Lytic and latent antigens of the human gammaherpesviruses Kaposi's sarcoma-associated herpesvirus and Epstein-Barr virus induce T-cell responses with similar functional properties and memory phenotypes

Abstract

The cellular immunity against Kaposi's sarcoma-associated herpesvirus (KSHV) is poorly characterized and has not been compared to T-cell responses against other human herpesviruses. Here, novel and dominant targets of KSHV-specific cellular immunity are identified and compared to T cells specific for lytic and latent antigens in a second human gammaherpesvirus, Epstein-Barr virus. The data identify a novel HLA-B57- and HLA-B58-restricted epitope in the Orf57 protein and show consistently close parallels in immune phenotypes and functional response patterns between cells targeting lytic or latent KSHV- and EBV-encoded antigens, suggesting common mechanisms in the induction of these responses.

FIG. 1.

EBV-specific T-cell activity exceeds responses to KSHV antigens. KSHV- and EBV-specific T cells were assessed by IFN-γ ELISPOT assay using peptide pools containing overlapping peptides spanning lytic and latent KSHV antigen sequences or previously described lytic or latent EBV antigen-derived CTL epitopes. The magnitude of positive responses in sera from 31 HIV-infected individuals tested is shown as SFC/10⁶ input PBMC. The cutoff for KSHV-specific responses is based on testing 20 KSHV-seronegative subjects that did not produce any positive responses (32). The majority (77%) of subjects received antiretroviral therapy and presented with a median CD4 count of 384 (range, 13 to 957) and a median viral load of 488 copies/ml (range, <50 to 440,000).

FIG. 2.

Mapping and HLA restriction analyses of an optimal CTL epitope in Orf57. Recognition of the optimal IF9 epitope, defined as the peptide truncation eliciting the strongest responses at the lowest concentration, is shown in a cell line from an HLA-B57-expressing subject (A) and an HLA-B58-expressing subject (B). The cell line was tested by ELISPOT assay using sequentially N- and C-terminally truncated peptides tested at 10-fold serial dilutions. (C) HLA restriction analysis was performed using an IF9-specific T-cell line from the HLA-B58-expressing subject, stimulated either (from left) by adding the IF9 peptide directly to the cell line or by pulsing single-HLA allele-transfected .221 cells and detecting intracellular IFN-γ by flow cytometry. Responses to HLA-B*5701- and HLA-B*5801-expressing .221 cells in the presence of the IF9 peptide are shown. Mock-transfected .221 cells served as negative controls.

FIG. 3.

Functional analysis of IFN-γ-producing CD8 T cells. CD107a expression and TNF-α release were assessed for individual responses detected in 17 samples from 11 individuals after stimulation with peptide pools derived from lytic or latent KSHV or EBV antigens. All analyses were based on gating on CD3⁺ CD8⁺ T cells that produced IFN-γ upon antigen exposure. The ratio of the percentage of cells that were CD107⁺ TNF⁺ to the percentage of cells that were CD107⁺ TNFa⁻ is shown for KSHV- and EBV-derived lytic and latent antigen-specific responses. The flow cytometry panels shown in Fig. 3 and 4 included antibodies against CD3 (Pacific Blue; Becton Dickinson [BD]), CD4 (allophycocyanin [APC]-Cy5.5; Caltag), CD8 (APC-Cy7; BD), CD45RA (phycoerythrin [PE]-Cy5.5; Caltag), CCR7 (PE-Cy7; BD), CD14 and CD19 (both peridinin-chlorophyll-protein complex [PerCP]; BD), CD107a (PE; BD), IFN-γ (fluorescein isothiocyanate; BD) and TNF-α (APC; BD).

FIG. 4.

Phenotypic analysis of IFN-γ-producing CD8 T cells. The expression of the effector-memory or the effector cell phenotype was compared between individual T-cell responses shown in Fig. 3. Cells responding to either KSHV (left panel) or EBV (right panel) lytic or latent antigens were gated on CD8⁺ CD3⁺ IFN-γ-producing cells. The ratio of the percentage of cells that were CCR7⁻ CD45⁺ to the percentage of cells that were CCR7⁻ CD45⁻ is shown.