Functional heterogeneity and high frequencies of cytomegalovirus-specific CD8(+) T lymphocytes in healthy seropositive donors

Abstract

Human cytomegalovirus (HCMV) infection is largely asymptomatic in the immunocompetent host, but remains a major cause of morbidity in immunosuppressed individuals. Using the recently described technique of staining antigen-specific CD8(+) T cells with peptide-HLA tetrameric complexes, we have demonstrated high levels of antigen-specific cells specific for HCMV peptides and show that this may exceed 4% of CD8(+) T cells in immunocompetent donors. Moreover, by staining with tetramers in combination with antibodies to cell surface markers and intracellular cytokines, we demonstrate functional heterogeneity of HCMV-specific populations. A substantial proportion of these are effector cytotoxic T lymphocytes, as demonstrated by their ability to lyse peptide-pulsed targets in "fresh" killing assays. These data suggest that the immune response to HCMV is periodically boosted by a low level of HCMV replication and that sustained immunological surveillance contributes to the maintenance of host-pathogen homeostasis. These observations should improve our understanding of the immunobiology of persistent viral infection.

FIG. 1

Staining of CD8⁺ T-lymphocyte cultures with HCMV-specific HLA-peptide tetramers. Staining of CD8⁺ T-lymphocyte cultures with the tetrameric complex HLA-A*0201/pp65^495–503. (a) A polyclonal CD8⁺ T-lymphocyte line specific for HCMV pp65^495–503 is stained with a high degree of fluorescence by the tetramer. A control HLA-A*0201-restricted CTL line generated against influenza virus matrix peptide 58–66 does not show staining with this tetramer. (b) Tetramer HLA-B*0702/pp65^417–426 binds specifically to a CTL clone grown on the pp65^417–426 peptide, but fails to recognize a nonspecific HLA-B*0702-restricted CTL clone.

FIG. 2

Representative data of tetramer staining of PBMCs from HCMV-seropositive and -seronegative donors. PBMCs from HCMV-seropositive and -seronegative donors were tested for their ability to bind the tetrameric complexes used in this investigation. Representative dot plots where tetrameric staining is represented along the x axis and CD8 staining is represented along the y axis are summarized. Both HLA-B*0702/pp65^417–426 and HLA-A*0201/pp65^495–503 tetramers display their ability to stain HCMV-specific CD8⁺ T lymphocytes in seropositive donors, yet display minimal binding to CD8⁺ T cells from HCMV-seronegative individuals. Plots a to c represent HLA-A*0201⁺ CMV-seropositive donors, and these samples were stained with A*0201/pp65^495–503 tetrameric complexes. Plots d to f are representative HLA-B*0702⁺ CMV-seropositive donors, and these samples were stained with the B*0702/pp65^417–426 tetramer. Plots g and h are HLA-A*0201⁺ CMV-seronegative donors which were stained with the A*0201/pp65^495–503 tetramer. Plot i is an HLA-B*0702⁺ CMV-seronegative sample which was stained with the B*0702/pp65^417–426 tetramer.

FIG. 3

Dual expression of CD45RA and CD45RO isoforms on tetramer-reactive CD8⁺ T lymphocytes from donor 13. Four-color FACS analysis, using a tetrameric complex in conjunction with anti-CD8, anti-CD45RA and anti-CD45RO MAbs, was performed with freshly isolated PBMCs from donor 13 in an attempt to estimate the proportion of cells which were positive for both the RA and RO isoforms of CD45. CD8⁺ tetramer-positive and CD8⁺ tetramer-negative cells were gated, and the percentages of cells coexpressing CD45RA and CD45RO within these subsets were estimated.

FIG. 4

Intracellular cytokine and chemokine in tetramer-reactive CD8⁺ cells. Freshly isolated PBMCs from donor 16 were permeabilized and stained with MAbs to CD69, IFN-γ, MIP-1β, TNF-α, and perforin either immediately following isolation or 6 h postactivation with cognate peptide. Both nonstimulated and peptide-stimulated CD8⁺ tetramer-positive cells were gated separately (a), and the proportions of cells within the given subsets expressing cytokines and chemokines (b) are summarized.

FIG. 5

Fresh cytotoxicity of PBMC fractions from HCMV-positive donors. Autologous ⁵¹Cr-labelled L-BCLs from donors 13 (a) and 17 (b) were incubated with antigenic peptide at a final peptide concentration of 5 μM and were used as targets in a 15-h ⁵¹Cr-release assay. Freshly isolated PBMCs were used as a source of effector cells. Cytolytic activity was assayed at various E:T ratios.

FIG. 6

(a) Cytotoxicity of CD28⁻ and CD28⁺ CD8⁺ subsets. HLA-A*0201/pp65^495–503-specific CD8⁺ T lymphocytes were sorted directly from PBMCs of donor 13 on the basis of the presence or absence of CD28 expression. Bulk CD8⁺ populations of CD28⁻ and CD28⁺ CD8⁺ T cells were sorted by FACS from donors 16 and 17. Sorted populations were incubated with peptide-pulsed L-BCL targets, and cytotoxicity was assayed after 15 h. The E:T ratios were 0.5:1 for donor 13 and 2:1 for donors 16 and 17. The final peptide concentration was 5 μM. The percentage of specific lysis represents CTL lysis of pulsed targets − lysis of nonpulsed targets. (b) Cytotoxicity of CD57⁻ and CD57⁺ CD8⁺ subsets. PBMCs from donor 13 were stained with A*0201/pp65^495–503 tetrameric complex in conjunction with anti-CD8 and anti-CD57 MAbs and CD8⁺ tetramer-positive populations were sorted by FACS on the basis of the presence or absence of CD57. Sorted populations were incubated with peptide-pulsed L-BCL targets, and cytotoxicity was assayed after 15 h. The E:T ratio was 0.5:1.

FIG. 7

Staining of HLA-A*0201/pp65^495–503-specific CTLs from donor 13 with a MAb specific for TCRBV14. Freshly isolated PBMCs and 12-day-old CTL lines were costained with anti-CD8-Tricolor, anti-TCR-FITC, and PE-conjugated HLA-A*0201/pp65^495–503 tetramer. CD8⁺ lymphocytes were gated, and the proportions of tetramer-reactive cells expressing TCRBV14 both in fresh PBMCs (a) and the CTL line (b) were assessed. Control antibodies included anti-BV16 and anti-BV17 TCR MAbs.

FIG. 8

Cytolytic ability of sorted CD45RO^high and CD45RO^high tetramer-reactive cells propagated in vitro. PBMCs from donor 13 were stained with tetrameric complex, anti-CD8 MAb, and anti-CD45RA and anti-CD45RO antibodies, and populations of tetramer-reactive CD45RA^high and CD45RO^high cells were sorted by FACS and cultured in the presence of peptide-pulsed autologous L-BCLs. Their ability to lyse peptide-pulsed targets by a standard ⁵¹Cr-release assay was assessed following 14 days of culture. Both populations lysed target cells with similar efficiencies at E:T ratios of 10:1.

FIG. 9

Segregation of tetramer-reactive cells expressing low, medium, and high levels of CD45RA and CD45RO isoforms. PBMCs were triple stained with anti-CD8 MAb, tetrameric complex, and either CD45RA or CD45RO isoform antibodies. CD8^high populations were gated, and the percentages of tetramer-reactive cells expressing the CD45 isoforms were estimated. A typical dot plot of tetrameric staining versus CD45RA and CD45RO antigen expression from donor 17 is summarized.