High frequencies of naive Melan-A/MART-1-specific CD8(+) T cells in a large proportion of human histocompatibility leukocyte antigen (HLA)-A2 individuals

Abstract

Using fluorescent HLA-A*0201 tetramers containing the immunodominant Melan-A/MART-1 (Melan-A) tumor-associated antigen (Ag), we previously observed that metastatic lymph nodes of melanoma patients contain high numbers of Ag-experienced Melan-A-specific cytolytic T lymphocytes (CTLs). In this paper, we enumerated and characterized ex vivo Melan-A-specific cells in peripheral blood samples from both melanoma patients and healthy individuals. High frequencies (>/=1 in 2,500 CD8(+) T cells) of Melan-A-specific cells were found in 10 out of 13 patients, and, surprisingly, in 6 out of 10 healthy individuals. Virtually all Melan-A-specific cells from 6 out of 6 healthy individuals and from 7 out of 10 patients displayed a naive CD45RA(hi)/RO(-) phenotype, whereas variable proportions of Ag-experienced CD45RA(lo)/RO(+) Melan-A-specific cells were observed in the remaining 3 patients. In contrast, ex vivo influenza matrix-specific CTLs from all individuals exhibited a CD45RA(lo)/RO(+) memory phenotype as expected. Ag specificity of tetramer-sorted A2/Melan-A(+) cells from healthy individuals was confirmed after mitogen-driven expansion. Likewise, functional limiting dilution analysis and interferon gamma ELISPOT assays independently confirmed that most of the Melan-A-specific cells were not Ag experienced. Thus, it appears that high frequencies of naive Melan-A-specific CD8(+) T cells can be found in a large proportion of HLA-A*0201(+) individuals. Furthermore, as demonstrated for one patient followed over time, dramatic phenotype changes of circulating Melan-A-specific cells can occur in vivo.

Figure 1

Detection of ex vivo circulating A2/Melan-A⁺ and A2/Flu-MA⁺, but not A2/Tyrosinase⁺CD8⁺ T cells by flow cytometry, in a melanoma patient (LAU 233) and a healthy donor (HD 329). Highly homogeneous CD8⁺ lymphocyte populations (>98%) were obtained from PBMCs of patient LAU 233 or from healthy donor HD 329, by two rounds of positive selection with magnetic cell sorting. The lymphocyte preparations were then stained with A2/Melan-A, A2/Tyrosinase, or A2/Flu-MA tetramers together with anti-CD8^FITC mAb, and analyzed immediately by flow cytometry. The frequency of up to 0.01% of A2/tyrosinase⁺ cells was below the limit of detection (≥0.04% for A2/Melan-A; ≥0.02% for A2/Flu-MA).

Figure 2

Phenotypic analysis of ex vivo circulating A2/Melan-A⁺ and A2/Flu-MA⁺ cells in melanoma patients and healthy donors. CD8⁺ lymphocyte populations were highly purified (>98%) from PBMCs, as illustrated in Fig. 1. The lymphocyte preparations were then stained with A2/Melan-A, or A2/Flu-MA tetramers together with anti-CD45RA^CYC and either anti-CD45RO^FITC or anti-CD28^FITC mAbs, and immediately analyzed by flow cytometry. (A) Pattern of expression of CD45RA/RO (top) or CD45RA/CD28 (bottom) in total circulating CD8⁺ T cells (left) from healthy donors (shown is HD 329). They were CD28⁺CD45RA^hi/RO⁻ in A2/Melan-A⁺ gated cells (middle) but CD28⁺CD45RA^lo/RO⁺ in A2/Flu-MA⁺ gated cells (right). (B) Circulating A2/Melan-A⁺CD8⁺ T cells detected in the majority (7 out of 10) of melanoma patients presented a CD28⁺CD45RA^hi/RO⁻ phenotype (left). In contrast, tetramer⁺ cells from 2 out of 10 patients displayed variable proportions of CD45RA^hi/RO⁻ and CD45RA^lo/RO⁺tetramer⁺ cells (middle) and a CD28⁻CD45RA^int phenotype for 1 out of 10 patients (right). (C) Summary of phenotyping data obtained for melanoma patients and healthy donors. Frequencies of CD45RA^lo cells detected in gated A2/Melan-A⁺ and A2/Flu-MA⁺CD8⁺ T cells were calculated with CellQuest™ software.

Figure 2

Phenotypic analysis of ex vivo circulating A2/Melan-A⁺ and A2/Flu-MA⁺ cells in melanoma patients and healthy donors. CD8⁺ lymphocyte populations were highly purified (>98%) from PBMCs, as illustrated in Fig. 1. The lymphocyte preparations were then stained with A2/Melan-A, or A2/Flu-MA tetramers together with anti-CD45RA^CYC and either anti-CD45RO^FITC or anti-CD28^FITC mAbs, and immediately analyzed by flow cytometry. (A) Pattern of expression of CD45RA/RO (top) or CD45RA/CD28 (bottom) in total circulating CD8⁺ T cells (left) from healthy donors (shown is HD 329). They were CD28⁺CD45RA^hi/RO⁻ in A2/Melan-A⁺ gated cells (middle) but CD28⁺CD45RA^lo/RO⁺ in A2/Flu-MA⁺ gated cells (right). (B) Circulating A2/Melan-A⁺CD8⁺ T cells detected in the majority (7 out of 10) of melanoma patients presented a CD28⁺CD45RA^hi/RO⁻ phenotype (left). In contrast, tetramer⁺ cells from 2 out of 10 patients displayed variable proportions of CD45RA^hi/RO⁻ and CD45RA^lo/RO⁺tetramer⁺ cells (middle) and a CD28⁻CD45RA^int phenotype for 1 out of 10 patients (right). (C) Summary of phenotyping data obtained for melanoma patients and healthy donors. Frequencies of CD45RA^lo cells detected in gated A2/Melan-A⁺ and A2/Flu-MA⁺CD8⁺ T cells were calculated with CellQuest™ software.

Figure 2

Phenotypic analysis of ex vivo circulating A2/Melan-A⁺ and A2/Flu-MA⁺ cells in melanoma patients and healthy donors. CD8⁺ lymphocyte populations were highly purified (>98%) from PBMCs, as illustrated in Fig. 1. The lymphocyte preparations were then stained with A2/Melan-A, or A2/Flu-MA tetramers together with anti-CD45RA^CYC and either anti-CD45RO^FITC or anti-CD28^FITC mAbs, and immediately analyzed by flow cytometry. (A) Pattern of expression of CD45RA/RO (top) or CD45RA/CD28 (bottom) in total circulating CD8⁺ T cells (left) from healthy donors (shown is HD 329). They were CD28⁺CD45RA^hi/RO⁻ in A2/Melan-A⁺ gated cells (middle) but CD28⁺CD45RA^lo/RO⁺ in A2/Flu-MA⁺ gated cells (right). (B) Circulating A2/Melan-A⁺CD8⁺ T cells detected in the majority (7 out of 10) of melanoma patients presented a CD28⁺CD45RA^hi/RO⁻ phenotype (left). In contrast, tetramer⁺ cells from 2 out of 10 patients displayed variable proportions of CD45RA^hi/RO⁻ and CD45RA^lo/RO⁺tetramer⁺ cells (middle) and a CD28⁻CD45RA^int phenotype for 1 out of 10 patients (right). (C) Summary of phenotyping data obtained for melanoma patients and healthy donors. Frequencies of CD45RA^lo cells detected in gated A2/Melan-A⁺ and A2/Flu-MA⁺CD8⁺ T cells were calculated with CellQuest™ software.

Figure 5

Time course analysis of PBMCs from immunized melanoma patient LAU 132 shows dramatic phenotype changes of A2/Melan-A⁺CD8⁺ T cells. CD8⁺ lymphocyte populations from PBMCs of patient LAU 132 collected at different time points were highly purified (>98%), as illustrated in Fig. 1. The lymphocyte preparations were then stained with A2/Melan-A tetramers together with anti-CD45RA^CYC and anti-CD28^FITC mAbs and analyzed immediately by flow cytometry. (A) The patient received five immunization cycles, consisting of subcutaneous injections of Melan-A_26–35, Tyrosinase_1–9, Tyrosinase_368–376, gp100_280–288, and gp100_457–466 tumor-associated Ags, and influenza matrix Flu-MA_58–66 peptide. All cycles with the exception of the first included treatment with GM-CSF. CD45RA phenotype of A2/Melan-A⁺ cells was analyzed before the first immunization cycle (left), during immunizations (middle), and at the end of immunizations (right). (B) Percentage of A2/Melan-A⁺ cells in CD8⁺ T cells was followed over time. Each arrow represents an immunization cycle. (C) Proportions of CD28⁺ (○) and CD45RA^lo (•) in gated A2/Melan-A⁺ cells were followed over time.

Figure 3

IFN-γ ELISPOT assay confirms the naive status of most Melan-A–specific cells. Frequency of Flu-MA– (A) and Melan-A–specific (B) CTLs in CD8⁺ cells from 10 healthy donors and 11 melanoma patients (LAU 240 and 267 excepted) was measured by both IFN-γ ELISPOT assay (x-axis) and tetramer staining (y-axis), as described in Materials and Methods. The phenotypes of tetramer⁺ cells are as follows: CD28⁺CD45RA^hi (○), CD28⁺ CD45RA^lo (▪), 30–50% CD28⁺ CD45RA^lo (▴) and >60% CD28⁻ (♦). The bars indicate the lower detection limits for both techniques. These were 0.4 × 10⁻³ for A2/Melan-A, 0.1 × 10⁻³ for A2/Flu-MA (see footnote to Table for details), and 0.09 × 10⁻³ for IFN-γ ELISPOT (see Materials and Methods for details).

Figure 4

Functional activity of PBMCs sorted according to their tetramer staining phenotype correlates with Ag specificity. Ex vivo CD8⁺ PBMCs from a healthy donor (HD 604) were sorted into A2/Melan-A tetramer⁺ and tetramer⁻ populations. After 2 wk in the presence of irradiated autologous PBMCs and PHA, each cell fraction and a specific CTL line was tested for its lytic activity against Cr-labeled T2 target cells pulsed with different peptides: T2 cells alone (○), T2 + Melan-A 26-35 (•); and T2 + Melan-A 26-35 A27L (▪).