285	KVMIGKPKAQNIRQGAKEPY
13	278–297	RQGAKEPYPEFVDRLLSQIK
14	290–309	DRLLSQIKSEGHPQEISKFL
15	302–321	PQEISKFLTDTLTIQNANEE
16	314–333	TIQNANEECRNAMRHLRPED
17	326–345	MRHLRPEDTLEEKMYACRDI
18	338–359	KMYACRDIGTTKQKMMLLAKAL

^aThese numbers are used to designate the p26 peptides in Fig. 5. 

^bPeptide position in the EIAV Gag protein (43). 

^cAmino acid sequence of synthetic peptides. 

TABLE 2.

Synthetic p15 peptides

Peptide no.a	Amino acid no.b	Peptide sequencec
1	1–20	MGDPLTWSKALKKLEKVTVQ
2	13–32	KLEKVTVQGSQKLTTGNCNW
3	25–44	LTTGNCNWALSLVDLFHDTN
4	37–56	VDLFHDTNFVKEKDWQLRDV
5	48–67	KDWQLRDVIPLLLEDVTQTLS
6	60–79	EDVTQTLSGQEREAFERTWW
7	72–91	EAAFERTWWAISAVKMGLQIN
8	84–103	VKMGLQINNVVDGKASFQLL
9	96–115	GKASFQLLRAKYEKKTANKK
10	108–124	EKKTANKKQSEPSEEY

^aThese numbers are used to designate the p15 peptides in Fig. 6. 

^bPeptide position in the EIAV Gag protein (43). 

^cAmino acid sequence of synthetic peptides. 

PBMC proliferation assay.

PBMC were isolated from whole blood from normal and EIAV-infected horses with Histopaque 1077. Cells were resuspended in complete RPMI medium. PBMC preparations were assayed in at least triplicate wells in round-bottom, 96-well microplates (Corning); 2 × 10⁵ cells per well were incubated for 7 days at 37°C in 5% CO₂ with purified p26 or peptides in 100 μl of complete RPMI medium. On day 6, cells were labeled with 0.25 μCi of [³H]thymidine (6.7 Ci/mmol; Amersham)/well for 18 h and harvested with a Tomtec cell harvester. Incorporation of [³H]thymidine was quantified by liquid scintillation counting in a β-plate reader (Wallac, Gaithersburg, Md.). Stimulation indices (SI) were calculated by dividing the mean counts per minute of p26 or peptide-stimulated PBMC by the mean counts per minute of PBMC incubated in medium alone (negative control).

Flow cytometry.

All cell analysis was performed with a FACScan equipped with a Power Macintosh 7100/80 computer and CellQuest software (Becton-Dickinson, San Jose, Calif.). Lymphoblast populations were defined based on log side scatter versus linear forward scatter, and subpopulations of the lymphoblasts were then identified based on MAb reactivity with single color fluorescence on side scatter versus log intensity of fluorescence scale. Cells were labeled for flow cytometry as previously described (47) with anti-CD2 MAb (HB88A), anti-CD8 MAb (HT14A), or anti-CD4 MAb (HB61A) (17). All stained cells were fixed in 2% formaldehyde and analyzed prior to and following 7-day stimulation with EIAV p26.

In order to determine the contribution of CD8⁺ T lymphocytes to the proliferative responses, CD8⁺ T lymphocytes were removed from isolated PBMC prior to 7-day stimulation with EIAV p26. PBMC were isolated from whole blood from horses H513, H529, and H532 with Histopaque 1077 following standard protocols. Isolated PBMC were washed three times with Hank’s buffered saline solution and then resuspended in 2 ml of PBS with 5 mM EDTA and 0.5% bovine serum albumin (PBS-EDTA-BSA) at a final concentration of 1.5 × 10⁷ cells/ml. The PBMC were labeled with 15 μg of MAb HT14A (anti-equine CD8) (17) and incubated on ice for 15 min. Labeled cells were washed three times with PBS-EDTA-BSA and resuspended in 240 μl of PBS-EDTA-BSA to which 60 μl of microbeads coated with goat anti-mouse antibodies (Miltenyi Biotec, Bergisch Gladbach, Germany) were added and then incubated at 4°C for 15 min. After preparing a separation column following the manufacturer’s directions, magnetically labeled cells were passaged through the column and effluent was collected. PBMC depleted of CD8⁺ lymphocytes were then stimulated for 7 days with EIAV p26 and analyzed by flow cytometry as described above.

RESULTS

Evaluation of purified recombinant p26 (capsid) protein.

EIAV gag encodes a polyprotein containing p15 (matrix), p26 (capsid), p11 (nucleocapsid), and p9 (whose function is unknown) (Fig. 1). p26 constitutes approximately 30% of the total virion protein mass and has B lymphocyte and CTL epitopes, some of which are conserved among EIAV isolates (13, 28, 49). Immunoblots evaluating purified p26 from a recombinant bacterial expression system are shown in Fig. 2. Lysate containing the p26-glutathione transferase fusion protein, control lysate, and thrombin-cleaved and purified p26 were probed with preinfection and postinfection serum obtained from EIAV-infected horse A-1924 (34), with MAb EIA6A1 reactive with p26, and with isotype control MAb. Preinfection serum and the isotype control MAb were unreactive in the immunoblots. Postinfection serum from A-1924 recognized a single reactive protein in recombinant bacterial lysate migrating at 55 kDa (Fig. 2, lane 5) and representing uncleaved p26-glutathione transferase fusion protein and recognized a 26-kDa protein (Fig. 2, lane 6) in the purified p26 fraction. MAb EIA6A1 also recognized the 55-kDa fusion protein in bacterial lysate (Fig. 2, lane 11) and only a 26-kDa protein in the purified p26 (Fig. 2, lane 12).

FIG. 2.

Immunoblot of purified EIAV p26 protein. Lanes 1, 4, 7, and 10 contain negative control pGEX-transformed bacterial lysate. Lanes 2, 5, 8, and 11 contain p26–glutathione-transferase fusion protein. Lanes 3, 6, 9, and 12 contain purified EIAV p26. Lanes 1 to 3, 4 to 6, 7 to 9, and 10 to 12 were probed with preinfection horse serum, postinfection horse serum, isotype control MAb, and MAb EIA6A1 to EIAV p26, respectively. Molecular mass is indicated in kilodaltons at left.

PBMC proliferative responses to EIAV p26.

To determine if PBMC from EIAV-infected horses recognized epitopes on isolated p26 protein, PBMC cultures from five EIAV-infected horses and three EIAV-uninfected horses were initiated and incubated with 0.2 to 20 μg of purified p26/ml. The level of p26 reactivity varied from horse to horse, but PBMC from all five EIAV-infected horses had a SI greater than 2.5 in all assays, with the SI ranging from 3.8 to 9.7 in the results presented in Fig. 3. The proliferative response to p26 increased with increasing amounts of p26, and a concentration of 20 μg/ml caused the highest and most consistent stimulation (data not shown). PBMC isolated from H532 invariably had the highest SI in response to p26, reaching an SI of 83 in one assay (data not shown). In contrast, PBMC isolated from H529 usually had the lowest SI, ranging from 2.5 to 4.0. PBMC from negative control horses H565, H567 (Fig. 3), and H545 (data not shown) had SI scores ranging from 0.4 to 1.3, which were considered nonsignificant.

FIG. 3.

Proliferation of PBMC from EIAV-infected (H513, H521, H529, H532, and H540) and negative control (H565 and H567) horses stimulated 7 days with EIAV p26 protein (20 μg/ml). Results are expressed as mean counts per minute of six replicate wells. Vertical lines represent standard deviations. The horizontal lines represent an SI score of 2 for each horse, and all SI scores above 2 were considered significant. The SI of 2 was calculated by multiplying the mean counts per minute of medium control wells by two.

CD4⁺ T lymphocyte proliferative response to p26.

To determine the phenotype of reactive cells in stimulated PBMC cultures from H513, H529, and H532, flow cytometry was performed after 1 week of stimulation with p26. Lymphoblasts were identified based on linear forward scatter and were defined as those cells greater than 500 on the forward scatter (FSC-H) scale (Fig. 4d). The percentage of lymphoblasts ranged from a low of 41% for H529 to 55% for H532, and the flow cytometry data for H532 are presented in Fig. 4. Based on anti-CD4 MAb reactivity, it was determined that the proportion of lymphoblasts expressing CD4 ranged from 78% for H529 to 90% for H532, demonstrating that CD4⁺ T lymphocytes were responsible for most of the proliferative responses noted in these cultures. This conclusion was further confirmed by demonstrating that proliferation in PBMC cultures from the three EIAV-infected horses stimulated with p26 was not affected by depletion of CD8⁺ T-lymphocytes. Following CD8⁺ T lymphocyte depletion and a 1-week stimulation with p26 less than 1% of proliferating cells expressed CD8 when examined by flow cytometry (data not shown).

FIG. 4.

Enumeration of CD4⁺ lymphoblasts in PBMC stimulated with p26. The results shown are for isotype control PBMC stimulated with medium only (A), CD4⁺ PBMC prior to stimulation with p26 (B), CD4⁺ PBMC stimulated for 7 days with p26 (C), all PBMC stimulated for 7 days with p26 (D), and lymphoblasts as defined for panel D which were also stained with anti-CD4 MAb (E). All cells to the right of channel 500 in panel D were defined as lymphoblasts. The cells surrounded by an oval in panel E were defined as CD4⁺ lymphoblasts.

Proliferative responses of PBMC to synthetic p26 peptides.

Eighteen overlapping peptides, enumerated as previously described (49), were used to map Th epitopes recognized by PBMC from five EIAV-infected horses. In preliminary experiments, an increase in response was noted in cultures stimulated with increasing peptide concentrations. Proliferative responses generally became evident at a 5 μg/ml concentration with maximal responses noted at 20 μg/ml (data not shown). PBMC from the five EIAV-infected (H513, H521, H529, H532, and H540) and one EIAV-negative horse (H545) were reacted with each of the peptides listed in Table 1 for 7 days at a concentration of 5 μg/ml, except for H540 PBMC which were stimulated with peptides at 10 μg/ml. In addition, the proliferative capacity of PBMC from each horse was determined by stimulation of cultures for 7 days in the presence of 20 U of recombinant human interleukin-2 (IL-2)/ml. The SI for IL-2-stimulated PBMC varied from 2.2 to 35.0 in all horses examined, including EIAV-negative horses H545, H565, and H567, with no significant difference in proliferative responses to IL-2 between EIAV-infected and EIAV-negative horses (data not shown).

PBMC from the five EIAV-infected horses reacted with 2 to 13 of the 18 synthetic p26 peptides (Fig. 5) with a variety of Th lymphocyte reactivity patterns. The patterns were similar to those described for human outbred populations (5, 12). Nevertheless, some epitopes were recognized by PBMC from the majority of EIAV-infected horses in this study. p26 peptides 149-168, 221-245, and 242-261 were each reactive with PBMC from at least four of five EIAV-infected horses (Fig. 5). p26 peptide 254-273 caused proliferation of PBMC from five of five infected horses (Fig. 5). Proliferative responses were generally lower in PBMC from H521 and H540. PBMC from H540 recognized only 2 of 18 peptides. This was in contrast to PBMC from H529 and H532, both of which recognized 14 of the 18 peptides.

FIG. 5.

Proliferation of PBMC from EIAV-infected (H513, H521, H529, H532, and H540) and EIAV-negative (H545) horses. PBMC were stimulated for 7 days with overlapping p26 peptides at a concentration of 5 μg/ml except H540 PBMC which were stimulated at 10 μg/ml. Peptide numbers are defined in Table 1. Results are expressed as counts per minute and represent the means of triplicate wells. The horizontal line represents an SI score of 2, and all SI scores above 2 were considered significant. The SI of 2 was calculated by multiplying the mean counts per minute of medium control (M) wells by two. Vertical lines represent standard deviations.

The maximum number of reactive epitopes cannot be discerned from data in Fig. 5 because of the overlapping nature of these peptides and the length variation of epitopes recognized by Th lymphocytes (39). PBMC from H545 (EIAV-negative) failed to react to any of the synthetic p26 peptides, indicating that the proliferative response was specific to PBMC from EIAV-infected horses (Fig. 5). This specificity was confirmed by the lack of reactivity of PBMC cultures from two other EIAV-negative horses, H565 and H567, when stimulated with peptides 149-168, 173-192, 221-245, 242-261, 254-273, 278-297, 290-309, 302-321, and 338-359 (data not shown). These peptides were each recognized by PBMC from at least three of the five EIAV-infected horses (Fig. 5). Of importance was the observation that peptides previously determined to contain CTL epitopes (49) were also found to contain Th lymphocyte epitopes. PBMC from H529 recognized both CTL and Th lymphocyte epitopes on peptides 197-216 and 209-228, PBMC from H521 recognized CTL and Th lymphocyte epitopes in peptide 278-297, and finally PBMC from H532 recognized CTL and Th lymphocyte epitopes in peptide 338-359.

Proliferative responses of PBMC to synthetic p15 peptides.

A panel of 10 overlapping synthetic p15 peptides were reacted with PBMC isolated from EIAV-infected horses H521, H532, and H540 to determine if this protein contained reactive Th lymphocyte epitopes (Fig. 6). Of the three EIAV-infected horses, H521 PBMC did not recognize p15 peptides in two separate assays. PBMC from the two remaining EIAV-infected horses, H532 and H540, each recognized multiple p15 peptides. H532 PBMC, which responded to the highest number of p26 peptides, had the highest SI and also recognized the greatest number of p15 peptides (9 of 10). H540 PBMC recognized 4 of 10 p15 peptides, i.e., peptides 37-56, 48-67, 72-91, and 84-103. Reactivity to regions spanning residues 37 to 67 and 72 to 91 could be due to epitope overlap and thereby represent only two unique epitopes, but several clustered epitopes, similar to patterns noted in Fig. 5, must also be considered. PBMC from EIAV-negative horses H565 and H567 were nonreactive to all p15 peptides, indicating that the response by PBMC from infected horses was to p15 and that peptides were not nonspecifically mitogenic (Fig. 6).

FIG. 6.

Proliferation of PBMC from EIAV-infected (H521, H532, and H540) and EIAV-negative (H565 and H567) horses. PBMC were stimulated for 7 days with overlapping p15 peptides at a concentration of 10 μg/ml. Peptide numbers are defined in Table 2. Results are expressed as mean counts per minute and represent the means of six replicate wells. The horizontal line represents an SI score of 2, and all SI scores above 2 were considered significant. The SI of 2 was calculated by multiplying the mean counts per minute of medium control well (M) by two. Vertical lines represent standard deviations.

DISCUSSION

The generation of humoral and cytotoxic T-lymphocyte responses critical to the control of viral infections can be augmented by Th lymphocyte activation (3, 37, 38). In this study, we have identified multiple Th lymphocyte epitopes within the p26 and p15 proteins of EIAV. Identification of these epitopes, encoded by the relatively conserved gag gene, complements results of recent studies in which we identified CTL epitopes within the same regions (49). Identification of epitopes recognized by Th lymphocytes and other effector lymphocytes is important to understanding the in vivo immune responses that contribute to EIAV control. In addition, recent studies have demonstrated that covalent linkage of both Th lymphocyte and CTL epitopes can greatly potentiate induction of CTL in vivo (32, 35, 42, 45). Others have determined that, for induction of HIV-1 gag-specific CTL, there may be a requirement for Th lymphocytes along with coexpression of Th lymphocyte and CTL epitopes on the surface of the same antigen-presenting cell, suggesting that a three-cell complex is necessary for CTL stimulation (44). Identification of Th lymphocyte epitopes that potentiate protective CTL responses will allow their evaluation in synthetic peptide and recombinant vaccine preparations.

In this study, it was noted that p26 stimulated marked proliferative responses in all five EIAV-infected horses examined, indicating that p26 had at least one Th lymphocyte epitope recognized by PBMC from each horse. The region encompassing EIAV p26 amino acids 221 to 273 was particularly reactive. This clustering of epitopes is consistent with the observations that regions of proteins, rich in overlapping MHC binding motifs for different MHC alleles, have multiple reactive epitopes (27). p26 peptide 254-273 caused proliferation of PBMC from all five EIAV-infected horses. Three other peptides were identified in this region, each of which caused proliferation of PBMC from at least four of five EIAV-infected horses. Peptide 254-273 likely contains a “promiscuous” or “degenerate” epitope, as defined by the ability to cause proliferation in horses with dissimilar MHC class II haplotypes. Promiscuous Th lymphocyte epitopes are candidates for inclusion in synthetic peptide vaccines directed against EIAV epitopes recognized by CTL and neutralizing antibody as they would be able to generate Th lymphocyte responses in the context of different MHC class II molecules. Four peptides that encoded both CTL and Th lymphocyte epitopes (multi-determinant peptides) were also identified. PBMC from H529 proliferated in response to peptides 197-216 and 209-228, PBMC from H521 proliferated in response to peptide 278-297, and PBMC from horse H532 proliferated in response to peptide 338-359. These peptides were previously recognized as containing CTL epitopes recognized by horses with different ELA-A haplotypes except peptide for 197-216, which was only recognized by A1/A5 MHC class I restricted CTL (49). Also of interest was the observation that amino acids 197 to 309 was an area where multiple Th lymphocyte epitopes were clustered that were recognized by EIAV-infected horses. This area of the Gag protein is highly conserved among not only EIAV isolates but has significant sequence similarity with isolates of puma lentivirus, feline immunodeficiency virus, Jembrana disease virus, visna virus, caprine arthritis encephalitis virus, HIV-1, and simian immunodeficiency virus. The region spanning amino acids 277 to 292 of EIAV Gag has 75% amino acid sequence identity and, when taking into account conserved amino acid substitutions, 94% similarity with the human endogenous retrovirus. This same region shares 63% amino acid identity and 81% similarity with the corresponding region in HIV-1. This region is defined as the major homology region in HIV (20) and contains a Th lymphocyte, CTL, and two B lymphocyte epitopes (31). B lymphocyte epitopes also occur in the major homology region of simian immunodeficiency virus and feline immunodeficiency virus isolates (21). In addition, major B lymphocyte epitopes of EIAV p26 are located within the carboxy-terminal region of p26 spanning amino acids 277 to 352 (12). This is the same region recognized in this study to contain both Th lymphocyte epitopes and recognized in a previous study to contain a CTL epitope (49). Specifically, peptide 278-297 and peptide 338-359 encode for both CTL and Th lymphocyte epitopes recognized by CTL and PBMC from H529 and H532, respectively. The carboxy-terminal region also contains the major B lymphocyte epitopes of HIV-1 p24 (6). Though B lymphocyte responses to Gag would likely be ineffective in protection, these observations highlight the possibility that conserved regions encoding multideterminant peptides containing Th lymphocyte, CTL, and possibly B lymphocyte epitopes may induce cross-protective immune responses against EIAV strains. In addition, peptides from similar regions of related lentiviruses might be useful in inducing protection against these lentiviruses in their host species.

Results from this study also demonstrated that multiple Th lymphocyte epitopes are present in the EIAV p15 (matrix) protein of EIAV. As with the p26 peptides, the patterns of PBMC reactivity varied among the three EIAV-infected horses examined, and in fact, PBMC from H521 failed to react with p15 peptides while PBMC from H532 responded to 9 of 10 peptides. This observation emphasizes the requirement for careful screening of peptides and the judicious selection of reactive peptides encoding Th lymphocyte epitopes which have the potential for binding diverse MHC class II molecules and that are conserved among divergent lentivirus isolates, if they are to be used in synthetic peptide vaccines.

Synthetic peptide vaccines containing peptides with multiple Th lymphocyte, B-cell, and CTL epitopes can elicit protective responses against microorganisms (1, 10). This strategy of linking epitopes in peptides has proven superior to using peptide-carrier protein conjugates because of the phenomenon of epitope-specific suppression. This suppression occurs when preexisting immunity or an immunodominant response to the carrier protein (16) prevents induction of epitope-specific antibody against a synthetic peptide, conjugated to the carrier protein. Furthermore, the use of smaller peptides can circumvent the problem of some large proteins eliciting inappropriate immune responses that may augment infection and/or disease progression (4, 23). Efficacious vaccines will need to induce not only strong Th lymphocyte, CTL, and B lymphocyte responses but also have the potential for priming the immune response so that a recall response will be elicited at the time of virus infection. The use of some promiscuous Th epitopes, such as those on tetanus toxoid which are effective in inducing Th lymphocytes that enhance CTL responses, are limited because Th lymphocyte cross-reactivity between the vaccine and the pathogen is lacking (10, 16). Therefore, identification of Th lymphocyte epitopes is necessary for subunit vaccine development for lentiviruses.

In conclusion, this study has identified multiple Th lymphocyte epitopes which are likely candidates for inclusion in a synthetic vaccine for EIAV. In particular, peptide 254-273 was recognized by all five EIAV-infected horses and may contain a promiscuous epitope. Another potential candidate is peptide 278-297, which was recognized by Th lymphocytes from three of five horses and also contained a CTL epitope. In addition, three other multideterminant peptides were identified which encode for CTL and Th lymphocyte epitopes. The majority of Th lymphocyte epitopes found in EIAV p26 cluster at the carboxy terminus, a region with significant sequence identity to other lentiviruses, including HIV-1. Furthermore, the identification of Th lymphocyte epitopes in the same conserved EIAV p26 region as CTL (49) and B lymphocyte (12) epitopes is in accordance with previous observations in HIV-1. Additionally, multiple Th lymphocyte epitopes have been identified in EIAV p15. EIAV infections are controlled with remarkable regularity by horses, and dissection of immune responses involved in this control, including identification of Th lymphocyte epitopes, may provide information necessary for induction of protective immune responses against EIAV and other lentiviruses.