Representative screen of individual K^b 9-mer peptides derived from the LCMV Armstrong proteome. The criteria for positivity were an SI of ≥2.0 and net SFCs/10⁶ cells of ≥20 in duplicate assays. The line at 20 net SFCs/10⁶ cells indicates the threshold used to define positivity. An asterisk indicates positive peptides. In total, 20 of 100 K^b 9-mer peptides were positive in the IFN-γ ELISPOT assay. The number of positive K^b 9-mer peptides was reduced to 18 after removing nested peptides containing the same core sequence but producing a lower response in the assay.

Frequency of CD8⁺ IFN-γ⁺ T cells against newly identified and known epitopes during LCMV infection. B6 mice were infected i.p. with LCMV. Day 8 postinfection splenocytes were harvested and stimulated in vitro with or without the indicated peptides at 0.1 μg/ml (boldface type indicates novel epitopes) in the presence of BFA and rhIL-2. Intracellular IFN-γ staining was used to enumerate the number of antigen-specific CD8⁺ T cells. A representative analysis for each epitope is shown, where the percentage of CD8⁺ T cells that stain positive for IFN-γ is indicated in the upper right quadrant.

Comparison of epitope identification methods. The 28 LCMV-specific epitopes are categorized by size (noncanonical versus canonical), method (predictive versus overlapping peptides), and reference (known versus new epitopes). The relative percentage of responding CD8⁺ T cells to each group of epitopes is shown, where the additive percentage of CD8⁺ T cells producing IFN-γ to each of the 28 individual epitopes is defined as 100% of the response.

Newly identified epitopes significantly contribute to overall magnitude of CD8⁺ T cell response against LCMV. (A) Splenocytes from day 8 LCMV-infected mice were stimulated with individual peptides at 0.1 μg/ml. The percentage of CD8⁺ T cells specific to either known or new epitopes was measured by intracellular staining for IFN-γ. The 9 previously characterized epitopes (9 known) consisted of the sum of responses against NP396-404, NP205-213, NP165-175, NP238-248, GP33-41, GP34-41, GP276-286, GP118-125, and GP92-101. The sum total response against the 19 new peptides was composed of GP44-52, GP221-228, GP365-372, GP166-173, L455-463, L349-357, L689-697, L2062-2069, L1369-1377, L156-163, L1878-1885, L1428-1435, L775-782, L338-346, L313-320, L663-671, L1189-1196, L743-751, and L1302-1310. (B) CD8⁺ T-cell responses were categorized by viral antigen. The NP response is against 4 NP epitopes, GP response against 9 GP epitopes, and L response against 15 L epitopes. (C) Total percentage of CD8⁺ CD44^hi T cells from day 8 LCMV-infected mice. Splenocytes were stained for CD8 and CD44, and the percentage of double-positive cells is indicated in the upper right quadrant. Results are representative of six independent experiments. (D) Comparison of the percentage of CD8⁺ CD44^hi T cells to the percentage of epitope-specific CD8⁺ IFN-γ⁺ T cells from LCMV-infected B6 mice on day 8 postinfection.

Representative screen of optimal determinants within an antigenic 15-mer peptide. (A) The optimal epitope was mapped within a 15-mer peptide, NP236-250, by analyzing CD8⁺ T-cell responses to a series of truncated peptides, including eight 8-mers, seven 9-mers, six 10-mers and five 11-mers (separated by dotted lines) in an IFN-γ ELISPOT assay using purified splenic CD8⁺ T cells from LCMV-infected B6 mice 8 days postinfection. The amino acid sequence of each peptide is indicated. Three highly antigenic 11-mer peptides, NP237-247, NP238-248, and NP239-249, were detected. (B) Dose-dependent CD8⁺ T-cell responses to the three antigenic 11-mer peptides were analyzed by IFN-γ ELISPOT assay to pinpoint the optimal 11-mer epitope. Only NP238-248 remained antigenic at low peptide doses and was therefore defined as the optimal epitope within the 15-mer peptide.

Evaluation of bioinformatics prediction accuracy and sensitivity. Each line corresponds to the prediction for one H-2^b allele and peptide size. The y axis represents the number of epitopes that were identified as a function of the prediction rank of each epitope on the x axis. Epitope ranking is based on the predicted IC₅₀s for each epitope generated by the prediction algorithms.

CD8⁺ T-cell dose-dependent response against known and newly identified LCMV-specific epitopes. Splenocytes from day 8 LCMV-infected mice were stimulated with candidate NP-specific peptides (A), GP-specific peptides (B), and L-specific peptides (C) from 10 pg/ml to 1 μg/ml. Intracellular IFN-γ staining was used to enumerate the number of antigen-specific CD8⁺ T cells. Data are presented as relative percentages of CD8⁺ T cells that produce IFN-γ following peptide stimulation, where 100% is defined as the maximal response regardless of the peptide dose tested. Closed markers indicate known epitopes, whereas open markers indicate newly identified epitopes. The dotted line indicates half of the maximal CD8⁺ T-cell response, which was used to define the EC₅₀ of each peptide. Data represent results from one of two or more independent experiments.

Functional role of L epitopes in vivo during acute LCMV infection. (A) Specific in vivo cytotoxicity of L peptide-coated target cells. Day 8 LCMV-infected B6 mice were intravenously injected with CFSE-labeled target cells coated with GP33-41, GP276-286, L2062-2069, L455-463, or L156-163 peptide (n = 3). After 5 h, spleens were harvested and analyzed for specific killing of peptide-coated targets. Representative histograms for each peptide are gated on CD45.1⁺ target cells in the spleen. Ratios of relevant peptide-coated (CFSE^high) and unlabeled marker (CFSE^low) donor cell populations were used to determine percentages of specific killing. Numbers correspond to the percentage of target cells killed. Data represent results from one of three independent experiments. (B) Comparison between cytotoxic activity and the percentage of antigen-specific CD8⁺ T cells producing IFN-γ. Each square represents the percentage of specific killing and the percentage of CD8⁺ IFN-γ⁺ T cells for 11 of the newly defined LCMV epitopes, GP33-41, GP276-286, NP165-175, and NP238-248. (C) Impact of L peptide vaccination on virus titer. B6 mice were primed with GP33-41 (open squares), L2062-2069 (closed squares), L156-163 (open circles), L455-463 (closed circles), or mock control (open triangles) with DMSO emulsified in incomplete Freund's adjuvant (n = 5). Twelve days later, mice were challenged i.p. with 1 × 10⁵ PFU LCMV. Virus titers were calculated per gram of spleen. When no plaque growth was detected, values were set at the threshold of detection (log₁₀ PFU/gram spleen = 4.3). Horizontal lines indicate mean values. Results are representative of two independent experiments.